You are on page 1of 131

Basics for OC's Original Design

1/17/08 At this point it appears that OC would like a device with the following features

-13/8 magnet configuration meaning


-13 rotating stator magnets, that can be latched, equally spaced in a circle
(27.692 degrees apart), 4 inches apart center to center producing
a diameter of roughly 16.7 in.
-8 bar magnets, that can pivot, equally spaced on the outside of the rotor
(ultimately the rotor magnet design should incorporate lag - slow in, fast out)

Parts for Alsetalokin Replication

1 -5/8” thick Perspex acrylic sheet for baseplate 7 3/4” x 8 1/2 “


1 -3/4” thick HDPE sheet for stator (Delrin may work) 6” x 6 “
(please see notes below may be 18 mm thick)
3 -R834DIA magnets for the stators
8 -1/4” dia x 1/2” long bar magnets N35 polarized end-wise for the rotor
(please see posts below, repulsion distance test just a shade over 30mm)
(also note there is some debate on the actual length)
1 -.625 OD Delrin rod for stator magnet/bearing holders 1 foot
2 -.500 OD 6061-T6 aluminum alloy rod 2 “
2 - 4-40 nylon screw 1 1/2” long (or use two golf tees) to hold dampers in place
3 - 1/2” OD x 1/8”ID x 3/16” H ball bearings (see below)
3 - 4-40 SHCS stainless steel screws 3/4” long
3 - stainless steel washers to keep the stator assemblies off the baseplate

(Tip: Make plenty of stator assemblies (magnet, bearing, Delrin holder). Only 2 out of
15 of Al's actually have good AGW properties)

The bearing system for the rotor was originally a ¼ “ brass shaft that ran thru 2 – 1/2”OD
x 1/4”ID x ??? H flanged ball bearings
The bearing system presently being used is described somewhere in the “Posts from Al
about some of the details” section below.

For blueprints and photos please go to www.ospmm.org/whipmag

Posts from Al about some of the details

(Editors note: This following paragraph is what should be considered a rough sketch
of the device, more detailed info follows below)

A dimensioned sketch of the bearing/magnet holder assembly is posted on the other site.
The baseplate that I am using is acrylic plastic, 5/8 in thick, 7 3/4 in x 8 1/2 in, with a
1/4-20 threaded hole in the center for the rotor axle, and 13 evenly-spaced 4-40 threaded
holes for the stator magnet bearing holders, on a circle of radius 3 3/8 in centered on the
axle.
The rotor is 3/4 in thick piece of HDPE plastic, 2 7/8 in radius, with a 0.500 in center
hole bored all the way thru for the flanged axle bearings. Magnets are held in 8 evenly-
spaced 1/4 in wide slots, 7/16 in deep, inner edge of slots at the 2 1/2 in radius from
center.

This is the basic layout for the 13x8 version that I am experimenting with. But I am
mostly just using 3 stator magnets and a couple of "dummy" aluminum pieces of the
same size as the magnets, also mounted in the same type of bearing holder.

Stator magnet holders. The Delrin rod stock is 0.625" nominal OD. ID is bored to just
a rch under 0.5000", so the walls are 0.125"/2, or 1/16" nominal. Height 0.500". One
side bored deep enough for the magnet, other side bored deep enough for the bearing,
with a little ring of material left inside (3/8" center hole) to separate the parts and to
provide clearance for the mounting SHCS (and washer if needed).

The stator magnets are mounted coaxially with and above their bearings. The bearings
are screwed down to the baseplate with a tiny spacer for clearance. The magnets are
diametrically polarized.

The stator magnets are mounted by a single trimmed 4-40 SHCS and a little washer to
make sure only the inner bearing race is contacting the screw or the baseplate.

Stator bearings -
I had 8, they are Stewart-Warner, and I had to buy 5 more from another source, they are
unmarked.
Bearing is generic ball bearing, ferromagnetic, 0.500 x 0.1875 x 0.125
(Editor's note: If you look at the photos it is obvious the bearing is
0.500OD x 0.125ID x 0.1875H )

Mass of stator magnet assembly, including holder, magnet, and bearing:


10 grams. Variation between the 13 units is below the resolution of my (home) scale.

Rotor shaft and bearings-


These that I'm using now are for 3.5 mm shaft size, in a paired housing that slips into the
rotor's central bore, extending down to a shock-mounted base bearing unit.

Also I have changed the rotor bearing/shaft bit: now the rotor itself has no bearings, but
is a light press-fit onto a precision centerless-ground shaft that itself is mounted in a
holder with 2 very fine ball bearings; this holder with shaft and bearings stays attached
to the base when the rotor is removed.

Also I have changed the rotor bearing/shaft bit: now the rotor itself has no bearings,
but is a light press-fit onto a precision centerless-ground shaft that itself is mounted in
a holder with 2 very fine ball bearings; this holder with shaft and bearings stays
attached to the base when the rotor is removed

And here would be a good place to point out that I replaced the original 1/4-20 brass
screw and 2 flanged ball bearings pressed into the rotor, with a system that uses 2
smaller bearings, in a housing, with a 3.5 mm shaft, spring-loaded axially to keep the
bearings at zero end-play. This unit is shock-mounted on the baseplate, in lieu of the
brass screw, by some rubber grommets with screws. I scavenged it from a similar
motor to that Andy describes.

"a system that uses 2 smaller bearings, in a housing, with a 3.5 mm shaft, spring-loaded
axially to keep the bearings at zero end-play. This unit is shock-mounted on the
baseplate".

------------

I think I got it from an old tape deck capstan drive system. Reel-to-reel, remember
those?
I am afraid you will have to go a bit further back in time.

Or just make one. Make a flange, like a plumber's floor flange. Extend a tube up
from it, mount the bearings inside, stick a shaft in there, put a collar and a thrust
washer on the bottom extension of the shaft, put a washer, spring, and another
collar or circlip on the shaft just above the top bearing, screw the flange down
with grommets and screws, slide the rotor on to the shaft. You will have to detail
the ID of the rotor so it doesn't drag on the bearing tube.
from Jan. 12th post

The assembly to which you refer was scavenged from a capstan tape drive
system. I described how I would fabricate one from scratch if necessary, but for a
couple bucks at the surplus store I saved myself 4 or 5 hours work.

What I had in there before were just some generic flanged 1/4 inch bore unshielded
bearings, and the shaft was a 1/4-20 brass screw. Not the most accurate arrangement.
The screw in the baseplate and bearings in the rotor was sort of ok but it wobbled too
much for my taste.

The rotor is made of HDPE which is a bitch to machine, and cuts usually come out a
little small. I used a 0.250 2-fluted end mill, one pass in width, to cut the slots for the 1/4
in nominal OD rotor magnets, and they press into the slots perfectly--that is, a medium
press fit. So if you specify tolerances of +/- 0.0005, which is what I worked to on this
project, all should be well.
The stator magnet/bearing holders are Delrin, and it cuts small too, but not as much as
the HDPE. I cut for exactly 0.5000, and stopped when the mike read 0.4997. I wanted the
magnets to be pretty tight so they wouldn't slip, anticipating the latch requirement.
With the Sherlines, I can always achieve +/- 0.0005 accuracy--they have digital readouts
and are very accurate. If I am careful I can get down to 0.0002 in most materials.

Here's an actual set of measurements made right now. (Note: It seems these were
made at home)

Material HDPE
rotor diameter 144mm (5.6692”)
rotor thickness 18mm (.7087”)
distance from OUTER EDGE of rotor, on a radius, to INNER EDGE of magnet slots
13 mm (.5118”)
slots are symmetrically arrayed 1/4” nominal width, 12 mm depth (.4724”)
slots cut in a single pass with a 2-fluted end mill 0.250”

But I do not recommend using polycarbonate (Lexan )for any part of the machine--it's
in the wrong place on the triboelectric series, if one of my theories is correct.

Nor do I recommend acrylic plastic (Lucite, Plexiglas, Perspex?) for the rotor, if the
base is acrylic. Different materials should probably be used. I used HDPE which at the
opposite end of the plastic tribo series from acrylic.

Mass of rotor assembly, including magnets and shaft adapter, but not including shaft
or bearings:
258 grams. (Note: This appears to be the “plugged “ rotor with the modified
bearing system not the original design. This rotor is best visualized in the
Al's second video that ends with the trash bin. It has cuts underneath that would definitely
effect weight calculations)

Misc Note:

I have little data on variants. If the system works, as I suspect it will, with only the one
stator magnet, you might try that first to see if you can reproduce the antigearwise
motion, as a start.

The rotor magnets are press-fit in place.

I wouldn't rely on glue to hold magnets to plastic or other metal; things can spin fast
and you don't want a strong magnet coming loose at 5000 rpm. Guess how I know this
stuff.
Magnokinetic Judson dampers-

The dampers are 6061-T6 aluminum alloy. I believe the silicon and copper content of
this alloy contribute to its characteristics. They are fixed in place but must be adjusted
for optimum orientation, as they are machined from extruded heat-treated 6061 rod and
the properties may be slightly anisotropic.
The dampers are tuned by trial and error. If too close they retard the magnet too much
and it won't synch. Too far and there's no benefit. Once the distance is found, rotating a
little at a time and trying the spin until I found a "sweet spot" is how I did it. The effect of
the dampers is small, in any case.

Posts from Al about getting it started after it is assembled

First, the spacing between the stator magnets and the rotor is critical. With the layout
I've got, 5 mm is the magic dimension. At 6 mm, the magnets lose synch. At less than 4
mm, air resistance between the parts seems to inhibit the free rotation of the stator
magnet assembly. That's from the outside of the holder to the outer edge of the rotor (not
to the rotor magnets).

Second, not all magnet/bearing pairs will produce the effect. I have 15 of the magnets,
and 13 of the bearings. I have not tried all possible combinations of these, obviously. But
of the 13 assembled magnet/bearing pairs, 3 don't seem to work at all, 2 are fairly easy
to get started, and the rest are somewhere in between. Even the "easy" ones usually
require several tries before successful engagement. I have blisters on my fingers from
spinning stuff over and over and over.

Third, the relative height of the stator and rotor magnets seems important. I don't know
if it's because of my particular geometry or whether it is critical to the effect itself, but
the magnets shouldn't be in the same plane--that is, I am getting best results when the
rotor magnets are rotating in a plane that is just slightly higher than the tops of the
stator magnets, rather than down and centered vertically.

Fourth, I have been trying to get three stators spinning, but it turns out that 2 can spin
gear-wise on their own, and the third is finger-flipped antigearwise, and when it synchs
and the rpms go up, the other two stators can actually be stopped, and the rpms will go
up even more!

Just slight differences: the dampers on the real thing are mounted lower. It looks like
everything is on the same plane in the [simulated] model; in the real unit the plane of
the rotor magnets skims the top edge of the plane of the stator magnets, and the
damper plane skims the lower edge, if that makes sense."

Oh, I guess this would be a good time to mention that any metal or magnets lying
around on the work surface can inhibit the ability to synch the stator. I was having fits
trying to start it last night, until I noticed a little precision screwdriver was lying there, 6
inches away--too close. After moving it away, I was able to synch the stator easily. Well,
relatively easily.

On perseverance:
--It took me literally days of spinning and experimenting before I had a hint that
something interesting might be happening--the decreased rotor drag that I first noted
before the holidays. After that, more days of trial and error and serendipitous mistakes
before I even found the AGW rotation modes. By Accident! So don't get discouraged if
nothing seems to work--you may not have the right components or something, or it's not
put together right. Recall also that about half my magnet-bearing pairs are ineffective.
--I would guess that most people will get decreased drag and increased rundown times if
they get the magnets right. But extended run times like I saw are probably a result of
some EM interaction that's happening in my apartment and at work. For sure it isn't
anything like OU.

Posts from Al about other tips

First, safety. Safety First!


--do not leave magnets lying around loose, even for a few moments. They are easy to lose
control of, and hot flying pieces and pinch hazards are to be taken seriously. I use metal
jar lids as magnet trays.
--do NOT grind, burn, etc. these NdBFe magnets! The dust is toxic and otherwise
obnoxious, the fumes are toxic and smelly, etc.
--do not rely on adhesives! use structure to retain magnets against fields and forces.

Next, machining acrylic plastic and polycarbonate:


--Use WD-40 or clean kerosene (NOT GASOLINE OR PETROL!) as a cutting lubricant
to keep drills, taps and other tooling cool and cutting well. Beware of alcohols--acrylic
crazes and loses strength on exposure to alcohol--even the alcohol in a Sharpie marker
will craze the acrylic over time, and cracks will occur. Wash the WD-40 off with
dishsoap and warm water.
--plastic supply companies sell special drill bits with a sharper tip angle that won't crack
or pull though the plastic. They are well worth the price.
--I usually use bottoming taps even when starting, in plastic, I rarely need the plug tap to
start a tapped hole. The bottoming tap makes a better thread I think.

For those who are machining acrylic plastic: water is the best coolant/lubricant but
who wants to get water all over the tooling; many commercial cutting lubes will wind up
crazing the plastic; I use WD-40, it works great. Clean kerosene is good too. Wash
with warm water and dishsoap when you're done. Also alcohols like methanol or ethanol
will really craze and weaken the plastic, so be careful what you clean it with. Meguiar's
Mirror Glaze is the best, in 2 grades, polishing and cleaning.
-------------------------------------------------------------------------------------------
First Post on New Thread The OC Magnetic Perpetual Motion Machine
OC- Oct 4th 2007
(I decided this thing deserves a thread of its own)
New PMM configuration and animation - someone please tell me why it won't
work!.
My previous animation was limited to a 5x8 configuration mostly because that was as
much as I could geometrically visualize. I finally took the time to draw a 8x13
configuration and animate that as well. This looks much more promising. In fact, I don't
see how it can fail.

Here it is:
http://www.imgbolt.com/files/view/100653/OC13-fast.gif
http://www.imgbolt.com/files/view/100653/OC13-slow.gif

If you want a detailed analysis of what's happening, I also have an annotated image of the
first frame:

http://www.imgbolt.com/files/view/100653/OC13-Frame1.gif

Is there a sticky spot? Sure, there's a couple of them. But with all the other rotational
forces in this rig, I don't see anything to stop it.

If you want to model this using some more sophisticated graphical and/or engineering
software, go ahead ... and post some links to your work. Or just build the damn thing.

If anyone knows of any previous attempts to do this, I sure would appreciate some
references.

OC- Oct 4th 2007

Response to sticky spot


.......A sticky spot in my terms is anyplace where there is a resistance to favorable
movement. In this case, we have 2 forms of magnetic resistance in addition to such well-
known things like friction and back EMF. The flip consumes a bit of energy, but not as
much as most people might think. The major energy consumer is at the completion of the
flip, the 2 magnets are in an attractive state, side by side, and we need to separate them
such that they are in a repulsive position. This takes a lot of effort. Fortunately there are a
lot of other magnets in this configuration contributing to that effort.....

OC- Oct 4th 2007


Note: I found a "flipping" magnet machine at the "Museum of Unworkable Devices".
http://www.lhup.edu/~dsimanek/museum/newacqui.htm#flip

There are some similarities. But there are also some significant differences.

1) My machine does not need gears to flip the magnets


2) Once past the sticky point, my machine uses both repulsion and attraction to
provide the motional force instead of attraction alone.

enginerd- Oct 8th 2007


Nice graphic. Neat machine.

I don't see any reason that it should turn unless some outside system was rotating the
stator magnets

OC- Oct 8th 2007

@enginerd,
Try it. It'll surprise you.

OC- Oct 10th 2007

Back to the topic

For those of you who don't quite believe that the magnets will rotate on their own, I took
the time to set up a small demo experiment and photograph it. I created a linear
representation of a small segment of the stator and rotor illustrated above. Now this is not
anything that produces the overall effect, I don't currently have the proper magnets,
materials, or time to do that. This is just a simple demonstration that the rotating magnets
will flip polarity all on their own, without any gears or motors, etc.

http://imgbolt.com/files/100653/Exp1/StatorRotor.jpg shows the two components I


created. They are flattened out segments of the 13 magnet stator and 8 magnet rotor
previously illustrated.

Here's the construction materials I used to build it with


http://imgbolt.com/files/100653/Exp1/Pieces.jpg, plus 9 half-inch cube neo magnets (I
got careless and my 10th one shattered). The nonrotating magnets are actually 2 cubes
each. The rotating magnets are 1 cube each. I attached roofing nails to the sides of the
rotating magnets in order to kluge some sort of axle for them to rotate. Lots of tape and
glue here. Pretty shoddy work, but maybe it will get the idea across.

http://imgbolt.com/files/100653/Exp1/RotatingMagnets.gif
In this animated sequence taken with a digital still camera, I am moving the stator instead
of the rotor simply so you can see the magnetic relationships and the rotating magnets as
the position changes. I actually only took one sequence of shots (17 frames from right to
left) and then looped it backwards and forwards 10 times.

Nothing fancy, just a visual aid for those who don't think the magnets will rotate
themselves. Have fun.

edit: I almost forgot to mention. In the static photos, you will see some toothpicks stuck
into the rotating magnets. They are removed in the animation so you can see more
rotation. The purpose of the toothpicks is to stop the rotation when passing through the
sticky spot into a repulsive orientation. If I had kept the toothpicks in place, you would
have seen only 1/2 flip.

OC- Oct 11th 2007

alsetalokin:
WTF? You know I am in favor of your ideas, OC, but here I fear you tread
perilously close to Omni-bus's S-N-OT.

@Al, I only did this to visually illustrate that the magnets will rotate of their own
accord. There have been comments in this and other threads claiming it couldn't
be done without gears, servos, or cams. In attraction, the rotating magnets will
actually align and keep themselves at an optimum angle. As we pass the sticky
point, the magnets will want to flip back into attraction (as they do in this little
experiment). The stop levers and ramps are designed to prevent this from
occurring by locking the magnet into repulsion.

OC- Oct 11th 2007

alsetalokin:@OC, I know, just ribbing you a bit.


But in your design, wouldn't the magnets need to rotate into position,
relatively, just a bit before they do when you move the assembly by hand? I
mean, the driving rotation has to occur at a point in the cycle a bit earlier, than it
occurs when the magnets are being driven, in order to get past the sticktion, I
think. And doing it sooner is less favorable energetically...like pushing uphill
instead of coasting downhill...?

That's very possible and it's something I hadn't considered. Since I haven't built it yet, the
answer to that question is unknown (at least to me). I really wish someone with a
magnetics lab at their disposal would actually build this thing and do the measurements.

OC- Oct 12th 2007


Back to the topic again

I'm going to try and continue the discussion about my PMM here by contrasting it with
the unworkable device shown at:
http://www.lhup.edu/~dsimanek/museum/newacqui.htm#flip and point out some
additional faults with that system, beyond what has already described on its web page.

Both systems have something in common, rotating magnets which require effort to flip.
No debate here, this does consume energy.

Stop the animation with one of the rotating magnets midway between 2 stator magnets.
Note that the flip has already occurred, the gear is no longer engaged, and the rotating
magnet is now free to rotate (but it doesn't seem to in the animation). If we let it rotate, it
will orient itself to most favorable attraction with the next stator magnet. Unfortunately, it
also aligns itself to most favorable attraction to the previous stator magnet, which will
tend to resist further rotation.

In my concept, the rotating magnet is oriented so it is simultaneously repelled from


one stationary magnet and attracted to the next, always in the direction of rotation.
In this way, we gain the cumulative forces of 2 magnets in the desired direction
instead of opposing forces.

In the unworkable device, the stator magnets are oriented with poles facing towards the
center. The greatest magnetic forces will also be oriented in the same direction, not
towards an approaching rotor magnet. The greatest magnetic force will be felt when the
rotating magnet is near the center of the stator magnet, at the point where the gear
engages. Very little of the force is oriented in the direction the rotor is turning and will
tend to oppose magnet rotation very strongly.

In my concept, the "flipping" of the magnet takes place in the area between poles,
where there is less resistance to rotation. And the non-rotating magnets are oriented
in the direction of rotor movement. A larger portion of the magnetic forces will be
applied in the desired direction, utilizing both repulsion and attraction
simultaneously to assist with continued rotor movement.

Comments?

OC- Oct 16th 2007

trim:@OC
I admire your unselfish attitude, giving away your ideas for free like you do. I
wonder, if they ever proved to work whether you will regret not being able to
patent them and seeing Asian countries making billions from them?

I have given things away before, some of them quite successful. Nothing with the
potential this has. But it's just a persistent vision. I have no idea whether it will actually
work. If it does work, I'm sure there will be moments where I'll kick myself for giving it
all away. But overall, I can live with that.

I just wish someone would take it seriously enough to build it, or at least model it
using a capable simulation software.

OC- Nov 7th 2007

@Alsetalokin,

In regards to the idea I had for reducing some of the resistance when passing through
the sticky spot, I just put together a small animation showing how the relationships
between the magnets change as a rotor magnet passes the (rotating) stator magnet. This
requires some sort of spring on the rotor. I did not draw the spring, only the magnets
(springs are a lot harder to draw than circles and rectangles).

Anyway, if you want to look at the magnet orientations as the stator magnet interacts with
the rotor magnet, check out this animated GIF. Animation is 5 sec per frame to allow
time to read the text.
http://imgbolt.com/files/100653/MagnetRelationships.gif
Rotor magnet (rectangular) moving up past stator magnet (round)
You can get a "feel" for it by just using a couple cube magnets in your hands and follow
the path I indicate in the graphic, and allow one of the magnets to rotate as shown.

Al- Nov 7th 2007

OK, that's clear enough, but I still don't see how it could possibly work--I mean, yes, if
you get past the "sticky spot" sure, but that's the rub--?
I do like the idea of the springs allowing the angle of the rotor magnets to change--
that's an idea I haven't seen before.
I won't let my skepticism get in the way of genuinely trying to implement your idea,
though. I'll try to order some of the magnets you linked--what's the minimum number that
you think will be needed?
I'm still not sure about how to implement the rotation of the stator magnets--I just can't
quite visualize them rolling on their own, and any linkage I can imagine will be a
loser, energetically speaking...

OC- Nov 7th 2007

I think the optimum would be 13 rotating stator magnets with 8 spring-mounted rotor
magnets. But that's just from the picture in my head. It has to be an odd/even
combination, where there are an odd number of rotating magnets. I'm trying to put
together something with a 5/4 configuration at the moment, my first crude attempt to
validate the idea.

I'm still not sure about how to implement the rotation of the stator magnets--I just
can't quite visualize them rolling on their own, and any linkage I can imagine will
be a loser, energetically speaking...

They really do rotate all on their own. You saw my crude cardboard demo animation,
right? Try it out with a couple handheld magnets.

Al- Nov 8th 2007


I looked at the K&J Magnet site. Which of these would be most suitable for the rolling
"stator" magnets, do you think?
R834DIA
or
D48DIA

I can see some construction advantages for the central hole, maybe. I wish I could afford
to get bigger ones, but if we need 13 or so I'll have to get the smaller cheaper ones.
And given that constraint, which magnets would you choose for the "rotor" magnets?

(ETA the link to K&J http://www.kjmagnetics.com/search.asp)

OC- Nov 8th 2007


@Al,

Either would work, but I would pick (I did pick) the R834DIA for 2 reasons: tiny things
don't agree with my fat fingers, and having the hole in the disc simplifies mounting it so
it will rotate. I'm using 3/16" wooden dowel for a shaft and it's a good loose fit.

I also ordered some B442 (http://www.kjmagnetics.com/proddetail.asp?prod=B442),


B444
(http://www.kjmagnetics.com/proddetail.asp?prod=B444), and B448
(http://www.kjmagnetics.com/proddetail.asp?
prod=B448) for the rotor, so I could see what effect different rotor magnet lengths might
have on the stator magnet rotation.

It looks like I might have my rotor done sometime next week. Then I can start putting
together a stator. My initial impression for best rotation of 1/2" diameter stator
magnets, the rotor magnets should be between 1/2" and 3/4" long. This may change
with speed though.

Al- Nov 20th 2007


Yeah, I plan to try it both ways.
Didn't get a chance to order magnets today, too busy with other things, hopefully will get
a few moments tomorrow.
Do you think the rotating magnets should be on ball bearings? I can get some non-
magnetic ones from Small Parts Co. if they are needed.

OC- Nov 20th 2007


The easier it is for the magnets to rotate, the better. They will be experiencing some
fairly strong radial loads as well. If you use bearings, make sure they can handle the
varying load. Nylon roller bearings might be pretty good. I definitely wouldn't use jewel
bearings. Teflon bushings might work pretty well.

Al- Nov 29th 2007


Still waiting for magnets to arrive (I ordered 15 of the R834DIA
http://www.kjmagnetics.com/prodimages/R834DIA.jpg), but I made a few "mock-
magnets" out of aluminum so I could start making the rotor and stator plastic pieces...will
be using slightly turned-down #8 plastic screws to hold these magnets to the stator
baseplate...

Al- Dec 5th 2007


Magnets arrived today--15 ea. R834DIA diametrically magnetized N42 ring
magnets.
These are really neat little magnets--the diametric magnetization makes them fun to play
with because they stack in all kinds of interesting ways. Also got some of the magnetic
field indicating film.
Next is to use the black magic marker to mark each magnet, by coloring one half
according to its polarity.

Al- Dec 6th 2007


Progress report on the OC MPMM:
Today I made a "jig", not really a motor yet, but a circle of 10 mounting holes for the
834s on a perspex baseplate, with a central mounting for a rotor of Delrin running on a
couple little flanged precision ball bearings.
The 834s are mounted on the circle of 10 nylon 8-32 screws threaded from the
bottom up thru the baseplate, and secured with nylon nuts, loosely so the magnets
can rotate freely. I haven't put any magnets in the Delrin central rotor yet, but it is
mounted on its bearings already. The central rotor is about 2 inches in diameter, a little
small for a reason, and I will be making a larger one as well.
I also examined the magnets with the field viewing film and marked them to show their
polarization and to make it easier to see them when they move.
I have already discovered a lot of interesting behaviors with this little rig. Nothing too
encouraging, I'm afraid, but pretty cool nevertheless.
Tomorrow I will most probably be able to make and post a short video on YouTube
showing what I've done so far, so that OC et al. can criticize and suggest, as appropriate.

Al- Dec 7th 2007


OK, I posted a video of the "test jig" on YouTube. This is not intended to "run"! I just
wanted to experiment with the magnets to get an idea how they would be behaving. Note
especially that the 834s do rotate like OC predicted, if the magnets on the rotor are
not too far away.
I can almost get the rotor to spin all the way around, by "driving" just one of the 834s by
hand.
http://www.youtube.com/watch?v=hP0QEj4BgiE

OC- Dec 7th 2007


1) The rotor magnets need to be oriented tangent to the rotor disc such that the
fields are parallel and opposite to the stator magnets when centered in closest
proximity. This allows for a balanced response to attractive forces on the way in
and repulsive forces on the way out. If the rotor magnets are oriented properly, the
stator magnets should rotate better even with larger gaps.
2) The rotor magnets in your video seem a bit large. Are the poles on the large
faces? I seem to get the best effects when the length (distance between poles) is
equal to or slightly greater than the diameter of the diametrically magnetized
magnets. My rotor magnets are http://www.kjmagnetics.com/proddetail.asp?
prod=B448 and I have tried extending them to 3/4" with 1/4" cubes.

Have you figured out a good latching mechanism to prevent the 834s from flipping back
to attraction once a repulsive force is applied on the way out (as the rotor magnet passes
the closest point)?

Al- Dec 7th 2007


Umm, I don't quite follow your Point 1--any magnets mounted on the rotor will either
have to be exactly radial (one pole toward center, other pole out, axis on a radius) or on a
tangent to some diameter of the rotor. Do you mean that, when the rotor magnets are at
their closest approach to the stator magnets, they should be oriented tangentially at that
point, that is, at 90 degrees to a line connecting the center of the rotor with the stator
magnet? "Parallel and opposite to the stator magnets"--I don't get this phrase. Can you
explain, or maybe draw a picture? Poles on the ends, or on the big faces?

As to Point 2--yes, they are just some ceramic magnets that I had lying about, I think they
are cabinet latch magnets. They are polarized on the large faces. I intend to use bar
magnets polarized end-wise, as soon as I can get some of the right strength and size. The
only bar magnets I could find today in the lab stash appear to be N52 grade--I couldn't
even pull them apart by hand, even though they were individually wrapped in foam. Too
strong!! I guess I'll be placing another order to KJ on Monday.

I'm going to make little mounts for the 834s that will incorporate a circle of holes,
that can hold little pins, that will provide an adjustable place for an escapement pawl to
latch. The problem here is firmly attaching something to the magnet that will be strong
enough to withstand the torques likely to be generated, yet small and lightweight. I have
some high-strength epoxy that might work. But I'm definitely taking the weekend off, so
it will be Monday or Tuesday before I'm back at it.

OC- Dec 7th 2007


Yes. In the little diagram above, the 834 magnet has already rotated almost 180 degrees
as the rotor magnet approached. This would be its stable position at this point if no
latching was performed. The 834 should actually be latched a few degrees before this
point, so the axes will actually never be quite parallel.

Al- Dec 13th 2007


I can see that the rotation of the stator magnets is extracting energy from the
rotation of the rotor, and so I am going to mount all the rotating magnets on ball-
bearings. That means I need to go to the LHS (local hobby shop) to get a handful of tiny
ball-bearings--RC helicopter spare parts.
Jaro K. tells me that I should only use one, or perhaps two, magnets in the rotor, until I
get the spacing and positioning roughly tuned.

Al- Dec 13th 2007


OK, now that I understand my polarity mistake and have corrected it--
I have to say this:

I am amazed. Overconfident's design comes closer to "working" than any magnet


motor design I have yet encountered (and I have seen a few, believe me.)
The little test jig I built is already pretty amazing, even without the mechanism for
latching the stator magnets.

Not that I have been converted, you understand--but if I drop out of sight and turn up
slumped over in an airport parking lot, you'll know what happened.

OC- Dec 13th 2007


You ain't seen nuthin yet. Wait until you get the latches in place on the stator magnets
and add the spring mechanism to the rotor magnets, then scale up the number of magnets
to 5x8 or 8x13 (5x8 was what I saw in my dream, but 8x13 looks even better once I took
the time to draw it).
OC- Dec 13th 2007
alsetalokin:Hmm--it would seem that the alternating rotor polarity would require there to
be an even number of magnets in the rotor?
Whereas the chained polarity could be done with odd or even numbers?
(alternating= n-s s-n n-s s-n etc. around the circle
chained= n-s n-s n-s n-s etc. around the circle)

You are welcome to try the chained polarity BUT I think the alternating polarity is
the only way there is a chance for it to work. As far as putting an ODD number of
magnets on the rotor, just take what you have now, hold the rotor still and spin the
stator ... voila, you now have a rotor with an ODD number of magnets.

Al- Dec 13th 2007


OK, good. For a minute there I thought I was totally lost!
So for the alternating polarity we agree that there must be an even number of magnets.
It turns out that I am really busy in the lab today so I may not get a chance to do any real
work.

(ETA did you see the note I put with the pics about the differences in the rotating magnet
rotations? I have confirmed--the rotating magnets rotate 2 full turns for each full
revolution of the rotor in the alternating config, and the rotating magnets rotate 4
full turns for each revolution of the rotor in the chained config.)

Al- Dec 14th 2007


27.692307692307......

which means that even I can't maintain that accuracy...


if I use 27.7, that means 0.1 degree of accumulated error around the circle, so 12 gaps
will be 27.7 degrees
and 1 will be 27.6 degrees.

Al- Dec 14th 2007


I began laying out the 13x8 baseplate today, and I have someone out buying more ball
bearings. This one will have a 6" diameter rotor.
Also I explained a bit of what was going on to the boss, who also has an interest in these
matters. He laughed, but it was very hard to pull him away from the Test Jig #2 assembly.
He was fascinated...

I made the 13x8 baseplate, laid out the hole pattern, drilled and tapped the holes, attached
the LRFs (little rubber feets). Also cut and trued a blank for the rotor. I didn't have any
Delrin of the right size so I cut the rotor blank from some UHMW plastic, or maybe it's
HDPE, but anyway it's nasty stuff, not nearly as nice to machine as Delrin. But I got it
done nevertheless.

OC- Dec 14th 2007


@Al,

Latching should occur to stop stator magnet rotation just before the 2 magnetic fields
become
parallel/opposite

----> <----

For the 5-magnet stator, maybe about 15 to 20 degrees before the magnets align. For
the 13-magnet stator, about 5 to 10 degrees before alignment. Sorry I can't give better
figures. The tolerances on my rig aren't very precise.

Latches should release when the approaching rotor magnet (in attraction) and the
receding rotor magnet (in repulsion) are equidistant from the respective poles of the
stator magnet. I'm afraid I haven't given much thought to this part yet. I'm just using
toothpicks and my finger to stop rotation and I have to remove them each time in order to
make more headway. I have considered some sort of magnetic triplever that releases the
latch when the rotor reaches a certain position, but haven't really done any experiments
with that concept. For initial testing some small solenoid latches could be used (doesn't
provide any real power to the mechanism, just latches at the appropriate point).

Al- Dec 14th 2007


re latching: I think I see where you intend the latch to occur, but wait till I have a chance
to post photos
showing the rotation of the various parts, with some reference marks, so you can see
exactly where, on this jig.
I was thinking of using some notches or teeth machined into the magnet housings, a
spring loaded pawl to do the latching, and a cam or "bumper" on the rotor itself to
release the pawl. That way there's only the mechanical motion of the machine doing the
work. But I like the solenoid idea too, it might make testing easier, as you say.
It's fascinating to put the field-viewing film over the assembly and watch the interplay of
the fields as everything is rotating.

OC- Dec 14th 2007


One thing I just noticed, the spacing for the stator magnets seems closer than it was on
the 4x5 rig. I hope it's not too close. There needs to be approximately twice the rotor
magnet length, maybe more, between the rims of the stator magnets.

We might need to go for a larger rotor. Oh well, let's see how it behaves

Al- Dec 15th 2007


Hmmm-yes, I see what you mean. I think there might wind up being about 3/2 length
between, now. Might be a little small.

OC- Dec 15th 2007


It needs some space to accelerate and gain momentum between stator magnets. You
are welcome to try this one, but I don't think it will be optimum. I can't give exact figures
because I haven't got that far myself. It will take some trial and error (or a good
simulaton) to determine what works best.

Al- Dec 15th 2007


re the size and spacing of the 13x8 model: I left it at the lab, and we are snowed in right
now, but >>

I estimate (from the photo) that the holes are 4.2 cm apart. The stator magnets are 0.6
cm radius. So that means they will be 3 cm apart, edge-to-edge. These rotor magnets (in
the 5x4 jig) are about 1.35 cm long. So that should be good--yes?

OC- Dec 16th 2007


Al,
I would consider that to be an absolute minimum. Go ahead and try it. You may want
to experiment a bit. Just set up a couple stator magnets and use an adjustable length
rotating arm with a single rotor magnet on it. Latch the first rotating magnet into place
and observe how the rotor magnet behaves with different radius and distance between
stator magnets. I seem to get (subjectively, I'm just going by feel) best response
(strongest forces, most gain of momentum) between 2 and 3" center-to-center, using
1/2" long rectangular neos on the rotor.

OC- Dec 16th 2007


@Al,
Since we may have some idle time to discuss things, due to the blizzard, let me take the
opportunity to do a quick review of where we are going and to throw another wrench into
the works. I don't know if this will be necessary, but it's possible it could add some
additional impetus.

We have discussed the required field orientations (latching of rotating magnets) and
the minimization of resistance in repulsion (hinged, spring loaded stator magnets).
There is one more topic I have not really covered. I briefly mentioned to Cloud Camper
at one point that I also had a way to introduce magnetic viscosity into the equation. I
have no idea whether it will make any difference, but there IS a way to add that as well.
Unfortunately, it will most likely require some major redesign of the rotor.

If we use a rotor that has spokelike disc instead of a simple, round disc, where each
spoke/arm has a magnet mounted at one end and is either hinged at the hub or has a
certain amount of flex in it, we can have some dwell time at the point where each magnet
provides maximum positive H to the other and reduce the time where the fields are in
opposition (maximum negative H).
1) As the rotor magnet approaches the stator magnet the arm will tend towards it,
compressing the spring in one direction, allowing the rotor magnet to get closer, sooner
than it would otherwise, leading the rest of the rotor and maximizing the time spent in the
attractive field. (slow in)

2) When the magnets are in closest proximity, the spring will relax and allow the rotor
magnet to stay close until the spring compresses in the other direction to a critical point,
lagging a bit. Then the combined forces of the rotating rotor and the spring action will
cause the arm to "jump" quickly past the sticky spot, minimizing the time spent in the
repulsive field. (fast out)

Current rotor

___
/
/
__ /_
\ |o| o
o | | |
__/ |_|zzz
^ \
\
\___

Proposed rotor for leveraging magnetic viscosity

z
z
__ _ z
\ |o|==========+===========o
o | | | z
__/ |_|zzz z
^ z

o = a shaft, axle, or pivot point


^ = pawl
zzz = spring
rotation is clockwise

The viscous effects can be enhanced (slowed) by adding some small soft iron discs or
washers to the ends of the rotor magnets.

OC- Dec 20th 2007


A good place to start would be with the pole facing directly towards the next stator
magnet on the repulsive side, then rotate a bit away from the rotor to allow the opposite
side to have a bit more effective attraction for the approaching rotor magnet. Each stator
magnet will be simultaneously repelling one pole and attracting on the other, so both
sides need to be considered. Once you establish a good response on the repulsion side,
rotate a bit away from the rotor to favor the attractive side a bit.

Then rotate 180 degrees and latch again with opposite polarity for the next passing
rotor magnet.

Some way to adjust the latches might be nice. If things start spinning really fast , we
might want to retune things a bit.

Latch release should occur when the rotor magnet is somewhere around midpoint
between stator magnets, maybe slightly before.

Al- Dec 21th 2007


I am already able to generate the kind of data with this rig, that makes people
believe that PMMs might be possible.
To wit: rotor drag is inversely related to rotor RPM, at least over a limited speed range
in preliminary testing.

Let that sink in for a while.

Al- Dec 22th 2007


I have described how the rotating magnets move in an oscillatory motion, like a washing
machine agitator, as they are rotating. I believe the "knee" shown in the data graph occurs
as the rotation of the stator magnets becomes so fast that they no longer "agitate" but
rather rotate more or less smoothly all the way around. The acceleration of the rotor must
be smooth and gradual so the stator magnets can stay locked in phase with the rotor
magnets.
To be sure of this I will have to view the unit under the StroboTach, which, of course, is
at the lab.

I'm sure this behavior will change when latches are implemented...

OC- Dec 22th 2007


Remember when I said the stator magnets looked too close together? I said then that I
thought they should be spaced 2 - 3" apart. I just sat down and did some experiments
using a couple of my "834" magnets and a 1/4" x 1/2" bar to see what the optimum stator
magnet spacing would be. I come up with about 4" center-to-center for an optimum
distance. This is far enough apart that stator magnets have a minimum influence on each
other and close enough for them to provide a significant continuous acceleration and
increase in momentum in repulsion and attraction from one stator magnet to the next. 4"
is considerably larger than what we have now. Looks like we will need a 17 or 18"
diameter rotor to get what I would consider optimum spacing.

One other other thing I noticed is the gap between rotor and stator magnets. If we can
cut this down to half, we'll be able to increase the available forces by 4x when magnets
are in closest proximity.

Al- Dec 23th 2007


Hmmm---bigger is not good for a number of reasons, such as cost of materials, stability
and balance, portability, and so forth, at this stage. It would be better to get smaller and/or
weaker magnets, if it comes to that.

When I incorporate the pivoting feature on the rotor magnets, the clearance to the
stator magnets will be smaller.

The smaller rotor for the 5x4 version behaves interestingly when all its slots are filled
with the flat magnets, using mutual repulsion to shape the fields a bit asymmetrically.

I forgot to mention that JK, the real magnet motor specialist in our lab, has been looking
over my shoulder with interest, and chuckling under his breath in some Slavonic dialect.
(He's also the one who believes the Earth is hollow.) He has pointed out that mutual
repulsion is stronger than mutual attraction at comparable distances. And of course most
magnet motors attempt to make use of this fact in some manner

OC- Dec 23th 2007


@Al, the vision I had in my head when I woke up that morning was of a 5x8
configuration, as illustrated in one of my earlier animations. Afterwards, I sat down with
pencil and paper and sketched out some other configurations. The only arrangements that
"felt right" were 3x2, 5x8, and 13x8. When I thought about it a bit I realized there was a
pattern here. They were all Fibonacci odd/even pairs, and the odd numbers were all
prime. I guess I may have let that, possibly irrelevant, insight influence my concept.

The only hard requirement is that one side must have an even number of magnets in order
to provide the alternating magnetic fields. The other side must have an odd number of
magnets AND there must not be a common factor between the two sides in order to
prevent more than one rotor/stator magnet being in the sticky spot at any given moment
(6x9 would definitely be a bad choice

OC- Dec 27th 2007


@Debaser,

I'm not familiar with the capabilities of the ANSYS software. Can it simulate the
magnetic forces and field vectors? Is it possible to create an animation from the
simulation that is viewable without requiring ANSYS software?
Alsetalokin has some images at:
http://freeenergytrackers.ning.com/photo/photo/listForContributor?
screenName=3pgkrtdln76us. The latest ones are of a 6" rotor with some 1/4" x 1/2" neos,
I believe they are N42s. The rotating stator magnets are 1/2" dia x 1/4" N42s
(http://www.kjmagnetics.com/proddetail.asp?prod=R834DIA).

A simulation of the current state of construction should already be fairly interesting. Two
additional features are being implemented:

1) latches to hold stator magnets into an opposing orientation while the rotor magnet
travels from close proximity to midpoint between stator magnets,

2) pivot points near the leading end of the rotor magnets to allow them to pivot away
from the stator magnets and more easily overcome the repulsive resistance when passing
through into a repulsive orientation.

Except possibly for the bearings, I don't believe any other ferromagnetic materials are in
use. We can probably ignore them for the time being.

If you need more exacting figures, I'm sure Al can provide them, but you should be able
to come pretty close just by studying his photos.

It would be interesting to compare the fields shown by ANSYS with those Al has seen
with viewing film, even at this stage.

Al- Dec 27th 2007


Well, I think it's jumping the gun a bit, but here are some specs on the 13x8 test jig, so
that Debaser can play around. Bearing in mind that there will be additions to the design
as OC has pointed out.

Magnets:
Stator Magnets: 13 ea. K&J p/n 834DIA N42 grade, ring, 0.5 in dia, 0.25 in thick, ~3/16
central hole, diametrically polarized.
# Surface Field: 5340 Gauss
# Brmax: 13,200 Gauss
# BHmax: 42 MGOe

Rotor magnets: 8 ea. generic N35 grade, bar, 0.25 in dia, 0.5 in length, axially polarized.
#BHmax: ~35 MGOe

Stator magnet spacing: on a circle, radius 3.375 in, spaced 1.6 in apart around the circle
(or every 27.7 degrees)

Stator magnets each mounted on (above) single ball bearing, 1/2 x 3/16 x 1/8
Rotor magnets equally spaced in rotor, with centerline (axis) of magnets 2.5 in from
center.

Rotor on two ball bearings at center axle.

riterX- Dec 27th 2007


@Al
Can you please give more detail on the stator magnet bearing/magnet holders and the RC
helicopter tiny ball bearings being used with it?

Al- Dec 27th 2007


Actually I didn't use the heli bearings there, I had some others at hand that were easier
(read cheaper) to come by. The heli bearings might wind up being used in the latches
and/or the rotor magnet pivots, but I may use bronze bushings instead.
Sometime later today I'll put a photo of a dimensioned drawing in the usual place. But
here's a verbal description (dimensions in inches, +/- 0.0005):

A Delrin cylinder 0.625 OD x 0.500 long, with axial 0.375 dia hole bored all the way
through.
Then one end bored out to 0.500 dia x 0.250 deep, to receive magnet.
Then other end bored out to 0.500 dia x 0.175 deep, to receive bearing.
Magnet and bearing light press-fit into the cylinder.

Magnet is K&J 834DIA, 0.500 x 0.250 x 0.1875


Bearing is generic ball bearing, ferromagnetic, 0.500 x 0.1875 x 0.125

Assembly is mounted to baseplate with a SS #4-40 SHCS with the head trimmed for
clearance, magnet up, with a little SS washer between the bearing inner race and the
baseplate, for clearance and stability.

I will be making some ratchet-type teeth in the OD of the Delrin cylinders for the latches
to grab. I think--I haven't tried this idea yet.

Al- Dec 30th 2007


Now, I agree that latches and holding the magnets back from where they want to go and
so forth must subtract energy from the system, no matter how smooth and precise the
mechanism can be made (I'm no robot Seiko watchmaker, after all).
I never liked the idea of latches anyway.

"Latches? LATCHES? We don' got to show you no stinkin' LATCHES!"


(Treasure of the Sierra Magnet,1948, dir John Hustler, starring Bogus Humphart)

I have described earlier how the stator magnets rotate as if geared to the rotor, but in a
slightly agitated motion, and in the asymmetrical 7x8 variation that I have been playing
with (and AFAIK in all variants with 8 rotor magnets) the stator magnets rotate twice for
each rotor revolution. In opposite directions. That is, rotor goes clockwise, stator magnet
goes counterclockwise.
Still with me?
Now, I ask you to imagine the stator magnet rotating in the same direction as the rotor.
Both going clockwise, very non-gear-like. Now, if things are turning fast enough, we
might be able to get into a situation where the timing sort of reverses, or something, and
extremely interesting things begin to happen, or at least become possible.
So, now I need to ask for someone with the interest to do some math.
I have already given the relevant dimensions in an earlier post. So, hypothetically
speaking if I were to get the rotor spinning at, say, 400 rpm by spinning it by hand; then
quickly spin a stator magnet in the same direction by flipping it with a finger, could there
be some resonant alignment of the fields that would make it "catch hold" and continue
spinning at 1600 rpm (4 x rotor rpm), both rotor and stator magnet going clockwise?
Now suppose this relationship continued locked in phase as the assembly gradually
accelerates until reaching a maximum measured rpm of 1904 rpm for the rotor, and 7633
rpm for the stator magnet, both rotating clockwise.
The math question is this: how long are the magnets in proximity, or how fast are they
passing each other, at these rotation rates?
Are we going fast enough to possibly be seeing the effects of the famous Sv?

Al- Dec 30th 2007


Uhh--I still get 8 milliseconds /rev.

7600 rpm / 60 = 126.7 revs per second = 0.0079 secs per rev or about 8 milliseconds

gee math is hard this time of night...

rX- Dec 30th 2007


alsetalokin:The slowest they will remain in "opposite synch" seems to be around 1160
rpm (or about 290 rpm for the rotor). As the system brakes from a high-rpm run, the
stator magnet loses synch and either stops dead or starts rotating in the "natural" or
gear-like direction, but now at 2 x rotor rpm rather than 4x as before. Once this happens
the rotor can no longer sustain rapid rotation and coasts to a stop.
I can start the thing by spinning the rotor to 300-400 rpm (that's as fast as I seem to be
able to get it going by hand) and then flipping a stator magnet. It usually takes 5-10
tries, and several re-spins of the rotor, before it "catches" and the magnets go into
"opposite synch" or whatever.

.....Now suppose this relationship continued locked in phase as the assembly gradually
accelerates until reaching a maximum measured rpm of 1904 rpm for the rotor, and
7633 rpm for the stator magnet, both rotating clockwise.....

It's pretty scary when the thing gets up to max rpm. The little stator magnet really
whiizzes at 7600 rpm. The first time I wasn't sure if it would stay together, or punch a
hole in the ceiling, or what.
I got pretty excited, I must admit.

Hypothetically speaking, of course.

So you spin the rotor to 300-400 rpm (as fast as you can get it going by hand) then
reverse the spin on a stator magnet and it reverses the spin on all the stator magnets.

Then you gradually accelerate the rotor with a dremel tool?


What determines the maximum rpm of the rotor and stator (1904 and 7633 as mentioned
above)?

Hypothetically speaking, of course.

Al- Dec 31th 2007


@rX: No, you have to reverse-spin each stator magnet individually by hand, and it
can be hard to get them all going. A little bearing drag causes the magnet to refuse to
synch in the reverse direction, or to drop out before reaching incredible speeds. Or
maybe it's variation in the individual magnets. I found one last night that would go 8030
rpm, and a couple that I couldn't get to synch at all.
Mostly I am working with one at a time, and the others spin gear-wise for a while then
give up and stop as the rotor spins up.
I'm not using the dremel. The spin rpm tops out, I think, because of air resistance and
bearing friction. The slightest touch on the spinning stator magnet will cause it to "drop
out" and decelerate to a stop incredibly fast, like from 8000 rpm to 0 in about 5
revolutions. At that point the rotor is freewheeling and slowing down, and because of its
inertia and good bearings, that takes a couple minutes to get from 2k or so down to the
300-400 rpm range where I can re-spin the stator magnet to get it to re-synch. And that
takes, as I said, 5-10 tries usually before it "catches" and then the cycle can repeat.
In my dreams, of course.

@OC, believe me I have done what you describe, in many possible variations, for hours
on end. That's why I built these gadgets in the first place. I've even taken them out on the
subway and the bus a few times.
Nothing I saw or felt would have enabled me to predict this behavior of spinning non-
gear-like and causing such rapid rotation. I expected, if anything, a rather stately motion
like a miniature Ferris wheel or something, but this phenomenon is in a completely
different class altogether.
I think it's time to abandon the idea of mechanical latches. We seem to have progressed
beyond that point, to where electronic latching will be needed. Maybe also some kind of
field shaping coils a la Helmholz.

(ETA I discovered the reverse-spin phenomenon completely by accident--I was


trying to see if I could get the magnet to re-engage in the gear-wise direction, flipping it
over and over, until it finally took off and scared the crap out of me. I even repeated the
thing several times before I realized I had been flipping the magnet the wrong direction
and when it engaged it was spinning backwards, ie clockwise like the rotor.)

OC- Dec 31th 2007


(I wrote the following before I wrote the preceding, go figure...)

Well, it's clear that I need more data. As you may know, for the last week or so I have
been in my basement apartment, experimenting with a few tools and the 13x8 (or
7Ax8) test jig. I have a laser tachometer and a DMM and a few wrenches and
screwdrivers and that's about it. Of course at the lab I have a bazillion bucks worth of
exotic test equipment, picosecond oscilloscopes, screen rooms, function generators,
machine tools, ceramics lab, you name it, I've pretty much got it at hand. But I haven't
been out there since before Christmas, and won't be working again until Wednesday.
So understand that recent reports are not even preliminary, they are more like
rudimentary, like from Teej's planet or something.

As I said, I am in a basement apartment. My workbench is in the tiny kitchen, a slice of


plywood up against a side (exterior underground) wall. The house's main electrical
distribution panel is embedded in this wall. I just realized this over morning coffee (I
haven't been sleeping very well the past couple nights...)

So. recalling RB's great vcr-head PMM, and some of the Steornish theorizing, I realize I
need to dash out to the lab, I guess, and pick up a couple easily portable EM field testing
meters and gadgets, to make sure that the damn electrical panel isn't causing the
hypothetical effects that I hypothetically could be seeing.

Wednesday I'll truck the whole affair out to the lab, and if I still have a job, I'll try to
sneak some more time to do some vids and of course more testing.

(Gee, as I recall we have been using the big screen room for storage, there are a few big
HV devices in there, sparky, EMP-type things, a Podkletnov anti-gravity beam generator,
a couple reverse-engineered ufo propulsors, etc...have to move all that stuff out for
testing...boss thinks I'm nuts...because Permanent Magnet Motors Can't Possibly Work
(PMMCPW!))

Oh, and Happy New Year to all!

(ETA there's always a typo somwhere)

OC- Dec 31th 2007


@Al,
I disagree. I think we should methodically explore every aspect of the thing we can and
document the results as well as we can. That's part of the reason I started soliciting for
someone to maintain documentation. Mechanical or electrically actuated latches (with
relay or solenoid far enough away that they don't affect the magnetic interactions) are
something that need to be investigated and I believe will be more important when the
stators are separated more (larger rotor). Right now, you are not seeing the necessity
because of the close stator spacing and the interaction of the stator magnets on each other.

The effects with smaller rotors, closer stators, and higher speeds were things I hadn't
really considered when I put this concept together. After hearing some of your results, I
have decided this needs to be investigated as well. There are likely some very interesting
things we can learn here. I believe the most likely anomalies we will see by going this
direction will be from magnetic reconnection phenomena rather than magnetic viscosity.

But I don't want to neglect the original idea. In fact, I think the original idea is where our
initial efforts should be concentrated. Let's see what happens there first. I really think a
larger rotor will be needed to (in)validate my original concept.

Too bad we can't seem to drum up more replicators to help with all the possible
variations. Actually, if we could post a couple videos of things you have already done,
I'm sure we could stir up some interest. A video showing the fields in motion, using the
field viewing film would be interesting. So would this latest behavior you are describing.

But we really need documentation and some interesting results!

These effects you have described here, which configuration are you seeing this with? Is it
the 7x8 or 13x8? In either case, I think the first change that should be made is the
addition of pivots for the rotor magnets. For a rotor of this size, spinning at say 50 rpm or
more, we shouldn't even need any springs, just allow the magnets to pivot and swing the
trailing end away, if and when it wants to. I would think it would be self-regulating above
a certain speed (speed gets too high, centrifugal force is greater than can be overcome by
the repulsive resistance which slows the rotor down until the magnets can pivot again to
reduce resistance and allow the rotor to speed up again).

Rx- Dec 31th 2007


@Al
Very fascinating results!
This phenomenon is, as you said, quite unexpected.
I can see why the "reversed" stator magnet speeding up would really scare you and why
you are not sleeping so well at night.
When you have the rotor spinning, how many of the stator magnets have you been able to
get into "reverse" spin?
Does the rotor lose rpm (spin down) slower with more "reversed" stator magents
engaged?

Sounds like history in the making.


Keep up the good work!

P.S. I also want to confirm you are saying the rotor is accelerating from say 400 rpm up
to 1160 rpm by only reversing the spin of one stator magnet.

Al- Jan 1st 2007

@OC:
Q: These effects you have described here, which configuration are you seeing this with?
Is it the 7x8 or 13x8?
A: 7Ax8

@rX:
Q: "When you have the rotor spinning, how many of the stator magnets have you been
able to get into "reverse" spin?"

A: Three is the most at once and that was really hard and only one kept in synch for
any appreciable time.

Q: "Does the rotor lose rpm (spin down) slower with more "reversed" stator magnets
engaged?"

A: I don't know for sure, because of the instability of more than one spinning.
I usually have been just spinning one stator magnet in the "reversed" mode while
letting the others do whatever they want to. Maybe only one stator magnet is
needed. Or maybe a bigger layout would help.

Q: "P.S. I also want to confirm you are saying the rotor is accelerating from say 400 rpm
up to 1160 rpm by only reversing the spin of one stator magnet."

A: I'll bet you do. So do I.

OC- Dec 31st 2007


Some ideas about things we can vary:

1) Number of magnets on stator and rotor: 3x2, 3x4, 3x8, 5x4, 5x6, 5x8, 7x8, 9x8, 11x8,
13x8, 15x8
For smaller rotors, we may want to see what happens when the ratio is greater than 2 to 1
(23x8?).

2) Rotor size and stator magnet spacing: various sizes and distances should be tried,
stator magnets in close proximity up to a distance where they have no noticable
interaction with each other.

3) Gap between rotor and stator magnets. Smaller gap produces greater forces, both
positive and negative. Gap can also affect interactions between rotor and stator magnets.
4) Magnet sizes and strengths. I have seen best results when rotor magnet length is
between 1 and 1.5 times stator magnet diameter. Longer rotor magnets and smaller rotor
diameters may also cause more complex magnetic transactions, incorporating field
interactions between 3 or more magnets simultaneously. Larger and more powerful
magnets are available. Different magnet thicknesses can provide different responses due
to field strength and distribution.

5) Pivot angle and magnet offset from pivot point can be varied for optimum results, far
from pivot point for maximum centrifugal effect, closer for less.

6) A latching mechanism may be required, especially at slower speeds with larger rotors,
and more so with longer rotor magnets.

7) Rotor magnet orientation (chained or alternating) can be switched for different effects.

8) Rotor magnet axis with respect to the stator magnet axis when in closest proximity.

9) Speed can be varied.

10) Ferromagnetic materials or coils can be introduced in a variety of places to note what
effects they have. Pole pieces can be used to concentrate and alter the field distribution.

11) Lead and lag effects (articulated rotor arms) that can possibly leverage magnetic
viscosity. Add viscosity by using different magnetic materials (iron pole pieces, Alnico
magnets, etc). Slow in/fast out.

12) Magnetic reconnection effects (small rotors/fast speeds, magnet orientations).

13) Reverse the roles of stator and rotor. Odd number of rotating magnets on the rotor
and even number of fixed magnets on the stator. Should change the dynamics a little bit.

14) Use different shape magnets. Do rectangular or cube magnets behave differently than
cylinders? Corners and sharp edges may influence the magnetic field interaction.

15) Heavier and/or lighter weight rotor/stator. Lighter weight would respond more
quickly to changes. Heavier would provide more momentum.

16) Dual rings of rotor and/or stator magnets influencing each other from both sides and
greatly complexifying the issue.

17) Dual, counter-rotating rotors or dual stator rings with sandwiched rotor in between.

Any other suggestions for things we might want to check?

edit: added 10 - 17
OC- Jan 1th 2007
@Al,

I appreciate your thoughts re electronic latches and Helmholz coils but I see downsides to
either one.

Electronically controlled latches could facilitate our initial analysis. They may be easier
to design and manage than purely mechanical latches and would provide more flexibility
in timing, testing, etc. However, any demonstration device we create would instantly be
suspect if any external power was required for operation. There is also the possibility of
EMI influence on the device. If used, the electromagnetic components should be kept at a
distance where EMI is unlikely and/or shielded from the device.

Helmholz coils are an interesting idea. Again, I have qualms about powered devices
being used (unless they are actually internally powered by the device itself, without any
batteries). Shielding would defeat the purpose of having the coils in the first place.
However, it may be possible to achieve the required effects using permanent magnets and
ferromagnetic materials. Halbach arrays have been successfully used to provide
Helmholz-like fields.

I do not have any objections to using passive coils or ferromagnetic materials to help
shape and focus the fields we are dealing with.

This project is as much yours as it is mine and I encourage you to do whatever you want.
You have already discovered at least one phenomenon that I never would have expected.
Who knows what you'll find next? I'm just kibitzing the best I can. Take it for what it's
worth. This idea would never have progressed beyond a couple drawings and some
random thoughts if you hadn't taken an interest.

If you look through the (growing) list of variations I posted earlier, you will see there are
enough possibilities there to occupy you for a lifetime. And you just broght up a couple
more.

I really think we should try to get as many replicators involved as possible. One way to
do that is to post some videos showing curious behavior and to have some detailed
documentation about how to build it, something any journeyman machinist could follow.
Once they see how easy it is to replicate what you have, then they may be inclined to
modify the experiment and explore other variations on their own.

I'm really happy to see RiterX take an interest. I hope he gets his supplies soon and gets
started on the larger rig he is planning to build. I also hope he gets a chance to flesh out
the documentation a bit more.

I wish some of the SPDC folks would participate. There are some talented folks in there.
Guess they're preoccupied with their orbotic efforts.
OC- Jan 2th 2007
@Al,

I made a rough drawing of my concept of a pivot. Something like this could be


interchanged with different size rotors, as long as the rotors have approximately the same
thickness. I'm sure you can design something better. Let me know what you think.
http://api.ning.com/files/L50DUpMAxO2FwLiCwTYXg3a-wRIA5hpJ71MVTyEhK5s_/
pivot.gif

How much would you charge to make me a couple of these devices?

1) 18" rotor with 8 pivoting magnet mounts, bearing and hardware.


2) a base with 13 stator holes
3) another base with 15 stator holes
4) 15 stator magnet assemblies with bearings and hardware.

I can provide my own magnets if necessary, but it might be better if you mounted them as
well. I'm likely to damage something if I have to use something more complicated than a
screwdriver.

After all, why should you have all the fun?

OC- Jan 2th 2007


@Al,

I really think you should get some nonmagnetic bearings for the stators,
nonconductive would be even better. Ferromagnetic materials will have a tendency to
distort the fields by pulling them towards the bearing and will reduce the magnetic
interactions with the rotor magnets. I don't know how much difference it will make ... but
who knows, it might just be enough.

Al- Jan 3th 2007


A dimensioned sketch of the bearing/magnet holder assembly is posted on the other site.
The baseplate that I am using is acrylic plastic, 5/8 in thick, 7 3/4 in x 8 1/2 in, with a
1/4-20 threaded hole in the center for the rotor axle, and 13 evenly-spaced 4-40 threaded
holes for the stator magnet bearing holders, on a circle of radius 3 3/8 in centered on the
axle.
The rotor is 3/4 in thick piece of HDPE plastic, 2 7/8 in radius, with a 0.500 in center
hole bored all the way thru for the flanged axle bearings. Magnets are held in 8 evenly-
spaced 1/4 in wide slots, 7/16 in deep, inner edge of slots at the 2 1/2 in radius from
center.

This is the basic layout for the 13x8 version that I am experimenting with. But I am
mostly just using 3 stator magnets and a couple of "dummy" aluminum pieces of the
same size as the magnets, also mounted in the same type of bearing holder.
The stator magnets are mounted by a single trimmed 4-40 SHCS and a little washer to
make sure only the inner bearing race is contacting the screw or the baseplate.
The rotor magnets are press-fit in place.

Al- Jan 3th 2007


I'd stay with either Delrin or HDPE for the rotor--good dimensional stability, OK to
machine, cheap, heavy enough for a good flywheel effect, strong enough to withstand
high speeds, won't crack.
3/4 will be thick enough.

Al- Jan 3th 2007


All right, I made some slight mods today and then tonight I've been doing some more
testing. I'm pretty shaky right now, it's a big adrenalin rush, already the idiots are coming
out of the woodwork, I expect this thread to be infected at any time.

First, the spacing between the stator magnets and the rotor is critical. With the
layout I've got, 5 mm is the magic dimension. At 6 mm, the magnets lose synch. At less
than 4 mm, air resistance between the parts seems to inhibit the free rotation of the stator
magnet assembly. That's from the outside of the holder to the outer edge of the rotor (not
to the rotor magnets).

Second, not all magnet/bearing pairs will produce the effect. I have 15 of the magnets,
and 13 of the bearings. I have not tried all possible combinations of these, obviously. But
of the 13 assembled magnet/bearing pairs, 3 don't seem to work at all, 2 are fairly easy
to get started, and the rest are somewhere in between. Even the "easy" ones usually
require several tries before successful engagement. I have blisters on my fingers from
spinning stuff over and over and over.

Third, the relative height of the stator and rotor magnets seems important. I don't
know if it's because of my particular geometry or whether it is critical to the effect itself,
but the magnets shouldn't be in the same plane--that is, I am getting best results when
the rotor magnets are rotating in a plane that is just slightly higher than the tops of the
stator magnets, rather than down and centered vertically.

Fourth, I have been trying to get three stators spinning, but it turns out that 2 can spin
gear-wise on their own, and the third is finger-flipped antigearwise, and when it synchs
and the rpms go up, the other two stators can actually be stopped, and the rpms will go up
even more!

I know I'm not being very clear about all this--it's pretty exciting to say the least. Next
step is to see if I can detect any stray oscillating EM fields (the RB effect) with this little
cheapo E-Field meter...

Evolvealready- Jan 3th 2007


Al, are you pulling our collective leg?
Hey, I had to ask!

Al- Jan 3th 2007


@Evolve: Every word I'm posting is the truth. I might be wrong, in fact I hope I am.
All the data is real. It might be wrong, but it's really data, taken to the best of my ability
with the instruments at hand.
Obviously I have constructed a real device, it's not virtual, I suppose you've seen the
pictures.
I have the skinned knuckles to prove it, the blood's certainly real.
Maybe the sweat and tears, too.

Don't forget, OC saw this design in a dream, and I discovered this interesting behavior
totally by accident--I was trying to flip the damn magnet the other way but I was so, shall
we say, distracted at the time, that I didn't realize which way I was flipping it.

Al- Jan 3th 2007

alsetalokin:... The mod I did today had to do with getting the rotor mounted on much
better bearings.........

Can you please give the manufacturer and model number?

No.
I mean, I scavenge most of this stuff from surplus junk equipment that I scrounge in back
alleys.
And you want manufacturers?

These that I'm using now are for 3.5 mm shaft size, in a paired housing that slips into the
rotor's central bore, extending down to a shock-mounted base bearing unit.

What I had in there before were just some generic flanged 1/4 inch bore unshielded
bearings, and the shaft was a 1/4-20 brass screw. Not the most accurate arrangement.

Al- Jan 3th 2007


riterX:
alsetalokin:.....I'm not even sure what I'm seeing is real,
yet....
But you have similar results at work... and in the bus.

Ahh, no. These tests have so far only been performed in my basement. I
discovered the effect over the holidays, today was the first day back at the lab, and
I had the unit disassembled for the bearing mod, and didn't put it back together till
I got home.

(ETA and for a few minutes I thought I had killed it--I couldn't get it to start, for
the longest time--then I realized I had installed one of the "dud" stator magnets.
After I already raised a blister...)

Al- Jan 3th 2007


One new pic up.

@OC: How do I make the animation go? It looks beautiful...

@rX:

Do the stator magnets look any different under the magnetic field indicating film?

I don't know, I hadn't thought of looking, but I will when I get a chance. Thanks for the
suggestion! (Although the film is sort of low-resolution, or rather contrast; it might be
hard to see small differences.)

Do you think this is caused only by the difference in the magnets or the difference of the
stator/bearing assembly? Could certain bearings be the problem?

It could be either one. I suspect the bearings first, as the magnets are obviously a high-
quality product, and the bearings just as obviously aren't. The bearings are from two
different manufacturers, although they are apparently exactly the same in dimensions and
components. I had 8, they are Stewart-Warner, and I had to buy 5 more from another
source, they are unmarked.

What is the strength of the rotor magnet you are using now?

They are generic N35 grade NdBFe, so BH product is approx. 35 mega-Gauss-Oersteds. I


don't know what the surface field strength is.

Did you remove the seals on the bearings you are using?

I tried to but I couldn't get them out easily, so I just left them in. These bearings are pretty
sloppy, they aren't really "sealed" but rather "shielded" and the system used doesn't
produce nearly as much drag as a truly "sealed" bearing does. That's actually another
reason why I didn't use the helicopter bearings--the seals on those are a bit draggier, and
even more difficult to remove.

Al- Jan 4th 2007


@rX: I have little data on variants. If the system works, as I suspect it will, with only the
one stator magnet, you might try that first to see if you can reproduce the antigearwise
motion, as a start.

The dampers are tuned by trial and error. If too close they retard the magnet too much
and it won't synch. Too far and there's no benefit. Once the distance is found, rotating a
little at a time and trying the spin until I found a "sweet spot" is how I did it. The effect of
the dampers is small, in any case.

@OC: I think the rotation rate will actually help the pivots move as we want them to,
perhaps requiring a spring for return, but that's OK because that energy is stored and will
be returned later in the cycle.

The dampers are 6061-T6 aluminum alloy. I believe the silicon and copper content of this
alloy contribute to its characteristics. They are fixed in place but must be adjusted for
optimum orientation, as they are machined from extruded heat-treated 6061 rod and the
properties may be slightly anisotropic.

Al- Jan 4th 2007 (Editors note: When all the trouble started)
At the risk of starting a brouhaha, I have posted a video of the stator magnet behavior. I
will only leave this video up for a short time.
Everything you see is real.
http://freeenergytrackers.ning.com/video/video/show?id=1638621%3AVideo%3A2526

Al- Jan 5th 2007


@HT: The rotor is made of HDPE which is a bitch to machine, and cuts usually come out
a little small. I used a 0.250 2-fluted end mill, one pass in width, to cut the slots for the
1/4 in nominal OD rotor magnets, and they press into the slots perfectly--that is, a
medium press fit. So if you specify tolerances of +/- 0.0005, which is what I worked to
on this project, all should be well.
The stator magnet/bearing holders are Delrin, and it cuts small too, but not as much as the
HDPE. I cut for exactly 0.5000, and stopped when the mike read 0.4997. I wanted the
magnets to be pretty tight so they wouldn't slip, anticipating the latch requirement.
With the Sherlines, I can always achieve +/- 0.0005 accuracy--they have digital readouts
and are very accurate. If I am careful I can get down to 0.0002 in most materials.

Al- Jan 5th 2007


Several folks have asked for specific blueprint-like details. They are already there! As
above, I answer questions, and have posted dimensioned drawings and descriptions that
should be fairly easy for any tinkerer to follow. Detailed machinist stuff is also readily
available if you know what to ask, like HT.
OC has posted some drawings and simulations of the original design and some
alternatives, as well as the basic motions, and / has made some really neat animations
showing field interactions.
I am hoping that someone will use some sophisticated mag model software to illustrate
just how the non-gearwise rotation synchrony is achieved. I just can't seem to wrap my
mind around that.
But in spite of that, when I accepted what my eyes were telling me, I realized that the
performance could be increased by adding just the barest retardation to the driving stator
magnet--really an extension of OC's latch concept, but something that could work at
higher speeds without extracting (or adding!) energy from the system, like an
electromagnetic latch or damper might. So once again, serendipity operated, and I
discovered that the aluminum alloy dampers did the trick--at least they seem to help.

Al- Jan 5th 2007


@rX: I have little data on variants. If the system works, as I suspect it will, with only the
one stator magnet, you might try that first to see if you can reproduce the antigearwise
motion, as a start.

The dampers are tuned by trial and error. If too close they retard the magnet too much
and it won't synch. Too far and there's no benefit. Once the distance is found, rotating a
little at a time and trying the spin until I found a "sweet spot" is how I did it. The effect
of the dampers is small, in any case.

@OC: I think the rotation rate will actually help the pivots move as we want them to,
perhaps requiring a spring for return, but that's OK because that energy is stored and will
be returned later in the cycle.

gafnaz- Jan 5th 2007


@Couldbe

The total amount of energy in the system is a sum of kinetic and potential energy. Kinetic
energy in the rotating masses and potential energy of work that can be done by the
magnetic field.

The geometry of the machine is very complicated which can result in multiple stable
modi of operating. When the machine is running in an unstable position it will
automatically converge into one of the stable ones. It tends to minimise the potential
magnetic energy in exchange for the kinetic energy (optimal entropy).

Start:
After the initial manual start is done 'right' the system has some potential energy stored,
and is in a non-stable position. It then spins up using the potential magnetic energy,
resulting in a stable position.

Stopping the two stator magnets:


Conservation of energy results in increased potential magnetic energy when the two
stator magnets are stopped. Immediately after stopping them, the system is in a non-
stable position. After that it converges to the new stable position, decreasing the potential
energy again and upspinning the rotor.

The magnets don't loose their magnetism. Like pulling two magnets apart increases the
potential energy (making them able to do work), the initial push to start the machine
increases to total potential energy of the system (which is then used to spin up the rotor).
After the machine stops the magnets are still the same.

Nice thing to think about is the possibility that the geometry of this machine results in a
lower potential energy while spinning then when it has halted. It could result in a set of
very complex transformations between different stable positions while spinning down
(maybe even with some upspinning between two positions!).

Al- Jan 6th 2007


Mass of rotor assembly, including magnets and shaft adapter, but not including
shaft or bearings:
258 grams.

Mass of stator magnet assembly, including holder, magnet, and bearing:


10 grams. Variation between the 13 units is below the resolution of my (home) scale.

Al- Jan 6th 2007


I keep discovering more interesting things. I am able to show a homopolar voltage--
alternating, of course--across radiuses of the rotating stator magnet. The oscilloscope I
brought home is a very basic analog dual beam scope and doesn't have any kind of
waveform storage ability, and it takes both hands holding very delicate probes to do the
measurement, so I can't show a picture--not enough hands! and I'm off pictures for a
while anyway. But I'm seeing a voltage, presumably generated by the Faraday homopolar
principle, between the inner edge of the central hole, and the outer rim of the magnet
(actually in a little on the top face, I can't really get to a good edge with the bearing
holder the way it is.) The voltage alternates in polarity in time with the magnet's poles, I
suppose. It's not much, at least 2, no 20 millivolts (I just remembered it's a 10x attenuated
probe) but hard to estimate as it has a very fast rise time, beyond the resolution of my
scope (10 MHz, I told you it was basic).

Al- Jan 6th 2007

Just slight differences: the dampers on the real thing are mounted lower. It looks like
everything is on the same plane in the model; in the real unit the plane of the rotor
magnets skims the top edge of the plane of the stator magnets, and the damper plane
skims the lower edge, if that makes sense.
Also I have changed the rotor bearing/shaft bit: now the rotor itself has no bearings,
but is a light press-fit onto a precision centerless-ground shaft that itself is mounted
in a holder with 2 very fine ball bearings; this holder with shaft and bearings stays
attached to the base when the rotor is removed.

Al- Jan 6th 2007


@pc:
I took down the dimensioned drawing of the stator magnet holders. The Delrin rod stock
is 0.625" nominal OD. ID is bored to just a rch under 0.5000", so the walls are 0.125"/2,
or 1/16" nominal. Height 0.500". One side bored deep enough for the magnet, other side
bored deep enough for the bearing, with a little ring of material left inside (3/8" center
hole) to separate the parts and to provide clearance for the mounting SHCS (and washer
if needed).

Stator mags are K&J Magnetics R834DIA, complete specs are on their website (although
the central hole is a little bigger than advertised).

I adjusted, or rather experimented at different spacings, by mounting the rotor


temporarily on the 5x4 baseplate, sliding it up against the larger base with the stator
magnet (1) mounted. I photographed this but I took the picture down. I had to use DVD
cases to block up the baseplate to compensate for height differences. I used the Dremel
with a buffing pad attached, to spin up the rotor to a known rpm, as measured with the
laser tach. I then spun the stator magnet by hand, trying to get it to synch as the rotor
slowed down. I played with this for several hours, it seems, before I found that about 5
mm seemed good--that is, easiest synch and longest rundowns--no sustained runs yet,
not even a hint, except that improbable speculation graph I showed.
And this was another stroke of absolute improbably luck, because I had already laid out
and drilled the holes for the stator mounts, at what seemed like too far away from the
rotor for OC's liking, because I was anticipating pivots and latches and all that
mechanism in the gap there.
Almost the same process for the damper location and orientation; but easier because I
could use the right baseplate which I installed with excellent bearings, and the runs were
longer and I didn't need the Dremel so I had more hands free.

Al- Jan 6th 2007


@photon--believe me I am verry familiar with the use of compressed air to spin things.
Early on I did use compressed air, but it's too hard to control.
Fast speed isn't necessary; if you get the rotor too fast it is impossible to synch the stator.
I think that not only does the 4-1 speed need to be established, but the phase relationship
must also happen correctly. I've had best success as the rotor is actually slowing down,
from 350-400 through about 250, and I estimate that below 200 rpm the stator will not
maintain synch in the anti-gearwise mode (gearwise is fine all the way to zero).
I have been able to get the stator to re-synch after dropping out, before the rotor slows
appreciably, but that requires a really good flip and a bunch of luck for the phase. It
usually doesn't work.

Al- Jan 6th 2007


Thanks, Mary. And that's exactly why I am chagrined at Sterling's disregard of my
wishes in this matter.
I just hope the boss can see the humor in the situation, tomorrow at lunch. Of course his
boss has probably already alerted him...

I understand your suggested protocol and I agree that the replicators should do this
exactly as you have outlined. I did something like that early on, and was able to
determine that the stationary magnets, and the gearwise rotation, extract energy from the
system, as you would expect, of course. Interestingly, if the stator magnets are "frozen"
by overtightening or leaving out the spacer washer, the rotor turns longer than if the
stators are stationary but free to vibrate. Again, as expected.
The unexpected behavior only shows up running with the one magnet spinning anti-
gearwise.
I have arranged a little variable-drag system using pulleys and thread; by starting up with
the thread loop slack, and then swivelling one pulley to tighten the thread once the system
is running, I can put a slight repeatable drag on things. It slows the equilibrium speed
down, but with only a little drag it doesn't stop it or cause it to run down faster, as far as I
can tell (these are VERY preliminary results, like from 2 trials).
I am trying to implement a pony-brake system but the requirement for absolute free-
spinning at the start is requiring a rather sophisticated clutch design. I may have to go
back to a magnetic eddy coupling, like in my old namesake's speedometer design (which
has only recently been replaced by digital stuff in autos, I am proud to note.)
I'm sure the thing could direct-drive a little aircraft propeller without difficulty.

Al- Jan 6th 2007


Oh, I guess this would be a good time to mention that any metal or magnets lying around
on the work surface can inhibit the ability to synch the stator. I was having fits trying to
start it last night, until I noticed a little precision screwdriver was lying there, 6 inches
away--too close. After moving it away, I was able to synch the stator easily. Well,
relatively easily.

Fergusthepup- Jan 6th 2007 (first confirmation of replicating the Alsetalokin


effect)
AL, OC, I'll put some details up in about an hour or so - been looking after the snapper
and eating dinner... there's not a lot to say other than the anti-gearing of the stator
definitely happens in a smooth and repeatable way. I havn't got any sort of self sustain
effect to report - but i'm not looking for that in the short term because the stator is poorly
mounted on it's bearings.

please feel free to post my whispered message.


Paul.

Hello alsetalokin,

Thanks for all your efforts on this - certainly is interesting !

I'm not expecting technical (or any) feedback or help from you - there is plenty on the
steorn thread. Just thought i'd let you know the following is what I and my brother have
found so far...

I've done a rough and ready bodge on my rig just to see if i could get the anti-gearing
effect that you have been getting. I am using one stator assembly and a 100mm dia. rotor
with 8 magnets in opposition. I definitely get the anti-gearing effect at higher RPM
that stays in Sync until lower RPM. It really is surprising that the stator is driven in the
opposite direction. I can get the stator to "gear" if I wind the rotor up slowly.

It does not self sustain in any meaningful way - but I'm not expecting it to as the stator is
bodged and very "lossy". I have stuck (hot glue) two 6mm dia x 3mm neos to the outside
of a small ball bearing to give the spinning poles. The flux pretty well seizes up the
bearing but the inner race runs on a 3mm shaft like a plain bearing. One advantage of the
6mm neo's is I can vay the diameter by adding 6mm dia x 1mm disks on either end to
vary the gap and strength. This arrangement works surprisingly well even at very high
stator RPM - none of the neo's has come flying off yet

Diametrally magnetised neo's will be ordered tommorrow and I will get a proper stator
housing assembly made during the week. My rotor is mounted vertically between 2 plates
so the stators will probably end up supported between 2 bearings like the rotor.

Regards

Paul

Al- Jan 6th 2007


My strategy has been to use the fingers to get both parts spinning, and to hope that they
will synch as they both slow down thru the critical rpm range which I believe is about
300-250 rpm on the rotor.
I will have to try the Dremel again and see if it gets easier. The Dremel is a commutated
motor though and kicks out a mess of EM on its own, which is the main reason I
tried to get away from using it...

Al- Jan 6th 2007


I rigged a model airplane propeller directly on the rotor shaft. Not very well, I can't
guarantee it isn't slipping a little bit.
It makes it much harder to start, for sure. Especially reaching under to spin the stator
magnet.

But once it starts, it looks like the max RPM is down around 875 rpm at the prop, if I can
trust the tach. Can't read anywhere else, like the stator magnet, because the prop is in the
way. But I'm sure the usual 4-1 relationship holds.

It's not making much of a breeze, the prop is a puller so in the current spin direction it's
"pulling" the air from around the rotor and pushing it upwards, so most of its intake is
blocked. But the breeze is easily perceptible.

The prop is GWS EP-1080, 25.5 cm from tip to tip.

Al- Jan 6th 2007


And I meant to add that I think they function, well, as dampers. They seem to reduce the
bearing wobble and add just enough drag to slow the system down to a more stable
rpm. The reason they are eddy-current dampers instead of mechanical drags is that, guess
what, the eddy current resistance is proportional to the speed, so at slow speeds it is
negligible.

OK the dampers are non-ferromagnetic, conductive, thick, alloyed and heat treated metal.
To wit, AS I HAVE ALREADY SAID several times, they are 6061-T6 aluminum
extruded rod or bar stock, machined to identical dimensions as the stator magnets,
and oriented and spaced by trial and error.

MrEntropy- Jan 6th 2007


alsetalokin:Far more significant is the anti-gearwise rotation of the stator magnet.
Whether or not anybody is able to replicate the long spins (that's harder than it seems
from my posted descriptions; I got lucky in the video), the stator spins are easy to
demonstrate, and that's the feature I would like to see explained in a sensible way that
doesn't require re-writing too many textbooks.

Hi Al,

I can help out with that one. Forget about the stator magnets for a moment, and
consider the magnetic field from the rotor at the point where the main stator
magnet is attached.

If the rotor is rotating at a constant RPM, the field at this point will vary
periodically, with a frequency of 4 times that RPM, since the rotor moves between
equivalent positions at that rate.

Any such periodic variation can be decomposed mathematically into the sum of
two amplitude-modulated fields -- one that rotates clockwise at that frequency and
one that rotates counter-clockwise at that frequency.
The stator magnet can be synchronized to either one of these fields, keeping itself
aligned with the field line through its center, as long as the rotor RPM is high
enough. At sufficient RPM, the other field has no appreciable effect. The net
torque it applies is zero, and any momentum added in one instant is quickly
counteracted.

Note, however: You may have stable synchronization, but it looks to me like the
field component you're synchronizing to isn't very strong, so the synchronization
might not survive sudden changes in the load on the rotor or stator. I read (in
another thread?) that you were opposed to belts. Why is that?

Cheers,

Mr. Entropy

MeggerMan- Jan 6th 2007


Hi Al,
I just noticed that your tag is Nicola Tesla backwards, good one.
When you take a flash photo of the running unit, is the stator magnet in the same relative
position to the rotor, you may need to take quite a few photos to get the various postions
of the stator?
If you could setup some white LEDS to a trigger from each of the rotor magnets, perhaps
using a hall effect IC. This way you could see in real time the relationship between the
stator magnet polarity and the rotor magnets.
Another way to get three stators to sync would be to drive each stator with a small pulse
coil(no iron core) then once they are in sync then stop the pulses.
The stator pulse coils would need to be pulsed from a pickup on the rotor provided there
is a relationship with rotor and stator as shown above.
[edit] ..or connect the stator to a small 12v brushless fan motor that matches at least 4
times the rotor speed.
When its up 4 x rotor speed(what ever that is at the time) reduce the power to the fan
motor until it syncs. If the rotor is turning slowly then this will be a viable option.
I have some spare 1" fans if you need any.
I wonder if torque is related to the number of synced stators then 8 stators would be
excellent.

I am hoping to build this myself and use it as a tool to develop the TPU.
I have made up a jig to cut the disks on a router table:
http://i100.photobucket.com/albums/m25/kingrs/disk_cutter.jpg

The end result is the clear polished disks in this picture of Mikes motor:
http://s100.photobucket.com/albums/m25/kingrs/?action=view¤t=DSCN4986.jpg
I do not have a cnc machine like some have, wish I did, just have to make do with what I
have.
[end of edit]

@Paul: look forward to you diode array device make the next news item.
@Oak: thanks for the heads up. :
Regards
Rob

Fergusthepup- Jan 6th 2007


@alsetalokin
The rotor on my rig is mounted on a 3mm steel shaft that extends a couple of inches on
one side - it allows me to wind the rotor up by dragging my finger repeatedly along it
building up the rpm in excess of 1500 rpm. At this point a deft flick of the stator in the
Anti-gear direction ( same direction of rotation as the rotor ) gets it sync'ed pretty easily
-- I'm guessing that it's the high rotor RPM that allows this.

Have you had a go at getting the anti-gear stator sync'ed without the other 2 stators?

Al- Jan 6th 2007


woo, fergus, great work, I have trouble synching at fast rotor speeds, your geometry or
bearings must be somehow better(?) than mine---

Yes, it will do without the idlers, but is more difficult. I know they are doing something
because of the way they behave when stopped. Again, those details posted elsewhere,
sorry.

Fergusthepup- Jan 6th 2007


@alsetalokin

The other thing about the stator I am using is that it can float side to side ( up and down if
it were on your rig ). It's not deliberate - I was just keen to get a look at what happened. It
does however become "centered" in the rotor field. I want to be able to fix the stator in
the next version. This floating may allow it to sync more easily.

The other thing is the stator magnet configuration - it is effectively a spinning bar magnet
with a longer narrower field which may make it "engage" with the rotor field more
positively and hence make it easier to sync.

Do you shout "clear prop" when you start a run ?

Al- Jan 6th 2007


It will do just that at low speeds as it "drops out", and yes it can be hard to tell especially
if the lighting is fluorescent. But you can feel the drag with your finger (just before the
rotation dies from the drag!) and tell the direction that way, and if you catch it right,
there's no doubt which way it's going. When I did the homopolar experiment I used tiny
little itty bitty wires wrapped around the tip of the scope probe, and it was very clear
which way things were turning from the way these wires behaved when the magnet was
contacted.

One bit of kit that I don't have readily available is a shaft torque encoder like Steorn used.
I have spoken earlier about designing and constructing and calibrating an eddy-current
damping-coupled pony brake dynamometer, that would allow the appropriate
measurement to be made.
Or I might rip apart an old VW speedometer, if I can find one in the local junkyard. It has
all the components and all I'd have to do would be to hook it up and calibrate it with a
known torque source.
I think. Maybe.

ETA But perhaps you are talking about electro- and magneto-dynamic forces. I made a
start on that with the Rogowski-coil oscilloscope probes of the "active region" if indeed
that is what it is. The waveforms and the coil placement locations were up for a while,
jeez, was it just this morning that I did that? Holy cow.
Anyway I took them down until the dust settles a bit. But I think somebody cached them,
but I don't have a link. Sorry.

@Clanzer: let's see--the position of the rotor when starting the manual movement of the
stators determines how the rotor behaves with a given stator input. If I swivel one of the
stators to just the right angle and hold it against the springy field, then rock it back and
forth about 15 degrees, and the rotor is in the right phase, I can keep the rotor rotating
continuously at about 1 rps or 60 rpm. If the rotor isn't in the right phase or I don't rock at
the right instant (same thing) the rotor stops or reverses direction.
So the fact that it didn't rotate for one magnet at the beginning of that hasty video isn't
significant, I don't think.
As I have indicated before there is variance in how the individual magnet/bearing pairs
behave. The "middle magnet" in the running rig is on the smoothest bearing out of the 13
that I have. I can't tell any variation among the magnets by the means at hand (viewing
film, guesstimations of pull strength, etc.) but there does seem to be something "sweet"
about the one I ended up using. Several of the stator magnets that I have don't seem to
work at all, but it might be the bearings. I haven't tried all possible combinations of the 13
bearings and the 15 magnets I have!

I think the dampers are working by keeping the bearing wobble down, and by providing a
gentle drag force that disappears when the parts are moving slowly. Eddy, Lenz, and all
those things.

Without the dampers the unit seems to speed up to instability, the stator chatters and
drops out, and the rotor coasts to a stop.

And one more time: 6061-T6 extruded aluminum bar stock, machined to the exact
dimensions of the stator magnets (except I really did use a 3/16 hole, the magnets are a
bit bigger here), spaced and oriented by trial and error, and in my unit, the pre-drilled
holes for the stator magnets seemed close enough to the sweet positions so I just used
them.

No I won't tell the story of how I found the sweet spots again, children, it's almost
bedtime and you need to say your prayers.
And please pray for me when you do.

Clanzer, the dampers on the cad rendering are toruses, but the ones I made are, as I said,
cylinders machined to identical dimensions as the stator magnets. My original idea was to
mount them in identical holders as the active magnets but then I realized they should be
stationary once properly positioned.
I think the toruses look way cool though, don't they?
Except the holes in the stator magnets are a little oversize, and the holes in the dampers
are 3/16.
I think if one used toruses that might be better performancewise, but almost impossible to
machine, especially without CNC tooling.

If you can find a copy of the picture sense_1.jpg that I removed from the ning site in a fit
of pique, that shows them quite well. I'd send it to you but for some reason I can't seem to
send attachments from that account. I won't be re-posting it until the fuss dies down but I
know someone has archived all that stuff somewhere.

Yes, that's the one. Thanks Mr Entropy for the analysis and to rX for reminding me of it.
Certainly just the slightest touch on the spinning stator will cause it to drop out of synch
and quickly stop--like from 5k rpm to 0 in about 4 revolutions. I mean when it stops, it
flat Stops. Then it just wiggles and growls (bearing chatter) until the rotor slows waay
down, then it will re-synch spontaneously gearwise, then it starts the washing-machine
motion, then as the rotor is about to halt, it stops too.

I'm not exactly opposed to belts per se, even though as I age I seem to prefer the comfort
of suspenders. Those elastic Spandex things are going too far, however.

No seriously, there seem to be many synch modes available to the system and belting
would cut out a bunch of those modes altogether. Now if one synch mode is somehow
optimum (in what way? I don't know...) then a belting arrangement might make sense, to
keep it in that synch mode during liftoff and orbital insertion.

OC- Jan 6th 2007


@Al,

1) Did you ever try the stator magnet temperature test?

2) Remember the first couple images you posted, OC_config_a and OC_config_b?
Remember, you asked what the difference was and I focussed on the rotor magnet
polarity difference, but you were actually talking about the rotor magnet positions in the
slot and where they lined up with the stator?
Have you ever tried moving the rotor magnets like that so they align a bit differently?
Specifically, I was curious how it would behave if the magnets were moved forward
(direction of rotation). Does it make any difference in A) starting the anti-gearwise
rotation B) speed C) length of run

With this rig, all you need to do is rotate the rotor 45 degrees and ... voila, opposite
polarity. Make sense?

Al- Jan 6th 2007


OK, at great cost to the fingers of my left hand (I for sure don't have the coordination to
wrap and pull a cord left-handed)---
Yes, it runs counterclockwise. If anything a bit more smoothly than clockwise. As far as I
can tell it doesn't care which way it goes--everything seems the same, except maybe the
rpms are a tad higher.

modervador- Jan 6th 2007


The aeronautically useful pressure gradients move a predictable mass of air per unit time
consuming a predictable amount of energy per unit time. Ergo, there will be a significant
drag imparted by the rotating prop.

The mechanical speedometer guts from an old car is a good idea for a load. Here's a
variation: you could mount an aluminum or copper ring (or cup) atop the rotor, and you
could control how deep a bar magnet is lowered into the cup to vary the braking effect.
Some kind of spring scale could measure the torque on the shaft holding the magnet, thus
with RPM and torque you'd know the power absorbed.

The ring would double as a sheave around which you could wrap a starter rope, like on an
old outboard motor or rototiller.

By the way, earlier somebody mentioned a "pony brake". I presume what was meant was
"Prony brake". One problem with it is the significant radial load it puts on the shaft
bearings, and the belt gets hot.

ETA: re legendre's post 42 seconds before mine: great minds...

Morgenster Jan 7th 2008


I think of another explanation:
KE input on the rotor at first makes it turn AND the stators turn. So in effect the kinetic
energy input is transferred to both rotor and stators on the first go. Afterwards more KE is
input by making one of the stators change direction and have it turn antigearwise. And
secondly KE is used for blocking the other two stators. Now imagine that these actions all
contribute to the KE of the rotor instead of diminishing it but without there being gain or
OU. I suspect that blocking the two last stators has the opposite effect of braking in that
changing the RPM to zero makes them transfer the kinetic energy of their motion (being
stopped) to the rotor. Same for the energy AL puts into stopping the stators. effectively
getting a gain in the rotorspeed albeit delayed. More like (KE first spin) + (KE reversing
stator) + (KE stator transferred to rotor) + maybe (KE stopping stator 2 and 3) = KE
rotor.
So now I'm pretty confident that the amount of energy put into the machine in total is
slightly more than what is put out by the rotor.
The only anomaly being the manner in which it happens.

Al- Jan 7th 2007


With rotor magnets installed, stator magnets installed, stator magnets free to move but
not spinning, rotor initial rpm 1250, rundown time ~120-125 seconds.
Same as above but with stator magnets (3) spinning gearwise, initial rpm 1250, rundown
time ~ 90-100 seconds.
Same as above but with stator magnets removed and far away, initial rotor RPM 1250,
rundown time ~ 130-140 seconds.
Same as above but with stator magnets screwed down hard to baseplate, not moveable,
initial rotor RPM 1250, rundown time ~110-120 seconds, with a distinctly "coggy" feel
and a reversal of direction at the very last revolution.
All 3 stator magnets installed and spinning, 2 gearwise and 1 antigearwise, max rpm and
rundown time UNKNOWN.

Too bad I got driven off from the ning site. Saturday I posted Rogowski-coil scope traces
from a little coil inserted in the "active zone" and just upspin and downspin of it. Very
interesting waveforms.
But I took them down, because nobody seemed interested, and that Omnibus-clone Kevin
was flaming me in a very insulting manner.

And before you come up with the suggestion, I also showed AC Homopolar activity in
the spinning stator mags. But I couldn't photograph the traces, my hands were full with
the scope probes and the tiny contact wires and all. Clear, robust, repeatable AC output
from the Faraday homopolar effect on a spinning stator magnet.

vibrator- Jan 7th 2007


BTW those waveforms are familiar. The instances i've seen previously were (inaccurate)
simulations but the shape of the gain pattern is the same, most spudders will recognise it
straight away. This is something i was trying to pick up on earlier (i wasn't trying to bang
my OE theory, honest - i'd be suprpised if it had any relevence here), my point was that
the rotor mag sequence is actually two complete cycles of the same pattern per revolution
- the interaction sequence is a 2:1 ratio, even tho the rotor/stator ratio is 4:1, since 4
rotations of the stator sync with 2 transits of the rotormag sequence.

It would be good to be able to plot the magnet positions to the trace at some point.
Trawling thru my cache now for a copy...

ETA: so squaring it off, each face of the stator mag must see magnets going in alternate
turns - ie. N face see a SN pass, S face sees a NS pass, invert & repeat...

Al- Jan 7th 2007


vibrator, keep thinking. You are coming up with good operational insights. I hadn't fully
thought about the alternate turns pattern. I wonder if it means there will be higher order
synch modes available to a fine-tuned and smooth system. Of course we don't know what
role the bearing chatter plays, either--it could be detrimental as one might at first imagine,
or it could actually be necessary for the effect. Only experiment will tell.

Dirt farmer, yes, it was my experience with homopolar motors that caused me to look for
the effect in the first place. But the Faraday homopolar motor/generator effect has
nothing to do with coil induction, if I follow your meaning.
And homopolar generators are usually DC, the AC variant is rarer.
And, lastly, what is moving, relative to what, in order to generate a voltage, in the
spinning stator magnet?
Think about this, it is far from a trivial observation.

scout- Jan 7th 2007


http://www.sciencedaily.com/releases/2007/12/071216134745.htm

"Applying a magnetic field to a magnetic vortex pushes the vortex away from the center
of the disk towards the frame. If one then turns the field off abruptly, the vortex moves
either clockwise or counter clockwise on a spiral like trajectory back into its initial
position in the center of the disk. This special movement is called gyration. In principal,
the perpendicular magnetization of the vortex core can point either upwards or
downwards, and four different kinds of movement can be found: right- and left rotating
magnetic swirls, combined either with an up- or downward directed perpendicular core
magnetization."

MrEntropy- Jan 7th 2007


Al,

Thanks for doing the counter-clockwise test.

About the rotor scope trace: If I remember correctly, it wasn't entirely symmetric. Do you
know why that is?

Also, can the machine sync and self-accelerate without the dampers?

I did some calculations of the time-varying rotor field at the stator locations as predicted
by the conventional laws. I found that the anti-gearwise synchronization mode does exist,
and its phase relative to the rotor is as follows: With the stator magnet to the right of the
rotor, and a rotor magnet as close as possible to the stator, and it's north pole pointing
upwards, the anti-gearwise synchronized stator will be in the same orientation as the
nearby rotor magnet, with it's north side poiting upwards, too. A gearwise-synchronized
stator would be in exactly the opposite position at this point.

After the stator rotates another 90 degrees, its south pole will point directly into the
middle of the space between two N poles on the rotor. A gearwise-synchronized stator
would be in exactly the same position at this point.

The anti-gearwise synchronization has about 1/5 the torque of the gearwise
synchronization. In that mode, the counter-rotating "noise" is also about 5 times stronger
than the synchronization force, so you will need high RPM to use it. I tried rotor magnet
lengths from 3/8" to 1", and this picture didn't change much.

The gearwise synchronization mode, on the other hand, is 5 times stronger and nice and
smooth. I was actually pretty surprised to find that the strengh of the rotor's field at the
stator point only varies by 25% or so over the whole cycle -- the field there is close to a
constant gearwise-rotating field. It's no wonder that the stator magnets easily self-
syncronize in this mode.

Unfortunately, in perfect alignment, the path by which the magnets come together with
an agw stator is exactly symmetrical to the path by which they separate. That means that
you're not going to be getting OU through any kind of Kedron effect unless the Judson
dampers or friction are significantly altering the phase relationship -- so I'd like to know
if the dampers are really important.

Cheers,

Mr. Entropy

Harvey- Jan 7th 2007


@AL Those were good tests. The taller peaks are derivitive of the opposing rotor mags
while the shorter more rounded impulse comes from the stator(evidently the stator field is
weaker than the two combined rotor fields even at these proximities). There seems to
quite a bit of jitter in precession location (scope trigger perhaps). The persistance of that
scope seems quite high. The samples clearly show syncronization with a trailing roll-off
(shear separation) of the smaller stator field. The symmetry indicates the rotation is
well centered on the fields which leads me to believe the inertia of the rotor is the
primary driver with the stator assist.

I calculated the stator frequency at 4.83KHz and the rotor frequency at 1.2075KHz

A test to check doppler compression would be a dual probe (or at least dual data log/time)
gauss reading in both the leading and trailing positions relative to the stator. The
precession location should show a higher density. If this is not the case, then we should
look to an outside source such as a nearby EM generator running at a harmonic of those
frequencies.
I believe it to be doppler compression coupled with inertia. Although you may not be
the first to demonstrate it you certainly do deserve accolades for sharing your discovery
and helping to get the concept recognition.

Cheers,

Harvey

Al- Jan 8th 2007


@Harvey: that's pretty darn good work for an invisible rabbit.

I'm not sure if the Bell gaussmeters we have have the dual probe option. However I could
construct a dual flat sense coil and use some kind of calibrated positioner to guide the
insertion into the active zone (is it getting warmer in here?)

Nice to see that the scope supports the high-speed 4-1 synch hypothesis.

I need a spectrum analyzer...

PC- Jan 8th 2007


Some numbers. Hopefully any mistakes here will be spotted (values are rounded) :

Al's rotor is .258 Kg with a Radius of 0.073m. He spins it by hand to 300-400rpm.

For MOI I'm treating it as a solid disc with uniform mass (not correct because of the
magnets) so k=0.5. MOI=k*M*R^2 so I get an MOI of 0.00068791. At 400 rpm the
kinetic energy of the rotor is 55 Joules with the stators, if they are locked 4:1 running at
1600rpm with 0.32J each. So Al's hand is able to impart at least ~56J to the system per
interaction.

With the rotor at 1200 RPM, the KE is 495J - ten times the energy, and at 1904 RPM it's
1241J.

Al says it slows from 1200 rpm (without interactions) in somewhere around 120 seconds,
so if the power drain were constant (it won't be) that would be 4.06 watts. In other words
- to keep spinning at 1200 rpm, a motor might have to consume at least 4 watts (the
actual value will be higher at high rpm because atmospheric drag is proportional to the
square of the speed AFAIR).

For the rotor to spin up from 400 rpm to 1200 rpm would take 440J. If that happened
over 60 seconds, the average power a motor would have to supply is 7.34 watts.

I hope someone will come along and show I've made some kind of mistake!
Well, if you are correct Mr Structures, then following 1560 seconds of reported windown
in err ... 'operating' mode we get :

Change in rotor RPM over 1560s = -11 rpm


Change in stator RPM over 1560s = -52 rpm
Chage in rotor KE over 1560s = 9.9 J
Rotor losses = 0.01w.

So compared to a run without any magnetic funny business - the system has gone from
losses of over 4w to one hundredth of a watt. If 4w were expended over 1560s we would
need 6240 J.

Course, there is not neccesarily anything remarkable about that since we don't have a
windown time for the system in a vaccuume. That would tell us the aproximate losses
that are involved in atmospheric drag..

Al's says he can spin his rotor by hand up to about 400rpm, I'm presuming one good flick
will do that. At that point, the rotor has 55 Joules of kinetic energy. 1 joule represents the
energy to move 1Kg 1meter - so methinks Al really does have sore fingers.

And another thought strikes me at this very moment - if the energy is supposedly being
stored, the when Al is winding it up he should be able to feel resistance to his pushing. So
if you are reading this Al please report on how it 'feels' when you flick.

After the rotor accelerates to 1200rpm, it has 495 joules of KE wrapped up. So if there is
an energy store, it's able to store 10 times the energy Al is able to impart with a single
flick. Does he always have to flick it 10 times or more?

modervador- Jan 8th 2007


Stop there. Moment of Inertia (MOI, or just I) is indeed 0.00068791 kg.m^2 or 6.88
kg.cm^2 for a disc, but then it falls apart.

KE= 1/2.I.w^2, where w = angular frequency = 2.pi.(rotation rate)

At 400 rpm, w = 2.pi.(400/60sec) = 41.9 radians/sec.

So KE = 0.604 joule.

You get 55 joules for KE is if w = 400 radians/sec, which is incorrect. 400 is the RPM,
which is less than 10 revolutions per second, easily attainable by hand with such a
lightweight disc rotor.

Any further analysis based on this incorrect substitution of RPM for angular frequency is
incorrect.
I was led to look for this error by pcstru's later post which contained:

"Al's says he can spin his rotor by hand up to about 400rpm, I'm presuming one good
flick will do that. At that point, the rotor has 55 Joules of kinetic energy. 1 joule
represents the energy to move 1Kg 1meter - so methinks Al really does have sore
fingers."

This contained an error in that 1 joule actually represents the energy to lift a 1 kg weight
by about 10 centimeters on earth, or to exert a 1newton force over a distance of 1meter in
all parts of the known universe.

But the bit about the sore fingers prompted me to compare AL to other athletes.
I looked up the mass of a baseball (~145g) and computed the kinetic energy of a 100mph
pitch (145joules) to see how pcstru's fugure of 55 joules would compare. 55 is only about
3/8 of 145, the performance of the world's faster baseball pitchers, but flicking a 5 inch
disk is considerably different from hurling a baseball. Nevertheless I didn't think 7
revs/sec for a 1/4 kg disc was out of the realm of even avewrage human performance, so I
was led to doublecheck pc's math.

Everybody, now please doublecheck me.

@1250 rpm the rotor has 5.89 J. If it's losing that over 120 seconds, that's an average of
0.05 J per second, so that would give a top end of 1.6 days. No testable perdiction out of
that though - if your 17,000 joules is delivered mostly at the top end of the rpm scale the
higher drag will shaft that.

If the device slowed at a constant rate of 11 rpm per 26 mins, your timed wind down test
of the other night would have lasted 2.5 days.

Regardless of my earlier blunder , the reported energy charactristics of the device are
remarkable. It does seem easily possible for your hand to put in enough energy to make
the thing spin at the claimed speeds but only provided it can store energy and either
reconfigure the effective mass distribution with it or feed the stored energy into the
system slowly.

modervador- Jan 8th 2007


It looks like less than 7J would get the rotor by itself up to 1300+ RPM and a fraction of
a milliwatt would sustain it there. These are pcstru's recent numbers (of 3 hours ago),
with which I concur.

WhiteLite- Jan 8th 2007


The anti-gearwise stator and the rotors are alternating between attraction and repulsion.
Between rotor magnets you have N N poles facing the S pole of the stator and then you
have the N-S of a single rotor magnet parallel to the N-S orientation of the stator as it has
made a quarter turn. Moving from attraction to repulsion takes a bit of force but you
get that back when the next stator quarter turn is made going from repulsion to
attraction, (this is all assuming the 4:1 rotation ration).

The gearwise moving stators will always be in attraction until the are stopped in which
case they will be alternating between attraction and repulsion with the rotor magnets, (but
not at the same speed as the anti-gearwise stator, I think).

Al- Jan 8th 2007


But that's not the remarkable result. The stator rotation modes, and the AC homopolar
generator phenomenon, and the long-distance effects of the idlers on each other when the
rotor is running, these results are much more important than a simple rpm increase in a
spinning system. There are so many things that could cause that, it could take years to
find out. But one thing is for sure and for true: there is no excess energy in this system. It
is a test bed and demonstrator for some ideas and phenomena that will become important
when someone builds OC's original idea. That is the direction to go, in my opinion. Build
his design, use what we have learned from this device, and go on from there.

vibrator- Jan 8th 2007


I still think the most likely explanation is an asymmetric distribution of angular
momentum - that of the decoupled vortices - and that this creates an inequilbrium balance
of KE.

The magnets may be bouncing up and down their BH curves, and the slow deceleration
over time may be attributable to domain changes, i don't know, but compared to the KE,
the MPE i think can be regarded as constant - IOW i don't see any reason to suppose
large quantities of PE are being pent up & slowly released - there's simply nowhere to
lock it up without obvious effects (ie. heating / glowing / raised mass etc.)

And i can't rationalise a "partially-closed open thermodynamic system". Can true infinity
have a finite start? Surely either a system is open or closed, not both?

One reason i couldn't accept Steorn's theory of "creation" is that a time delay just acts as a
one-way valve for elementary electric charge, and i think the same point applies here.

In a situation where we have an unbalanced wheel with more or less constant PE but
rising KE, we have an input of energy. Perhaps the KE of these vortices are transfering
some kind of environmental PE - and elementary electric charge seems the most obvious
motive force.

This must create a gradient vectored towards the machine. Would it be possible to test for
such by running long wires radially around the assembly? There should be a measurable
current when the machine is in operation....

ETA - six wires could cover all directions, but one would be enough - just move it
between positions and re-measure...

Al- Jan 8th 2007


According to my own weird pet theory that I have never discussed here, there may be
about 17000 Joules available for momentum change in the device. At the calculated rate
of power dissipation, which I can't seem to find right now, was a thread removed?, how
long would that energy content predict the device would run for?

While my model does not rely on new physics, it is "new" in the sense that I haven't seen
it formulated before. It may be the only original idea I have ever had. That's why I haven't
discussed the details.
I hope it will suffice to say that the model says the energy is there from the beginning,
and is dissipated over time as the device runs.

Now, why I can seemingly restart it, well, that's a different question, that may indeed
invalidate my model.

But let's get a good number for the 17000 Joule rundown first.

PaulLowrance- Jan 8th 2007


LOL, hey, if the conversation is turning to la-la physics then please don't stop there ...
may as well add E=mc^2. There, that's plenty of energy to explain the acceleration, LOL.

I'm certain most conventional physicists are getting a good laugh over this thread, if
they're still around. Come on folks, this video is either a ***fake/lie*** (bingo) or it's
capturing energy from an unknown source or by means of an unknown method.

Al- Jan 8th 2007


"@1250 rpm the rotor has 5.89 J. If it's losing that over 120 seconds, that's an average of
0.05 J per second, so that would give a top end of 1.6 days. No testable perdiction out of
that though - if your 17,000 joules is delivered mostly at the top end of the rpm scale the
higher drag will shaft that."

Thanks for going back and doing it again.


Now we are getting somewhere. It is in fact testable. Straight lines to first order,
remember? So model the situation as a constant rate of power dissipation. After all, the
rpms stay pretty constant in the short (compared to 1.6 days) runs that I have seen.
So, to first order, if it runs for 1.6 days, that would be strong support for the 17000 Joule
figure, IMHO.

But of course if it is getting a constant supply of energy from some external source, that
isn't depleting appreciably, then it will run indefinitely, and my pet theory is out the
window.

One thing for sure, it is not creating energy, or angular momentum. These quantities are
conserved.

"The key is the silver tape "


Hah. I guess that poster doesn't care about the bit of silver tape that actually is on the
edge of the rotor, and would rather hallucinate his own. On the black stripe of electrical
tape (which itself could trigger a photodetector-pulse motor arrangement if I so desired) I
had written, in Silver Sharpie, the letters OC MPMM. For the video I blocked them out
with a Black Sharpie. You are seeing the glare from the differential albedo of the tape vs.
the marker.

Al- Jan 8th 2007


And here would be a good place to point out that I replaced the original 1/4-20 brass
screw and 2 flanged ball bearings pressed into the rotor, with a system that uses 2
smaller bearings, in a housing, with a 3.5 mm shaft, spring-loaded axially to keep
the bearings at zero end-play. This unit is shock-mounted on the baseplate, in lieu of
the brass screw, by some rubber grommets with screws. I scavenged it from a similar
motor to that Andy describes. I had to make an adapter out of Delrin to press into the 1/2
dia. center hole in the rotor. Now the rotor slides on and off the shaft, it's just tight
enough not to slip in rotation. Much smoother and more accurate than the screw/flanged
bearing combo.
Later post
The only thing I would add or change is the rotor axle. I know it's what I showed in most
of the pictures. But if you look at the later ones, you will see a 3.5 mm shaft instead of
the 1/4-20 brass screw, and the 1/2 in hole in the rotor has been "plugged" with a Delrin
insert to adapt to this shaft. The shaft is mounted in a dual ball bearing holder with a
spring compression end-play eliminator, and the holder in turn is mounted on the base, in
lieu of the brass screw. The rotor now simply slides on and off the shaft. For upside down
runs I secure the rotor with a little retaining collar with a setscrew.

Thanks in advance.

Al- Jan 9th 2007


Two tips:

I wouldn't rely on glue to hold magnets to plastic or other metal; things can spin fast and
you don't want a strong magnet coming loose at 5000 rpm. Guess how I know this stuff.

For those who are machining acrylic plastic: water is the best coolant/lubricant but who
wants to get water all over the tooling; many commercial cutting lubes will wind up
crazing the plastic; I use WD-40, it works great. Clean kerosene is good too. Wash with
warm water and dishsoap when you're done. Also alcohols like methanol or ethanol will
really craze and weaken the plastic, so be careful what you clean it with. Meguiar's
Mirror Glaze is the best, in 2 grades, polishing and cleaning.

Clanzer- Jan 9th 2007


Quick Movie showing my Stator Mounts.

Used the small 6mm Stator Magnets too test if loose enough, if they will spin with these
small magnets then I am more than confident the bigger ones will be okay.

Forgot to add in the video that I placed a Nut on the end of the 3mm rod and did it up
tight do it press fitted the V-lip into the bearings, well solid :)

Am wondering whether to press fit the rod ends directly into the Magnet Holes rather
than cutting out a mount in the Polycarbonate Rod shown in video.

< http://www.overunity.org.uk/ocpm/CLaNZeRSStatorBearing.wmv >

Al- Jan 9th 2007


@RB: you might be able to find some plastic tubes, like at the hobby shop or the garden
store or the hardware store, that has an exact or close to 1/2 inch ID. Then you could just
use some kind of improvised spacer to go between the magnet and the bearing, inside the
bit of tube. This would almost duplicate my Delrin holders but without requiring
machining.

Also most drills will drill Delrin and Nylon slightly undersize. So if you can find some
solid Delrin rod (the best) you could simply drill it out with a 1/2 inch drill bit, and it
probably would be enough undersize to provide the correct tight fit.

Finally, Small Parts Company. http://www.smallparts.com/?020


And Small Parts Mfg. Co. http://www.smallpartsmfg.com/

korkscrew- Jan 9th 2007


Crystal clear. Expected even. It happens with any phase locked system. It also suggests
some kind of resonance which would be why the system would prefer a narow range of
speeds and fall apart quickly as it moves away from that range.

That's why I think the relationship between the stator mass and the force of the magnets
is critical. I expect most of the replications to fail due to not getting this relationship right.
I'm thinking you got very close by accident.

What it doesn't quite explain, and what I was hoping the simulations might reveal, is the
source of the energy for spin-up. It seems it must be a balancing of the forces as the rotar
and stator synch up. Your observations also suggest that there are several synchronous
modes and only one produces the anomalous spin-up.

It would be incredibly interesting to have some visualization of the relative motions and
how they change during spin-up.

hdeasy- Jan 9th 2007


@Alsetalokin
I think Gdaigle might have seen my gravity theory on another thread - I likened it to a
satellite getting a gravity assist in a slingshot to the outer solar system: that takes energy
from the planet's rotation or rahter orbit. It struck me that this might not only explain a
gain in your sim but in the normal static Orbo: Prediciton there - run the rig on a surface
suspended from torsion wires with strain gauges etc. and measure how the stresses get
transmitted if at all. It's a bit vague and wooly. But there you have it. Might even relate to
OC's Kekule-like vision of gyrating hurricanoes ("strike flat the thick rotundity of the
Earth!" - King Lear)

Later post
It's quite clearly labelled - the horizontal axis is always time in seconds and the vertical
axis is always Rotor Rpm. By the way, I will put a few more plots up tomorrow. One
other interesting one is increase of KE vs. time on the same graph as torque.

I had suspected that I hadn't corrected for the KE of the stator - but even with that
spinning up to 5000 RPM, as the weight is tiny compared with the stator, it will only ever
be a third or so of the rotor KE. That's a correction factor of about 0.6666 - gave about
0.6 J at about 350 RPM.

So curiously in the sim, which is fairly Newtonian with just a little movement in the BH
graph a la BR's prescription as shown in umpteen papers on the subject apparently, (not
even viscosity to bulge the linear BH curve) all you need is that 4:1 resonance or synch
and then even if you pay the energy bill for spinning the stator it seems to spin up the
rotor. I'll look a bit closer at my sim tomorrow to make sure it's kosher.

RB- Jan 9th 2007


Lakes:I sorta like (I think it RB`s idea?, not sure now)the idea of using loaded coils (a
coil with a variable resistor connected) to somehow fine tune the thing to keep it running
in sync longer.

My idea has the dampers replaced with a coil, the coil effectively becomes the
damper, but keep in mind that the aluminum dampers are very low resistance in
the order of mOHM's.

Al- Jan 9th 2007


Re static charge: as some of you might know, SE is a fascination of mine, and I have built
many SE generators and motors and other devices in my time.
Suppose I were imparting a HV SE charge to the device--not by simply spinning it up,
but, say, just for argument's sake, that I was insulated from the floor, and was hooked up
to a 500 kV van de Graaff machine. The charge would be harmless to me, but could
easily be transferred, and stick to, the dielectric materials of the device. If I imparted such
a charge to the device's parts, is there a possibility that this could account for the unit's
behavior?

Now, the VDG is an extreme example, but around here on a very cold dry day it is really
easy to get a charge of 20-50 kV just by taking your parka off.

Did anybody bother to view or cache my OTHER videos on youtube, before I took them
all down? I illustrate the power of SE, and the ease of obtaining it, in several of those.
No? I didn't think so.

And another thing, on the ning site, I posted several nice artistic renderings of Julia-sets
and associated chaotic dynamical strange attractors. I'll bet most folks thought those were
just pretty pictures.
Well, I recommend going back and looking at the ball-in-a-bowl over the air-powered
Tesla turbine with spinning magnets on it, video that someone linked earlier. Notice the
path of the ball....
It is a classic chaotic dynamical system, following a strange attractor path.
Now here's the clinker: in a chaotic dynamical system, while it is fully mathematically
determined, a tiny change in initial conditions can make a huge, even qualitative,
difference in the output behavior of the system.

Someday folks might even realize that the Conservation of Sham'poo thread isn't what it
seems, either.

Later post
Every time I have actually seen it stop, it does so by the "dropping out" of the spinning
stator. It seems to run at an almost constant rpm until that happens. Spontaneously? Due
to some perturbation? They turn off the Jacuzzi upstairs? I don't know.
But I can say that the slow linear decay to zero rpm, isn't the way it seems to behave. Of
course I do not have any data at all past about 3 1/2 hours, and I expect it to be able to run
for about 1 day 15 hours, or 39 (say 40) hours, thanks to the math models that have been
worked out, and based on my conjecture that there may be as much as 17000 Joules
available for momentum change, per run.

As for the acceleration when I stop the 2 "idler" stators: I feel certain that this is due to
the reduction in drag. It takes energy to rotate each part of the unit, clearly. When the
stators are stopped, the energy that was expended in keeping them rotating is now
available to the rotor, and it speeds up until again, the power output is equal to all drag
and friction sources.
Like I said, the acceleration in spins isn't anything to get excited about. It's the normal
response of a spinning system to the reduction in overall drag.
@gaf:
I realize you joined us late, so you probably aren't aware that the dampers are there
precisely to slow the stator (I think) a little bit and to help stabilize it on its bearing.
You are indeed correct: the unit runs faster without the magnokinetic Judson dampers.
Typical stator speeds are 8000 rpm, rotor 2000. But the unit is quite unstable at those
speeds, maybe due to the cheap bearings I used for the stators. But that's unknown, it
could be that the bearing chatter actually contributes to the effect by some chaotic
dynamical interaction. Don't ask me with what!
Anyway the unit will run up to high speed but only stays for a short time before the stator
"drops out" and the rotor freewheels to a stop. I figured that a little eddy-current damping
at high speeds (and proportionally less at low speeds, so the damping at startup speeds is
essentially zero.) would do the trick. There was an adjustment process that took a couple
hours to find the sweet spot and the orientation, and it seemed so close to an already
drilled and tapped set of holes in the base, that I just used the existing holes, even though
the positioning wasn't totally optimum. It seems to do what I intended. I never anticipated
the need for these dampers when I was laying out the unit.

Al- Jan 9th 2007


riterX:@Al
I am trying to replicate your unit as closely as possible.

Where can I get this?

"a system that uses 2 smaller bearings, in a housing, with a 3.5 mm shaft, spring-loaded
axially to keep the bearings at zero end-play. This unit is shock-mounted on the
baseplate".

I have already dismantled 6 CD ROM or DVD drives some dating back to 1998.

I think I got it from an old tape deck capstan drive system. Reel-to-reel, remember
those?
I am afraid you will have to go a bit further back in time.

Or just make one. Make a flange, like a plumber's floor flange. Extend a tube up
from it, mount the bearings inside, stick a shaft in there, put a collar and a thrust
washer on the bottom extension of the shaft, put a washer, spring, and another
collar or circlip on the shaft just above the top bearing, screw the flange down
with grommets and screws, slide the rotor on to the shaft. You will have to detail
the ID of the rotor so it doesn't drag on the bearing tube.
from Jan. 12th post

The assembly to which you refer was scavenged from a capstan tape drive system. I
described how I would fabricate one from scratch if necessary, but for a couple bucks at
the surplus store I saved myself 4 or 5 hours work.
The screw in the baseplate and bearings in the rotor was sort of ok but it wobbled too
much for my taste.

Jag- Jan 9th 2007


Oooh that gave me a question to ask AL,

We know the speed prior to the AGR (Anti Gearwise Rotation) we know the speed post
one syncronisation has occurred. pcstru4 make a very valid point that we can calculate
the rate of acceleration if we know what period it takes to ramp up to a stabilized rotation.

You've stated in response to another question that there is "torque" during the ramp up
phase. (hello blistered fingers)

1) can we extend the period of acceleration by placing a load on the stator?


2) If there was a load on the stator would you need the J dampners?

I just had this image of one of those little bicycle lights that you used by attaching to your
front tyre with a little flywheel to generate a light as you rode along. Sure hell of a lot
more torque on a bycycle wheel, but I also never got my bike tyre to 8000 rpms

Al- Jan 9th 2007


The stator rotation is very sensitive to perturbation.
That's a fancy way of saying if you mess with it too much it quits.
I would put a load, if any, on the rotor instead.
If it is accelerating, of course theres a torque. I don't recall saying that, but I agree with it,
sure.

-"We know the speed prior to the AGR (Anti Gearwise Rotation) we know the speed post
one syncronisation has occurred"

I would say that only the latter is known accurately. I don't actually know with precision
when the stator synchs. I do know it can be between 400 and 200 rpm.
I just don't have enough hands to measure all that simultaneously and still start the silly
thing.

Harvey- Jan 9th 2007


As far as the speed differential is concerned we can derive this accoustically from the
video so long as the time base has not deviated much. If your ears can hear it we should
be able to put it on a scope.

MrEntropy- Jan 9th 2007


Here's an easy experiment to help determine where the acceleration is coming from:
- put it on a turntable (you're old enough to have one of those :-)

- sync it up

- let it go and watch it accelerate.

If the machine is being accelerated by an externally applied torque (electromagnetic or


whatever), then some of this torque will be transferred to the turntable through the rotor
and stator bearing friction, and the turntable should start to turn in the same direction as
the rotor and stator.

If, on the other hand, the rotor and stator are being accelerated by an internally generated
torque, then in order to conserve angular momentum (since the rotor and stator are
rotating in the same direction!), there should be a significant reaction torque transferred
to the turntable through the base of the machine, and the turntable should start to turn
pretty quickly in the opposite direction.

Either way, it narrows down the candidate effects.

plasmad- Jan 9th 2007


When you use a flash, a typical camera will run at least a minimum of a 1/60 shutter
speed to remove any motion blur cause by hand movement. It can go faster according to
focal distance or other light in the enviroment. Hard to tell on this particular picture as
there is glare off of the DVD cases resulting in a +3 or +4 EV white blotch in the center
of the image and off of the white marking on the rotating stator. I really want to film this
device with one of my HD Cameras or on Super 16mm film. :D
Al and OC -
Kent- Jan 10th 2007
In the 24 hours since I have become aware of your accomplishments, I've been able to
grind through about a thousand posts on the matter, and wish to toss in an idea on your
product development.

I've been fascinated with stuff like Fibonacci sequences and Golden Ratios, et al, since
long before the forum began.

The following quote is interesting to me, and I hope, to you.

"The best clues for a possible dynamical cause of phyllotaxis came not from botany but
from experiments in physics by L. S. Levitov (in 1991) and by Stephane Douady and
Yves Couder (in 1992 to 1996). The experiment by Douady and Couder was particularly
fascinating. They held a dish full of silicone oil in a magnetic field that was stronger near
the dish's edge than at the center. Drops of a magnetic fluid, which act like tiny bar
magnets, were dropped periodically at the center of the dish. The tiny magnets repelled
each other and were pushed radially by the magnetic field gradient. Douady and Couder
found patterns that oscillated about, but generally converged to, a spiral on which the
Golden Angle separated successive drops." (The Golden Ratio by Mario Livio, 2002)

The Golden Angle is about 137.51 degrees.

I wonder if you are willing and/or able to place the rotating stators at 0, 137.51, and
275.02 degrees, or thereabouts and to spiral them out a bit...

pcstru4- Jan 10th 2007


More fun with numbers -

After stopping two stators the rig accelerates. Let's say it takes 10 seconds (it's that sort of
order from the video). In 10 seconds the rotor gains 4.83 J, so around .5 joules per second
(watts) on average is being delivered to the rotor during the windup time.

Now Al's theory (not that I know the detail) says that at that point the input energy is
being balanced by the normal retarding forces - drag, friction, eddy currents etc. So let's
see.

The rotor takes about 120 seconds to wind down from 1200 rpm (these are not al's exact
numbers - but should be close enough). So it loses it's 5.43J of KE over 120 seconds, i.e
if the loss is constant, then it's losing 0.05 joules per second, not 0.5 joules per second.

Neo- Jan 10th 2007


I wouldn't expect any depletion in the current set up. The materials stability in terms of
shape, coercivity and airgap will make them virtually impossible to lose any power. Only
elevated temperatures above 120 deg c and high, stray, electro-magnetic fields in the
order or 2 Tesla reversed through the sample will relax the domains. The action of the
stator's alternating polarity on the rotor will have no effect in terms of losses.
Anisotropic Sintered diametrically magnetised NdFeB is relatively new in terms of
particle alignment during the pressing cycle in that phase. Isotropic Bonded NdFeB
materials (10 mgo) are more common for applications requiring this type of orientation.
This may be relevent in terms of power to weight (torque) ratio of 35 + mgo materials.
Anyway, well done so far, I'm sure the effect will help with continued study in this area.

There will be Bonded Neodymium version / experiment happening in the next couple of
days.

br- Jan 10th 2007


jcims: Not sure it's good for anything, but it's fun to look at:

http://img256.imageshack.us/img256/7133/spectrumqo7.jpg

looks like something I was going to do, only better!


Tried to measure the dominant frequencies at the right hand edge of the graph by
noting the pixel numbers. I got

Lower curve = 175 Hz = 1662 RPM


Mid curve = 867 Hz = 8279 RPM
Upper curve = 1432 Hz = 13675 RPM

The lower curve seems to agree with Al's number, but the mid curve (presumably
the stator fundamental) is 8279/1662 = 4.98 times the lower curve. A factor of 5
instead of a factor of 4.

The upper curve is a factor of x8.2 the lower curve, but what is at this frequency?
It's a very strong harmonic. Could be the stator rattling twice on it's bearing, as it
is first pushed then pulled to the rotor as it rotates?

Added (@br - there are 8 cut-outs for the rotor magnets, which might
account for this. )

Also, looking at the banding at the left of the figure - the 'dominant smudge' here
also seems to be pretty much the same frequency as the clearer line at the right of
the figure. So it seems the 'acceleration' is homing in to this natural frequency,
which is already present?
Later post
@Al,

Also a note on the two 'idle' stators. You mentioned briefly that when the rotor is
spinning and they are stationary that they interact with eachother. This sounds like a
viscosity effect to me. As the rotor passes the first stator, the rotor magnet gets
magnetised slightly. Then as it swings around it holds on to that magnetisation due to
viscosity, and interacts with the second stator. Even though the rotor mags are oriented
longitudinally N-S-S-N, they all get the same lateral magnetisation due to the idle stator.

Does the effect hold just as well when you tweak the second stator - does the first one
move?

Also you mentioned they interact in 'repulsion'? If the viscosity explanation is true, I
would expect the stators would have the same sides pointing in to the rotor.

It could also have something to do with the system being easier to start when all three
stators are in place, as you mention, though that is harder to see.

@br - 'Lower curve = 175 Hz = 1662 RPM'

Interesting figures, but how are you converting from Hz to RPM?


wow, you're right. I have been so used to converting angular frequency (radians
per second) to RPM (factor of 60/(2*pi)), I just used the wrong formula.

So that makes the equivalent RPMs

175 Hz = 10500 RPM


867 Hz = 52020 RPM
1432 Hz = 85920 RPM

Now I'm confused again - these don't seem to correspong to anything (I liked the
wrong values better!)

Maybe jcims can comment on what his software is plotting? (Or I might have to
do it myself after all!)

vibrator- Jan 10th 2007


Just drawing doodles of the geometry of the interation... wondering if 4 stators in a square
would seem to optimise what's currently happening. I think 4 would sync easier than 3,
because at 4:1 each stator turn interacts with two rotor mags.

This would also even out torque distribution, perhaps increasing stability...?

ETA: of course if it is 4:1...

Al- Jan 10th 2007


@k: An interesting thing happens when I manage to spin the rotor extra-fast by hand on
starting. If I can get it to 500 or so, then flip the stator, sometimes the stator will synch,
but it's going too fast or too slow by a bit, so it sort of rocks back and forth in speed until
both units are fully synched. Am I making myself clear here, it's kind of hard to describe.
One can even hear it, thanks to the chattery bearings. It goes
HmmHmmHmHmHmmmmm,,,,
as the stator magnet "phase locks" with the rotor. I imagine it is happening in the rotor
too but since the moment arms and the masses are so different, the eye (and ear!) only
pick it up on observing the stator magnet.
I think this observation supports your main point, and also this kind of thing is why, I
believe, simple belts or gears are not a good idea. Any mechanical linkage would have
to have a sophisticated method for allowing the phase to vary as the stators come up to
speed.
I think.
Of course this, I believe, is what OC saw in his vision: the mechanical latches, and the
rotor magnet pivots.

Al- Jan 10th 2007


But I do not recommend using polycarbonate (Lexan )for any part of the machine--it's in
the wrong place on the triboelectric series, if one of my theories is correct.

Nor do I recommend acrylic plastic (Lucite, Plexiglas, Perspex?) for the rotor, if the base
is acrylic. Different materials should probably be used. I used HDPE which at the
opposite end of the plastic tribo series from acrylic.

Grimer- Jan 10th 2007


WhiteLite:
Grimer:There certainly will. How else is the "free" energy
going to be generated but by a BH power loop formed by
the interaction of the stator/rotor combo. They are taking
bites our of each others knees in the second quadrant.

I haven't worked out how the AGW stator affects the rotor magnets
yet but if I do I will bear that in mind.

Take two second quadrant knees. Flip one of them twice so they overlap -
hdeasy- Jan 10th 2007
Well, I did the BH plot for my sim. Even without viscosity I get a loopy BH curve,
where for static stator got straight line in the 2nd quad:

Also, note that I only use one stator and one rotor magnet as before, but still get a spin-
up. Thus I think all those extra magnets (8, 3 or whatever) are irrelevant - the key issue is
the spin on stator.

The BH curves, along with plots of B vs time and H vs time (see plot attached here),
show that an effective lag is introduced, though there is no viscosity modelled here. The
lag is in the region of 0.01 sec - i.e. much longer than before where microsec were the
longest
considered for Sv. This would seem to explain the KE gain seen.

http://www.geocities.com/deasyart/steorn/OCAL_B_H_vs_t.JPG

And to remind: the corresponding plots of spin up (RPM of rotor vs. time)
are in the plot below, where different plots just focus on more detailed regions
of the RPM-time graph;

http://www.geocities.com/deasyart/steorn/OC_sim_one_stator_a.JPG

Well, as I said to BR on the SPDC already - sorry! There WAS indeed viscous lag in my
sime before. I took it out though, after being made to realise my mistake by BR but now I
STILL get the same energy gain and spin-up, though the BH curve is different,
namely as below:

http://www.geocities.com/deasyart/steorn/OCAL_BH_curve_nolag.JPG

Tomorrow I want to get to the bottom of this. I'll focus on 1 or 2 revs...

hdeasy post from the following day

I confirmed that all lag was eliminated from my sim - the parallel lines in the BH curve
were just artifacts of the precision of my printout - when
I added 2 decimal places this disappeared and I had a perfectly straight line - no area
swept out in BH. The plot below looks at rotor
speed, torque and B (H looks the same just larger scale due to no lag). It could be that the
gain is due to a slightly
stronger kick going out.
http://www.geocities.com/deasyart/steorn/OCAL_gain_nolag.JPG

vibrator- Jan 10th 2007


As far as i can see, all OU requires is a temporal or spatial asymmetry. What we're seeing
is what we'd expect from Sv (from what we've been led to believe anyway), but OC's
right in that it's not the only possibility. But whatever, if there's an unexpected asymmetry
then we may have to accept the system is a either a KE feedback loop between the EM
and mechanical domains, or some kind of transducer for environmental electric charge or
something...

..an asymmetric transaction across a composite shift symmetry.

Whassat mean? Conservation laws seem to start out from the presumption that
information is covariant between time and space. So it always has one foot in each, and is
conserved between them. If 99% of the info about an entity is in space, 1% of it is in time
and so forth. Energy is associated with transits of information between time and space,
and transformation symmetries show how different fields, forces and charges etc.
transform between these two domains.

So if OU is possible it may well imply a conservation symmetry we currently regard as


elementary can be broken down into two or more symmetries, thus re-establishing CoE.

An example of a temporal asymmetry might be a lossy or gainy Sv reaction. An example


of spatial asymmetry might be a synchronous but asymmetric division of electric charge
(positive and negative or internal vs external etc), MPE, angular momentum (as per OC's
vortices) and so forth.

I don't know yet exactly which symmetry OCAL is breaking, but i believe it's working
because of conservation symmetries, and not in spite of them...

ETA: an example of a potentially composite symmetry might be conservation of angular


momentum...

Morgenster- Jan 10th 2007


That's what I was thinking but the real question is what will it's actual behaviour be? If
the effect Al observed is indeed putting out energy and it's not an artifact your assertion
should hold true.

I suspect that there is not only a limited RPM range within which it can be observed but
also a tight synchronisation required between stator and rotor to get the spin-up and I
wonder what would happen if you would keep these two fixed artificially.
On the one hand by applying a load of 0.5 watts (or any other figure that is correct) and
on the other by linking the stator to the rotor in a fixed ratio as proposed by others.

Maybe I should just post this in the replicator's thread. I'm not skilled enough to produce
a test rig myself and my knowledge of physics is pretty basic so don't shoot me for
making these suggestions if I make no sense at all

Grimer- Jan 10th 2007


I extrapolated to the best straight line fit to the data Al gave on that hour run which took
place on the web and that gave between 3 and 4 hours.

But I'm sure if you put confidence limits on the data you wouldn't be able to say it wasn't
a horizontal straight line. I bet that when a decent rig is built and run under controlled
conditions the relation between power and time will be of the form:
pppppppppp P = zero.t + a constant
ppppppppppwhere P is power and t is time.

Harvey- Jan 10th 2007


pcstru4:
When the load (loss) is equal to the energy in, then the system will be in equilibrium and
the speed of the rotor should not change.
As best as I can tell this occurs between 4700 and 5200 RPM on the 'stator'. This
is the 'sweet spot' where momentum (of both rotor and stator), resistance and
magnetic attraction find equilibrium. As time goes on the losses increase. This is
due in part to very gradual warming of the bearing surfaces and the loss increases
as they expand. Perhaps we should use magnetic bearings

vibrator- Jan 10th 2007


Someone mentioned the mass-H relationship back there (i forget who -Victory?) and after
some experimenting it seems it's an important aspect. One thing it might suggest is
another mechanism for the slowdown - if the 'magic' comes from an asymmetric
hysterisis rate, perhaps the response freq. characteristics of the magnets drift over time...

ryandinan- Jan 10th 2007


alsetalokin:
As for the acceleration when I stop the 2 "idler" stators: I feel certain that this is due to
the reduction in drag. It takes energy to rotate each part of the unit, clearly. When the
stators are stopped, the energy that was expended in keeping them rotating is now
available to the rotor, and it speeds up until again, the power output is equal to all drag
and friction sources.
Like I said, the acceleration in spins isn't anything to get excited about. It's the normal
response of a spinning system to the reduction in overall drag.

This is a very interesting comment - and I think it needs to be examined a bit.


Supposedly, the only energy input into this system was by the initial rotating of
the rotor (and to varying degrees, the smaller stators). I'm sure there has to be a
way to accurately measure how much "startup" energy there is, and at the very
least, a way to estimate it at this point - let's call it "X" for now.

Most of this X energy is used to overcome the various sources of friction in the
system (bearing surfaces, air resistance, magnetic, etc) - Al refers to this as drag. I
think it's important to note however, that all the drag is already present in the
system, even before it is started. Once the unit gets the right combination of rotor/
stator spins and actually "starts up", it is apparent that the RPM of the rotor
accelerates past the level obtained solely by the X energy, and levels off (there is
no way Al spun the main rotor up to 1500 rpm by his hand). Is this caused by a
reduction in drag somewhere? Somehow, somewhere, extra energy is going into
the main rotor to make it spin - I'll explain my reasoning below (feel free to pick it
apart).

When Al stopped the two stators, the main rotor quickly increased in rpm over
2X. If this is caused by the reduction in drag, then that must mean there is some
force being applied to spin the rotor back up. Is this extra available force equal to
what was required to spin those two stators? And more importantly, what is
providing this force? It can't be his hand - and it can't be the kinetic energy that's
already in the system... Is there some other potential energy stored somewhere?
Here is my analogy: It would be something similar to driving a car on a flat road
while simultaneously applying the brakes and the accelerator; the car would
move, but only go so fast, and waste a lot of energy (as heat) in the process. As
soon as the brakes are let up (with the accelerator kept at the same position), the
car would accelerate up to a certain speed, and then stay there. This obviously
requires force imparted by the engine (by burning gas), and is what we see in this
device.

If there was no additional force in this device, then the analogy would change.
It would be more like pushing a car up to speed with the brakes applied. The car
would gradually slow down. As soon as the brakes were removed, the car would
not accelerate. It would simply slow it's rate of deceleration, relying only on it's
momentum and the remaining friction (road, wind, etc) to carry it forward, until it
ultimately came to a stop.

Which analogy seems to better fit the observations?

jcims- Jan 10th 2007


For what it's worth, grabbing the bright fundamental in this chart in Sonic Visualizer, I
show an increase from 152.306Hz to 175.047Hz in a period of 20.040 seconds. Might be
useful for folks that want to do some power calculations (presuming we can determine
the relationship of the fundamental to the rotor)...if i had to guess i would say it's what br
recommended above, the stator rattling twice each revolution. If that were the case, we
would be going from approximately 4570 RPM to 5250 RPM in the stator (and what,
1140 to 1312 in the rotor?) in 20 seconds.

br - Jan 10th 2007


jcims:
br:Any chance of plotting the bottom 100 Hz, and
boosting the contrast? 1300 RPM corresponds to 21.7 Hz.

That lowest fundamental gets lost after it passes some environmental


hum at about 85hz (although it looks like the 'harmonic' or whatever it
is at 88 hz gets a bit of a boost afterwards)

http://img184.imageshack.us/img184/153/bottomzs6.png

thanks! that's great, 88*60 = 5280 RPM, just what we needed for the stator
fundamental.
I cranked the gain way up and tried finding something at ~21Hz, and
there just isn't much of anything there. There's a whisper of something
at ~18Hz and at ~28Hz, but these could just be artifacts.

http://img176.imageshack.us/img176/4921/21rm8.png

indeed. Any chance of increasing the bin-length of the audio software - the longer
the time chunk that is used for the FFT, the clearer the signal will be. Maybe not
worth it, and 20 Hz is at the low end of what mikes and sound cards are good at
anyway.

Any other frequency components have to be harmonics of the 88 Hz and 22 Hz


signals, but that gives a lot of freedom to assign which harmonic is due to which,
especially when the dominant harmonics are in the kHz range!

In harmonics of 88 Hz, we have


175 Hz = x2
867 = x10
1432 = x16

not sure what they refer to though, especially the x10 one. Could be 21.7Hz
x40 ???

What gets me at this point is that the acceleration at the right of the figure to the
three lines of 175, 867 and 1432 Hz are still at the same frequency as the
'dominant smudges' at left of the figure. Something to do with natural frequencies
I guess.

gafnaz - Jan 10th 2007


Assuming 2 minutes run for 'normal' configuration and like 30 hours run for 'OCAL'
configuration, the difference is 1000.

When the 2 minute run finishes some between 10 and 20 Joules of kinetic energy are
dissipated. Accordingly the 30 hour run would use between 10.000 and 20.000 Joules
(Alsetalokin mentioned 17.000 Joules).

20.000 Joules is the equivalent of 10 penlite AA sized batteries. That is a whole lot of
energy.

Al,
I'm very interested in your theory about the source of this energy and how you calculated
the 17 KJ amount.
NB: do you have a pace-maker? what type of mobile phone? do you have a big watch,
made in Russia?

Al - Jan 10th 2007


That sounds about right, or at least ballpark. Although in the vid I seem like a total dolt
when I'm trying to get a stator rpm reading, the max rpm is usually 1250 rotor, 5000
stator. So the 175 Hz numbers make sense to me.
I don't know about the higher freq. bands. I am puzzled too about the 2 broadband noise
bands that are there from the beginning and then seem to "coalesce" into the formants of
the bearings (if indeed that is what they are.) I just downloaded a bunch of software from
them, thanks jcims for the link. I've been trying to find shareware for these things for a
long time.
My impression is that the noise bands have nothing to do with any effects, and the
software is pulling the power into the bearing formants in error.
But I haven't played yet with the software so I don't really know.
This is good work, though, and it shows that sophisticated analyses can and should be
done, if the data is available.

OC- Jan 11th 2007


@Al,

When I picture Kent's golden angle in my head full of counter-spinning vortices:


If the golden angle has anything to do with this I would expect 1 or more of the
following: a) easier to sync stator; b) greater torque while spinning up to high RPM; c)
higher top end speed

The high-end speed is a little shaky in my mind. We might need to adjust the angle for
optimum speed.

Grimer- Jan 10th 2007


alsetalokin:
Grimer:One could easily calculate how much
the effect mass of the rotor had increased and
test it experimentally by having a hole in the
rotor through which one dropped a steel rod
and saw what the difference was between the
OCAL rotor and a rotor turning at the same
speed with non-magnetic ballast.

One would be advised to stand well clear. A


good experiment for under Sean's stairs.

I don't follow what you are suggesting. Could you please


expand?

I don't want to lead you down a false trail. I shall have to think about it
and review Laithwaite's demonstrations and the way precessing gyros
behaved.

If one can polarise inertia it opens up the possibility of a gravitational


mill where the "mass" of a ring on the downward side will be greater
than the "mass" on the upward side.

They will have to make you Lord of the Rings.

I'll sleep on it.

Al- Jan 11th 2007


I'll lay it out on paper tomorrow, if I can get my actual work problem straightened out
(literally).

WhiteLite- Jan 11th 2007


Magnetic viscosity is the time delay of movements on a BH curve.

hdeasy- Jan 11th 2007


Viscosity for magnets is the sluggishness of the domains in slewing around to adjust to a
strong field imposed by an external source. It means that the B field of the weaker
magnet will follow that of the stronger one, but with a delay of, typicaly, micro- to milli-
seconds.

Just Maybe- Jan 11th 2007


DO you remember the old google spreadsheet days with the fast in mu metal?

I suspect that we will find that repulsion is the key since a "fast in" (above 1600 rpm for
stator) will allow the stator to sneak in at lower force to the rotors field as north faces
north, but as the stator rotates and the souths now face off, the rotor magnet has had the
time to notice the stators repulsive nature and says, let's get the heck out of here...
QUICK and the full force of repulsion kicks in. The rotor accelerated away and the
stator receives an equal, but opposite force, which kicks it's anti-gear rotation into a
faster spin, sufficient for it to meet the next magnet at the right time to repeat this process
of acceleration. It all maxes out when the bearings and air resistance have had enough of
the funny business

korkscrew- Jan 11th 2007


This morning in the shower, I was pondering this contraption that alsetalokin built, trying
to reduce it to it's simplest form. I have often found reductio ad absurdum a useful
exercise when I'm trying to understand something that should be simple, but seems
complex.

In it's simplest form, this thing is two magnets rotating in the same direction (AGW). One
(corresponding to the rotor) is rotating at a more or less constant rate. The other (stator) is
rotating at a variable rate. While they both have the same net RPM, the stator is
alternately speeding up and slowing down over the course of one revolution

So rather than build and study the same thing everyone else is. I think I’ll build a rig that
basically looks like two of the stators on Al’s rig. But I’ll add an axel to each so I can
attach flywheels of different weight, and I’ll make the distance between them adjustable.

I don’t know if I’ll learn anything new, but it will be way easier to start. Pull string
should work marvy!

blueletter- Jan 11th 2007


In summary...
alsetalokin:For some reason it seems to be working rather well
tonight. The rotor rpm is 1334, the stator rpm is 5378. It was difficult
to start, it took maybe 20 flips of the stator magnet in all before it
"caught". But I am developing a thumb technique that is easier than
the finger.

I started it at 02:14. blueletter note: 11 Jan 2008


It is pretty noisy with the bearings chattering and growling, but I think
I might be able to sleep.

alsetalokin:Shut down by stopping stator mag at about 09:43


RPM just before shutdown: Rotor 1257, Stator 5030
Al's sleep time = 0

7 hrs. 26 min.

vibrator- Jan 11th 2007


@OC - re. reconnection, was looking at Slashman's FEMM anim....

...and noticed this - it appears the fields may be reconnecting to the outer side of the
stator, such that the closing reconnecting field is pushing the stator, and the exiting
reconnecting fields are pulling it, both torques going in the AGW direction.

IOW, the fields themselves may be reconnecting in the gear-wise direction

Al- Jan 11th 2007


First the misconceptions. The rotor has never gone over about 2100 or 2200 rpm in any
testing I've done. It is important to distinguish between the rotor and stator speeds. In the
current config the max rotor rpm is 1250-1300 or so. Small discrepancies in such things
as my measuring technique and the tach calibration will result in the computed ratios
based on the data, to vary somewhat from exactly 4:1 as Frank has found. Is this
unexpected? I hope not.
The exploded view is not the 13x8 basic unit, of course--most all who are following
know that it is an exploded view of the design of one of the first test beds--the 5x4
version. Do they think I just built what I've shown here, and voila it works? There are
several test beds and versions.

Now for new stuff. I scanned the basement kitchen with the TriField meter, and I can
detect a fluctuating magnetic field of 25-100 milliGauss all over the kitchen. It is
especially strong ( > 100 mG ) in the area where I usually run the tests. This is without
the unit running, with it going I need to back completely out of the kitchen before the
meter reads below 50 mG.
The meter tells me that 2 or 3 milliGauss can begin to cause biological effects. The meter
shows fluctuations, I guess about 1/4 second time scale and irregular in magnitude. No
appreciable electric or RF/microwave fields were detected. But there's a lot going on in
the EM for sure.

I was able to get a 12-to-1 resonance going. With the unit operating, presumably on the
one stator, I spun another stator magnet really fast gearwise, and it caught, and ran very
smoothly, I could tell the stator was spinning faster than usual, and I was amazed when
the tach confirmed it. 2 examples: rotor 1050, stator GW 12586. Rotor 1040, stator GW
12454.

I put setscrews in the rotor shaft adapter and on the GWS 1080 prop to eliminate all
chance of slip. I ran the unit, clockwise, and the rpm stabilized at 909, idlers stopped. I
can't get a stator reading easily with the prop in place, but assume 4:1, and AGW for sure.
I stopped the third stator, and the unit took 29 seconds to come to a complete stop from
909 rpm with the prop attached.

My fingers hurt.
I need a night off, so I may not post anymore tonight.

Svein, some of the pictures are taken with a flash and show the motion frozen. I carefully
examined each stator with the magnetic field viewing film that K7J sells, and the Bloch
wall is very distinct and right down the middle of each magnet, and my markings are as
precise as I could do it. (EDIT On the stators, that is. The rotor magnets are just identified
as to polarity, I made no effort to identify the Bloch wall location in these, so please don't
start reading significance into the variations in markings on the rotor magnets.)

Harvey- Jan 11th 2007


My hypothesis is: Since permanent magnet flux density is is reduced by 1/r^3 the points
along B are subject to doppler shift. This shift creates a relative flux compression
between approaching magnets and a relative flux decompression on retreating magnets.
The differential between these two densities can become significant enough to overcome
load resistance. In other words, two opposite passing fields will have a stronger attraction
when approaching due to the increased relative density and a weaker attraction when
retreating due to the decreased relative density. If this hypothesis (which I admit I made
up) is correct, then it explains the phenomenon under discussion.

Dirtfarmer- Jan 11th 2007


Expanding on Harvey's doppler theory,.
Although it may imply in some sense that speed alone is key, The geometries of this
configuration expedite this theory immensely, We have a fantastic closing speed
(afforded by relative counter-rotation) and an appropriate angular entry and exit, possibly
making the "shear" (Harvey) easier... (The effect would be far lesser, theoretically with a
much larger "stator", due to angle and duration of exit...)

I very much believe Slashmans' Visimag to be correct (sans Doppler effect), proper to
ordination...

Al- Jan 11th 2007


I got this far, all ok but I need to clear this one up again I guess. The position that the
rotor is in is critical for the way it responds to the stators. I can't really tell any difference
in the rotor's behavior with any of the magnets when the parts are manipulated by hand.
Clearly in the video I get the best response from the (edit:rotor) from the leftmost idler.
And yes, the Middle stator does flip easily--it's on the best bearing of all 13. But it is jut a
coincidence that the rotor wasn't in just the right position to respond well to the stator
when I was wiggling it. I can hold any one of the stators, turn it to just the right angle,
and from there, if the rotor is also just right, I can get the rotor to move around quite well
as shown in the video--with any of the stators.

boogie woogie- Jan 12th 2007


Harvey, where did you get this from "Since permanent magnet flux density is is reduced
by 1/r^3 the points along B"...I would like to see a reference towards this quote if you
have one in particular for '1/r^3 '?

Harvey- Jan 12th 2007


@ boogie woogie
Inverse Cube Law

Here's a kiddie link:


http://image.gsfc.nasa.gov/poetry/activity/l23.pdf
cyrilsmith- Jan 12th 2007
In response to boogie-woogie’s question about the 1/r^3 law Harvey quoted the following

Here's a kiddie link:


http://image.gsfc.nasa.gov/poetry/activity/l23.pdf

The guy who wrote this should have his knuckles rapped (and if he teaches magnetics,
which seems to be the case, they should be rapped even harder!). You will find a
description of the magnetometer at
http://image.gsfc.nasa.gov/poetry/activity/l21.pdf
where it is seen to be a suspended bar magnet with its axis horizontal, having a mirror
attached so that tiny angular deflections can be measured using a light beam. It is
designed to measure changes in the vector direction of the local earth’s magnetic field.
Note that it does not measure amplitude.

In the kiddie link given by Harvey the suspended magnet is made to turn from its stable
position by having a relatively large lump of soft iron close to it. The author found that
the angular deflection of the suspended magnet was approximately related to the inverse
cube of the separation distance between magnet and soft iron. He then claims that as
evidence that “magnetism follows an inverse-cube law”.

The so-called magnetometer was really just a field-direction meter, it did not measure
field amplitude. It could be used to determine field laws using the local earth’s field
amplitude as a reference, but the field to be compared must be normal to the local field.
Then for small angles the angular deviation is proportional to the additional field
magnitude. However the experiment did not follow that procedure, the device compared
not linear force but the torque on the magnet due to the nearby soft iron. The actual field
direction that it indicated came from the vector sum of (a) the local earth’s magnetic
field, (b) the field from the soft iron as magnetized by the earth’s field and (c) the field
from the soft iron as magnetized by the proximity of the magnet. Because of the vast
difference in size between magnet and soft iron, range laws from (b) and (c) would be
quite different, so even a true magnetometer reading would not directly indicate the “laws
of magnetism”. Had the author included other variables (such as the size of the soft iron)
he would have found his inverse cube relationship not to be a universal law.

FWIW the “range laws” of magnetism can be almost anything you want depending on
where you do the measurements and over what distance. The 1/r^2 law applies to isolated
point poles, which do not really exist but can be approximated by long thin bar magnets.
If you get close to a practical pole face the inverse square law doesn’t apply, it can be
1/r^1 and even 1/r^0 when you get very close. At distances large compared to the
separation between the poles the bar magnet can be considered as a dipole obeying the
1/r^3 law. Somewhere between the distant 1/r^3 region and the close 1/r^0 region there
will be a region where say 1/r^2.5 applies, or 1/r^1.3, or 1/r^2.7,----you choose it
Harvey- Jan 12th 2007
I believe the Inverse Cube Law relates primarily to normalized fields (more spherical)
beyond the pole cantenoids. The actual density curves of a magnatized object more
closely resembles a ring toroid and the inverse cube center should be very close to the
minor radius center. In some cases we may find the field laps into a spindle toroid or even
resembles a near sphere throughout. Obviously with such a large variety of field shape
interactions applying the inverse cube law in these regions is impractical. My reference to
this law is simply to show the gradient exist and as such is subject to doppler shift.

Terry Lingle- Jan 12th 2007


AL : I have enough nylon to machine a rotor and base plate , More than sufficient
bearings for this and the equiptment to build it . I have not started machining yet as I have
a couple of questions.
1 do you feel that the choice of base material and rotor material is critical?
2 Do you suspect that the spacing between the rotor and base is important ?

Al- Jan 12th 2007


Nylon is about like congealed asphalt as far as machinability goes. It is usually molded to
shape.

The Delrin (or Acetal in the civilized world) is what I used for the early rotors and all the
stator bearing/magnet holders, and the 3.5 mm to 1/2 in shaft adapter. It is by far the
nicest plastic to machine. It has a sensual quality about it that is somewhat indescribable.
Why, when I first met Delrin...but that's another story.
Oh, and the mechanical properties are nice too.

I don't think the base-to-rotor distance is important, any more. I did until last night. My
mount doesn't allow me easily to vary that distance so I haven't experimented directly.

But, and please everybody note this well, I have a startlingly intense electromagnetic
radiation situation in my primary work area in the basement kitchen. I discovered this last
evening. My meter shows at least 25 milliGauss, and in most places 50-100+ mG, all
over the kitchen, especially from the wall (which contains, as I have tried to emphasize,
the power distribution panel for the house) near my tiny workbench. The high range of
the meter pegs out at 100 milliGauss!! And as I reported last night, if the device is
operating, I have to get completely out of the kitchen before the field drops below 50 mG.

My first guess about why this thing might do what it seems to do, since I am absolutely
certain that it cannot be anything that violates any conservation laws, is that it works by
the "RB effect". Others have pointed out problems with this theory, and I have not been
able to secure an adequately shielded environment to test it.

The other location where I have observed extended runtimes, and where the video was
shot, is a couple km from a ground-surveillance radar installation.

So, my friends, caveat constructor, but we are most likely chasing artifact.
pcstru4- Jan 12th 2007
What does reduction of flux density have to do with doppler? Surely a doppler effect first
and foremost is about the speed of propoation of the field (supposedly c) in relation to the
speed of the source. Any doppler should be as difficult to spot as sticking a bulb on the
circumference and looking for a red/blue shift in emmitted light.

Harvey- Jan 12th 2007


Doppler shift occurs in any changing value emitted from a moving source relative to a
reference.

@pcstru4
EM is just one example of doppler shift.
Moving your finger up and down in a pool as you move your hand produces doppler
compression in front of your finger. Sirens and train horns are often used to explain
doppler shift.

In this case we are talking about magnetic field density. Since the density changes the
farther away from the magnet we get, moving the magnet creates a change in relative
density. And in simple terms, the greater the density the stronger the attraction

sveinutne- Jan 12th 2007

“I was able to get a 12-to-1 resonance going. With the unit operating, presumably on the
one stator, I spun another stator magnet really fast gearwise, and it caught, and ran very
smoothly, I could tell the stator was spinning faster than usual, and I was amazed when
the tach confirmed it. 2 examples: rotor 1050, stator GW 12586. Rotor 1040, stator GW
12454.”
----------

This is interesting:
I bet the next sweet spot or ratio might be 20:1, so you got 4:1, 12:1, 20:,1 and then
28:1.
Or (1, 3, 5, 7 …) x 4
This indicating that it is one spot that need to be in sync, and between the sync spot the
“stator” magnet can flip 90° or 3*90° or 5*90° to show the same magnetic direction.
It would be very interesting to see more high speed picture taken to find the sync spot,
and compare with the 4:1 and 12:1 speed and see what position that is the most critical.

I guess the power that the magnets are providing is equal with the friction. So at a stable
speed the sync position is moved a little to compensate for the friction. If we put a load
on the system, the speed will go down some and the sync spot will move some degrees
like an electrical motor.
Harvey- Jan 12th 2007
@sveinutne
That is an interesting sequence.

I imagined that we could get a sync on any multiple of 4, however since the stator would
have to make complete rotations during this period there would be drag associated with
the shear and I suppose the rotor rpm would suffer eventually decaying below momentum
to sustain the sync.

But it may support my hypothesis for flux compression in that since the density would
increase with the rotational speed the extra drag is probably overcome by the increased
density. Theoretically then this could continue to saturation. Obviously conservation of
momentum plays a big part in the stators ability to keep rotating at these speeds and
demonstrates a sort of symbioses whereby the rotor gives up potential speed to the stator.
And this may be where Dirtfarmers dwell becomes important.

I think I would like to try a 4:1 belt & pully and see if we can get it sync'd. If so, then we
could do some experiments with 8:1 and so on. Plus, it could save some thumbs

sveinutne- Jan 12th 2007


Harvey,

I think 8:1 will not work if you have eight rotor magnets. Also a belt might be a
disappointment. I think the “stator” magnet is not running at a constant speed. The speed
might vary a lot depending on where it is in the cycle. If you can spin the stator magnet
up with an air gun or something, and then let the system find the sync by itself.
Also I think 4:1 will be easier to find then 12:1 or 20:1.

WhiteLite- Jan 12th 2007


Sterling @ OU.com made this comment which I thought was amusing. Not sure if he is
in the USA or EU:

Doug said that when he was calling the magnet guys they commented something to the
effect that all the sudden people were ordering this one kind of magnet, and they ran out.
They didn't know why there was this sudden rush. I guess no one told them.

Terry Lingle- Jan 12th 2007


My game plan includes a groove for a starting string in the perimeter , pockets for the
rotor magnets and a noncontact thermometer as part of the setup. I may also fill the
pockets with epoxie .
What I want to create is a finger friendly version with an easy to use consistent input set
up and minimal wind drag /noise.
If it works the audio should be much cleaner. starts more reliable and my fingers
remaining blister free.
I hope to learn from Al's mistakes. I will not start machining until I have magnets in hand
as I intend each part to be completed in a onetime set up.
I will probably test spin the rotor to about 6000 rpm for balance and noise checks.
which is one reason I intend to epoxie the magnets into the pockets.

Al
one of the things I am considering is cutting the slots at an angle from the vertical so
the magnet must move against the applied forces to leave the pockets.
Again the epoxie is meant more as a filler than a glue.
I have also pointed out on a different thread that it is possible to machine in mechanical
strength to the glue(filler)/plastic interface

Sveintune & RB - you need more mass on the stators by the sounds of it, i've been
playing around today and the stator needs to carry a fair bit of inertia to reverse-sync. I
tested it ad absurdum using a tiny stator and no matter what, the thing immediately
reverses back to gearwise lock. Conversely, heavier stators seem to get progressively
easier to reverse-sync. Haven't hit an upper limit yet...

korkskrew- Jan 12th 2007


I would think at slower RPMs you would need higher mass, and at higher rpm's you
would need lower mass.

The forces accelerating and decelerating the stator are not changing, but the amount of
time the stator is subject to the force are changing with RPM, so as the device speeds up,
the stator starts to rotate at a more uniform rate. At a slow speed the stator will get VERY
slow at it's slowest point and the system will become unstable.

hdeasy- Jan 12th 2007


On the composite plot below, you can see that as the stator and rotor S's charge at each
other at close quarters, the RPMs sink to a minimum and the negative torque becomes
most negative. But once past each other, where doppler would have the trains go from
scream to baritone, in principle there should be symmtry and the kick got going out
should just balance the resistance encountered going in. But the RPM curve gives maybe
a clue to the assymetry - i.e., have slowed down, the rotor will now interact at the going
out phase for a longer time than going in (scream phase) and so the kick will be more
effective. That's why the torque curve shows a taller positive peak than the negative dip.

See the gain2a image for a close-up of the magnet position sketches, indicating which
time is in question.

http://www.geocities.com/deasyart/steorn/OCAL_gain2a.JPG
http://www.geocities.com/deasyart/steorn/OCAL_gain_composite.JPG

Note that my sim may not be quite comparable with Al's and OC's system - since I have
the rotor magnet aligned radially - I don't plot the N of the Rotor out of laziness - it lies
toward the centre: but it is the Rotor S interactions that are the main ones as shown
Harvey- Jan 12th 2007
@ Dirtfarmer
You are correct. There is a definite relationship between the magnetic mesh of the two
radii. As most will notice during 'gearwise' movement the 'coggy'ness demonstrates a
long tangental slip of the stator between axial magnets where the stator tends to 'flip'
along the Bloch wall.

During AGW operation this slip area is traversed by the stator in a smooth rotation and
shears at the Bloch Wall. Thus the stator field diameter and the rotor field diameter are
important to proper sync.

@ Dirtfarmer

Just a quick eval of the radii: 3.375 and .25

Rotor Magnetic radius 4


Stator Magnetic radius .5

Ratio 8:1

Result: 50% shear providing a 4:1 ratio.


The stator slips 50% during 4 to 1 sync.

Dirtfarmer- Jan 12th 2007


Although I liked this, your math is wrong here Harvey...
magnetic radius 2.625
You quote 3.375 when it should be 2.375 and then add .25 wrong...
Also the stator has a Diameter of .5 , a radius of .25, and none of that includes the gap,
mag to mag...

Harvey- Jan 12th 2007


Regarding Magnetic Diameters:

A. We observe a 4:1 ratio between them.


B. We know physically the spindle distances and size differentials prohibit 'locked'
rotation esp. in AGW operation.
C. There must be slip (aka shear).

My numbers above are quickie ballparks to show just one possible configuration.

EDIT: I took the radii from http://www.ospmm.com/whipmag but my ratio's off the pics
gives some other numbers. It appears from 3 different angles I keep coming up with the
stator mags at .625 if the rotor is 8". Hmmm.
OC- Jan 12th 2007
It IS 8 to 1 IF the rotor magnets are mounted in a chained fashion, all pointing the same
direction. But in the alternating fashion required by my design, they only rotate 1/2
revolution per rotor magnet. Part of the WhipMag mystery.

Harvey- Jan 12th 2007


@Vibrator
Doppler cannot happen unless some mechanism is provided for the
rate of propogation to increase and decrease on either side of a
transaction.

http://en.wikipedia.org/wiki/Doppler_effect#Analysis

I fail to see why the 'propagation' of magnetic density needs to change for the
effect to hold but I suspect it occurs at or near the speed of light so its really a
non-issue.
The important thing here is that the 'relative density' of each magnet on approach
is greater than the 'relative density' on retreat. The frequencies were are dealing
with here is the MGOe/Br * approach speed. This is in the realm of accoustics not
light.

Harvey- Jan 14th 2007


Mock Up:
I've succeeded in the AGW sync of my Mock Up.
Please keep in mind that my Mock Up is powered so I can control the tests I am doing.
Just for laughs I did remove power during one of the AGW sync's and both fan motors
degraded rpm rapidly as expected. I attempted to make some video with my cell phone
but realised quickly that it was set to a 15 second max The stupid thing went off right in
the middle of a rundown in AGW mode. Eventually I got all that sorted out and made
some really bad video that needs edited or scrapped entirely since the lens is off center of
the screen and I kept taking out of frame shots. All of my syncs were at a 4:1 ratio as
detected on my scope.

Of special note is the relative attraction of the stator to the rotor during both AGW and
GW sync. I am convinced the Bloch Walls of both the stator and rotor mags were
syncronized in my tests. I will have to try and rig up some stroboscopic tests to confirm
however. This places the rotor and stator in a very high ratio of attraction. by having a
portable stator I was able to test out different zones. Sync would normally occur at about
1.5" stator field center to rotor field center. The pulse width was about 45ms so the
frequency was near 22Hz or 1320 RPM on the stator and about 330 rpm on the rotor.
When I get my videos done you will see that stator at rest is in opposition to the rotor in
all postions and actually moves away from it. Once sync'd it is just the opposite, it
gravitates toward the rotor and today, I confirmed this effect is valid for both AGW and
GW modes.

vibrator:@loreman - yep i remember your question re. magnets closing at C. The funny
thing is, even within a hare's whisker of C, in free space the field shape remains
symmetrical in front of the magnet as behind it, and the same goes for two magnets in
head-on collision at such speeds. I hadn't thought of Doppler then for the same reasons i
still don't see it now.

It's still a good thought experiment tho.

To continue the experiment: In this case we have two magnets approaching with
opposite poles so they are attractive. The flux density of each begins to interact as
they approach. First we have low density interaction, then higher and higher
density interaction until the poles are tangent and physical restraint prevents any
further approach. If the approach is slower, is the density interaction over time
different? If this interaction is based on the density gradient, then does changing
the speed of approach alter the timing and thus frequency of the interaction? If,
then, each magnet becomes an observer for the other magnets approaching field,
does doppler shift occur for the foregoing? And if so, how would it be
manifested?

vibrator- Jan 14th 2007


There seems to be some confusion re. "Bloch walls" here - i think perhaps started by Al.

AFAIK, the Bloch walls are domain walls - relative inhomogeneities in the lattice, whose
motion nucleates domain pooling as B increases. It doesn't refer to the null zone between
poles.

Harvey, unfortunately the discussion's too close to the bone for me - i just can't contribute
the points i see as relevant and which might support your hypothesis, due to NDA
constraints.

Obviously, we all know Sv limits the energies exchanged, however if Sv is a purely


materialistic property then it isn't going to effect the field shapes ex situ. I can't tell you
there's nothing here worth pursuing... but i can't go into details.

What i will say is that Doppler applies to emmitted signals, which have their own
velocity relative to the source. As fas as i'm aware an object's magnetic field shape is a
static property of the material - a spatial extension of it, so its velocity in free space is
dependent only on the relative source/observer velocities. In other words, if H can
experience Doppler shift, all matter should be likewise amenable.
andromeda- Jan 14th 2007
In a nutshell (and relatively plain English) ...

'Domain' wall refers to a 'demark' between any two domains of any given polarity.

'Bloch wall' refers specifically to the domain wall between (isotropic) domains of
opposing polarity.

Allegedly, theoretically, as per several university level physics programs I studied a year
ago in preparation for the SPDC ... during application of a reverse B field to an
(isotropic) lattice, the Bloch wall between opposing domains will "push" or "shift" as one
domain begins to shrink and the other grows. Up to a point, this effect is reversible, ie if
the B field is removed, the domains will return fully, or near so, to their original state.
Depending on the "hardness" or "softness" of the magnetic material (magnetic hardness,
not mo's scale of material hardness), after a point of criticality, the domains will
irreversibly "flip", ie. adjacent opposing domain(s) will fully reverse polarity and remain
reversed after removal of the induced (applied) B field.

Interestingly, when working with especially hard magnets, such as Nd based materials,
the intensity of field required to fully flip the domains will be relatively high and the
material can actually be driven into the third quadrant of opposing polarity at sub-critical
levels of induction, ie "the knee" of the BH graph falls below the x-axis. In other words,
hard magnets can flip polarity and return their original state without a reverse field
applied.

Everyone with me so far?

Keep in mind that the current or previous state of polarity or magnetization is important
to defining just what constitutes a "reverse" field. In other words, assuming a "poorly"
magnetized or partially magnetized or somehere less than saturated magnet, any field
might be "reversing" as seen by some domain somewhere.

Typically, broadly opposing domains will be found in a previously saturated and/or then
partially reversed magnet...which is not easy to accomplish with hard materials such as
Nd. In other words, if you apply enough field to reverse some domains of an Nd, you
pretty much guarantee an avalanche of all the domains. Hence the "square" characteristic
of the j graph of intrinsic magnetization ...as opposed to merely the induced B field
characterisitic (which works fine as an assumed magnetizationfor softer materials with a
high knee and shallow fall line).

That fall line can also be extrapolated to indicate a time factor leading to your viscosity
issue which some people seem to be trying to get at here, if I understand what they were
trying to say ...

In any case ...


IF the relative field strengths between rotor and stator mag's were high enough, and
considering the fast viscosity or harder magnets, I suppose it MIGHT be possible that
domains were shifting about in these whipmag things and that some sort of time delayed
complex phasing could have some bearing on a compound synch rate ...

but nobody has even begun to do the math or conduct real experiments to test for any
hypothesis that I can see much less backed off a tight framed shot of their rig which begs
all sorts of questions ...

In any case, in all cases, to my ear, it sounded like the stator that was making the noise
that everyone sems to be keying on, not the main rotor. In other words, the relatively low
mass stator being spun up (or possibly maintained) by what may or may not be a steady
state rotor.

At the end of the day, until someone hooks up a tacho and samples it all for real...your
eyes and ears cannot say what is even happening for sure. And I think Al should know
that.

Show me some real data.

hdeasy- Jan 14th 2007


Both Greg with his FEMM sim and I with my own Fortran one have seen the same effect:
I.e. it makes no difference with the spinning stator OCAL- WhipMag system if you have
viscosity switched on or not. Sure - you can have straight line BH movement in the B-H
plot, but no loops a la lag are needed to get a KE gain here.

Have a look at this latest annotated plot of mine - I think it shows rather well how the
dips and peaks in the RPMs correspond to the stator N and S approaching and receding,
at their higher spin rate, w.r.t. the slower moving rotor pole (N in this case dominant as I
have a radial sim and the S is so far away it's negligible).
http://www.geocities.com/deasyart/steorn/OCAL_frames_110.JPG

Just Maybe- Jan 14th 2007


IN the Chat thread RunningBear made this comment, I thought it best in the development
thread.
RunningBare:Just like to add, there is no magic in stator AGW lock,
this is explainable by momentum, as long as the stator has enough
momentum from the last passing rotor magnet it will swing around
enough to catch the next set of magnets on the rotor.

I'm with you there RB, As I see it, once the stators are set to spin AGW and the
right speed is reached, the rotor magnets lightly stroke the stator with there
repulsive fields at reduced force due to lag. The momentum of the stator needs to
be sufficient to resist this reversing force, and once it has rotated the field around
enough it received the full repulsive kick which accelerates the stator away from
the rotor.
This, by nature also accelerates the rotor, as Al found out. :big grin:

COnclusion, The mass of the Stator and the loosness of the bearings is critical..
There is a similarity with the original stop start device of Steorn.
It too required two axis. Sean said so in this public forum may moons ago.
He also said (paraphrasing) that through cunning engineering, the excess joules had to be
ferreted away for a short time (during the stop phase) and then some of them used to kick
start the next movement around the cycle. The excess were available as "free energy".

Is seems that Al's AGW stator is doing just that. Accidental Cunning Engineering

New name for WhipMag. ACE

modervador- Jan 14th 2007


To assess the drag forces in this system, I have analysed the experiment: “stator magnets
removed and far away, initial rotor RPM 1250, rundown time ~ 130-140 seconds”.

Other assumptions gleaned from other posts: rotor mass = 258 grams, diameter = 5 ¾”,
radius = 2.875” = 73.0mm. The moment of inertia I = 6.88 kg.cm^2. At 1250 rpm, the
rotational kinetic energy of the rotor disc is 5.89 joules.

The simplest case is that all drag force is bearing friction. The resulting motion equation
is

w(t) = Wo + Ao.t

which describes a linear decay of angular velocity from Wo=131 to 0 radians/second.


Solving for w(t)=0 at 140 sec yields rotational acceleration of Ao = -0.935 rad/sec^2, so
the drag torque and initial power dissipation at t=0 is

To = I.Ao = -0.643 millinewton meters.


Po = To.Wo = -84.2 milliwatts.

The negative sign on power indicates a loss. To maintain 1250 rpm requires a steady 84.2
milliwatt input.

The more complicated case assumes that bearing friction is 20% of the value estimated
above, and that air friction (a velocity-dependent viscous drag) of the spinning rotor
accounts for the remaining loss. The resulting motion equation is provided analytically by
solving a linear non-homogenous differential equation with initial condition w(0) = Wo:

A(t) = dw(t)/dt = Ao - b.w(t)/I = C.w(t) + Ao


w(t) = (Wo + Ao/C).exp(C.t) - Ao/C

where b is the coefficient of rotational viscous drag and C = -b/I. This describes a quasi-
exponential decay of angular velocity from Wo=131 to 0 radians/second. Solving for
w(t)=0 at 140 sec with Ao = -0.187 rad/sec^2 (20% of the earlier figure) yields C =
-0.190 rad/sec, b = 0.0131 millinewton meter seconds. So the drag torque and initial
power dissipation are

To = I.A(0) = I.[Ao - b.Wo/I] = -1.84 millinewton meters.


Po = To.Wo = -241 milliwatts.

To maintain 1250 rpm requires a steady 241 milliwatt input. When power is removed,
50% of the kinetic energy is lost within about 20 seconds.

Compare these results to those which assumed bearing friction alone. This shows that it is
essential to measure velocities at several points as the system winds down, not just the
total run time, so that the losses can be accurately partitioned between viscous and
velocity-independent components.

Numerical solving (4th order Runge-Kutta) provided the same results and decay curves
for w(t) to better than 1ppm of the analytical method.

Now I’ll comment on the other experimental conditions, relative to the condition
analysed above.

“With rotor magnets installed, stator magnets installed, stator magnets free to move but
not spinning, rotor initial rpm 1250, rundown time ~120-125 seconds.”

Run time is about 10 seconds less than when stators are omitted. The rocking of the
stators causes slightly increased friction loss. Perhaps some eddy current is induced in the
stators, another loss.

“Same as above but with stator magnets (3) spinning gearwise, initial rpm 1250, rundown
time ~ 90-100 seconds.”

Increased bearing friction and air drag on the spinning stators, and eddy currents induced
in the Judson Dampers by the stators rotating nearby cause losses that decrease run time
to its shortest value.

“Same as above but with stator magnets screwed down hard to baseplate, not moveable,
initial rotor RPM 1250, rundown time ~110-120 seconds, with a distinctly ‘coggy’ feel
and a reversal of direction at the very last revolution.”

Same basic result as when stators were ‘free to move but not spinning’, but couldn’t get
over the last ‘cog’ to run unhampered for another 10 seconds in the same direction.
Over to you for questions and/or comments. (Criticism is discouraged from those who
don't know the difference between a newton meter of energy and a newton meter of
torque )

To assess the drag forces in this system, I have analysed the experiment: “stator magnets
removed and far away, initial rotor RPM 1250, rundown time ~ 130-140 seconds”.

Other assumptions gleaned from other posts: rotor mass = 258 grams, diameter = 5 ¾”,
radius = 2.875” = 73.0mm. The moment of inertia I = 6.88 kg.cm^2. At 1250 rpm, the
rotational kinetic energy of the rotor disc is 5.89 joules.

The simplest case is that all drag force is bearing friction. The resulting motion equation
is

w(t) = Wo + Ao.t

which describes a linear decay of angular velocity from Wo=131 to 0 radians/second.


Solving for w(t)=0 at 140 sec yields rotational acceleration of Ao = -0.935 rad/sec^2, so
the drag torque and initial power dissipation at t=0 is

To = I.Ao = -0.643 millinewton meters.


Po = To.Wo = -84.2 milliwatts.

The negative sign on power indicates a loss. To maintain 1250 rpm requires a steady 84.2
milliwatt input.

The more complicated case assumes that bearing friction is 20% of the value estimated
above, and that air friction (a velocity-dependent viscous drag) of the spinning rotor
accounts for the remaining loss. The resulting motion equation is provided analytically by
solving a linear non-homogenous differential equation with initial condition w(0) = Wo:

A(t) = dw(t)/dt = Ao - b.w(t)/I = C.w(t) + Ao

w(t) = (Wo + Ao/C).exp(C.t) - Ao/C

where b is the coefficient of rotational viscous drag and C = -b/I. This describes a quasi-
exponential decay of angular velocity from Wo=131 to 0 radians/second. Solving for
w(t)=0 at 140 sec with Ao = -0.187 rad/sec^2 (20% of the earlier figure) yields C =
-0.190 rad/sec, b = 0.0131 millinewton meter seconds. So the drag torque and initial
power dissipation are

To = I.A(0) = I.[Ao - b.Wo/I] = -1.84 millinewton meters.


Po = To.Wo = -241 milliwatts.

To maintain 1250 rpm requires a steady 241 milliwatt input. When power is removed,
50% of the kinetic energy is lost within about 20 seconds.
Compare these results to those which assumed bearing friction alone. This shows that it is
essential to measure velocities at several points as the system winds down, not just the
total run time, so that the losses can be accurately partitioned between viscous and
velocity-independent components.

Numerical solving (4th order Runge-Kutta) provided the same results and decay curves
for w(t) to better than 1ppm of the analytical method.

Now I’ll comment on the other experimental conditions, relative to the condition
analysed above.

“With rotor magnets installed, stator magnets installed, stator magnets free to move but
not spinning, rotor initial rpm 1250, rundown time ~120-125 seconds.”

Run time is about 10 seconds less than when stators are omitted. The rocking of the
stators causes slightly increased friction loss. Perhaps some eddy current is induced in the
stators, another loss.

“Same as above but with stator magnets (3) spinning gearwise, initial rpm 1250, rundown
time ~ 90-100 seconds.”

Increased bearing friction and air drag on the spinning stators, and eddy currents induced
in the Judson Dampers by the stators rotating nearby cause losses that decrease run time
to its shortest value.

“Same as above but with stator magnets screwed down hard to baseplate, not moveable,
initial rotor RPM 1250, rundown time ~110-120 seconds, with a distinctly ‘coggy’ feel
and a reversal of direction at the very last revolution.”

Same basic result as when stators were ‘free to move but not spinning’, but couldn’t get
over the last ‘cog’ to run unhampered for another 10 seconds in the same direction.

Over to you for questions and/or comments. (Criticism is discouraged from those who
don't know the difference between a newton meter of energy and a newton meter of
torque )

Al- Jan 15th 2007


What I've seen look really good, I think people are doing some impressive work, and I
hope having some fun too.

I don't have much time


But here are a few tips

First, safety. Safety First!


--do not leave magnets lying around loose, even for a few moments. They are easy to lose
control of, and hot flying pieces and pinch hazards are to be taken seriously. I use metal
jar lids as magnet trays.
--do NOT grind, burn, etc. these NdBFe magnets! The dust is toxic and otherwise
obnoxious, the fumes are toxic and smelly, etc.
--do not rely on adhesives! use structure to retain magnets against fields and forces.

Next, machining acrylic plastic and polycarbonate:


--Use WD-40 or clean kerosene (NOT GASOLINE OR PETROL!) as a cutting lubricant
to keep drills, taps and other tooling cool and cutting well. Beware of alcohols--acrylic
crazes and loses strength on exposure to alcohol--even the alcohol in a Sharpie marker
will craze the acrylic over time, and cracks will occur. Wash the WD-40 off with
dishsoap and warm water.
--plastic supply companies sell special drill bits with a sharper tip angle that won't crack
or pull though the plastic. They are well worth the price.
--I usually use bottoming taps even when starting, in plastic, I rarely need the plug tap to
start a tapped hole. The bottoming tap makes a better thread I think.

On perseverance:
--It took me literally days of spinning and experimenting before I had a hint that
something interesting might be happening--the decreased rotor drag that I first noted
before the holidays. After that, more days of trial and error and serendipitous mistakes
before I even found the AGW rotation modes. By Accident! So don't get discouraged if
nothing seems to work--you may not have the right components or something, or it's not
put together right. Recall also that about half my magnet-bearing pairs are ineffective.
--I would guess that most people will get decreased drag and increased rundown times if
they get the magnets right. But extended run times like I saw are probably a result of
some EM interaction that's happening in my apartment and at work. For sure it isn't
anything like OU.

blue_energy- Jan 15th 2007


I've been trying to educate myself on magnets so that I can understand some of the things
being said and put them into perspective. Here's something to add to the discussion of
how powerful the magnets in Al's rig are. I read on overunity.com that someone
remembers Al stating somewhere that he used N35 neodymium magnets that were one
quarter inch in diameter and a little over a half an inch long on his rotor. I get what a
quater inch by a little over half an inch cylinder looks like. But - what about "N35"? How
powerful is that? This site:

(http://www.monstermagnete.de/catalog/advanced_search_result.php?keywords=Z06)

...sells an N35 cylinder magnet that is 6mm (.24 in) in diameter and 15 mm (.59 in) long.
Here's what it has to say about it:

<<<
Diameter(d) = 6 mm
Height(h) = 15 mm

Material/grade: N35
Residual induction KG(mT) = 11.7-12.2 (1170-1220) = 1.1 Tesla
max.operation temperature = 80 °C

Dead weight: 3.16 g


Holding force on a steel plate: 15.43 Newton
Weight, which the magnet can lift: 1.57 kg
>>>

So, each of these eight little magnets can lift over 3 lbs. That gives me an idea. So, what
does "N35" mean anyway? This page:

http://www.stanfordmagnets.com/magnet.html#grac

...discusses it here:

<<<
9. How to Choose the Correct Grade of Permanent Magnet Materials
Selecting a grade is the next step, once you have decided which permanent magnet
material is best for your application. Generally, a grade indicates the Maximum Energy
Product of a magnet. For instance, Grade 32 implies the (BH)max is about 32 MGOe. A
higher grade of permanent magnet has a better performance. However, higher grade is
usually associated with a higher cost. Taking sintered Nd-Fe-B magnets as an example,
the price of Grade 45 is twice and even more of that of Grade 33. Other property
parameters, such as Br and Hci, also need to be considered in selecting a grade. One way
to select the suitable grade for your application is "trial and error". You can purchase the
several magnets with different grades (some suppliers have these magnets available on
their shelves) and try each grade until you find one right for your application.
>>>

...and here:

<<<
Maximum Energy Product, (BH)max.: There is a point at the Hysteresis Loop at which
the product of magnetizing force H and induction B reaches a maximum. The maximum
value is called the Maximum Energy Product. At this point, the volume of magnet
material required to project a given energy into its surrounding is a minimum. This
parameter is generally used to describe how "strong" this permanent magnet material is.
Its unit is Gauss Oersted. One MGOe means 1,000,000 Gauss Oersted.
>>>

So - N35 is a grade of magnet that is about 35 million Gauss Oersted.

Most of the overunity.com replicator crowd seems to be using N42 magnets from
http://www.kjmagnetics.com. I wonder whether that is throwing off the effect
considering that some effort has been made to, otherwise, keep the same dimensions as
Al's rig?

Whoops - now I'm seeing that gauss and gauss Oersted are different animals altogether
and likely can't be contrasted.

WhiteLite- Jan 15th 2007


Blue_energy. I'm usually a bit flakey when it comes to detail but here's my simplification
of what a magnets grade is all about:

The N in the N35 just means it's a Neodymium magnet. I think Samarium Cobalt
magnets start with S, Alnico's start with A etc. There are 2 components needed to work
out the rating, (which is also known as BHmax). One is B(r) which is the strength the
magnet holds when not in the presence of another field and I think H(c) which is the
strength of an opposing, (repelling), field needed to demagnetise the magnet. The 2 kind
of plotted together on a graph form a square and the greater the area the greater the grade
of the magnet. The grade can also be measured as KJ/cubic metre as well as MGOe.

MGOe = megagauss-oersteds. I don't know how this relates to Gauss values to be honest.

I think someone recently stated that a magnets field strength decays by an inverse cube
amount but I only know the basics. No doubt some one will educate you, (and me),
further.

Dirtfarmer- Jan 15th 2007


Has anyone noticed that in relative motion, there is a display of spin, , as well as orbital
angular momentum......
Strong analogy to the earth, moon, solar system, and in the other direction, (and probably
more meaningfully) atoms?(electrons)...
Orbo might not be such a bad name after all...
Expect some thoughts on this...
Bohr, Fermi energy...
We may actually be seeing quantum effects being brought to the very "macro" scale...
Differing interactive trajectories would tend to display angular magnetic moment...
I think I am scaring myself...

Harvey- Jan 17th 2007


Al has been absolutely swamped trying to stay on top of things and help the replicators as
much as possible. I did see a post regarding a dual trace scope but didn't follow up on it.
I'm sure he's spending his time trying to pin down the cause of this 'artifact'. Theories,
Hypotheses and Ideas abound but ultimately they all boil down to two simple necessities:
1. Reproduction 2.Geographic independance. If you want to see how the unit is made
check here:
http://www.ospmm.com/whipmag/
I've reproduced the waveforms Al posted using my Mock Up, and many have achieved
AGW sync. Some are attempting exact replicas (something I recommend both AL and
OC should do to pin down tolerance anomolies). Others have created wide variants based
on principles alone. As of yet none have acheived sustained operation or acceleration as
AL has and everyone involved is anxious to identify the cause.

kdelaere- Jan 17th 2007


From http://www.mceproducts.com/knowledge-base/article/article-dtl.asp?id=90

Magnetic field effects

A magnet can be partially demagnetized based on the magnetic load placed upon it. This
effect is normally studied by looking at the second quadrant of the hysteresis curve of the
magnetic material, also referred to as the demagnetization curve. Typically, this is the
only section of the hysteresis curve that is reported for a hard magnetic material. This
curve shows the response of the magnets delivered flux into the space around it (B) to the
demagnetizing force (H) imposed on the magnet. To assess the performance of the
magnet material in a given situation, the user should calculate the ratio of B/H (taking
care to ensure the conformity of units). Once this ratio is calculated a line is
superimposed on the demagnetization curve, as below for a B/H ratio of 0.8 (note that the
illustration shows four demagnetization curves to reflect different operating
temperatures):

http://www.mceproducts.com/_img/intranet/DemagGraph.gif

The curve of interest that we are comparing our blue line to is the diagonal red line, the
demagnetization curve. Notice that each of the diagonal lines has a bend in it,
colloquially referred to as the “knee” in the curve. If the blue line, which shows the
working condition of the magnet crosses above the knee in the curve, then the magnet is
operating in its safe linear region and should perform as expected. If the magnet’s load
line is below the knee on the curve, the magnet will become demagnetized and damaged.
It is also important to notice that the knee gets higher with increasing temperatures,
which reflects the material’s increasing vulnerability to demagnetization at higher T.

Harvey- Jan 16th 2007


Well, I took some stroboscopic video of GW and AGW sync today. They are on VHS-C
and now I have to get them on a computer some way.

Here are the results:

The poles on the stator seek the opposite pole on the rotor. The magnetic equators align
with very little shift.

In GW sync there is no shear, the fields mesh clean.


In AGW sync there is a definite shear as the two equators pass each other. This is a very
important point: The inertia of the stator must allow for sync'd traversal through this
repulsive zone. Too much mass produces strobes like these:
http://www.youtube.com/watch?v=Q-McMAqt04Q

Too little mass an you won't be able to sync as the repulsion overcomes the inertia.

From these we also realise there are limits to the speed at which the device will operate.
When and if this ever becomes a production device special stators would need to be
constructed to adjust the mass as needed to allow for various ranges.

As more data is gathered I believe we will find this to be a CoM machine that shapes
greater field density into the leading zone and lower field density in the trailing zone.

OC will be happy to see that the stator does 'flip' slightly as it pushes through the narrow
repulsive zone into the wider attractive zone if the stator is imbedded into the field.

I found that things smooth out when a true 4:1 magnetic circumference is obtained, and in
this position the stator traverses with a relative smoothness through the repulsive zone.

Now, what to do to get a 'self-runner'. Is the key in this traversal zone? Is the shape of the
fields in this area create a differential in Mass-H relationship that has the effect of adding
KE in a resonant manner?

I believe it is possible to establish a magnetic equivalent of a satellite orbit and that this is
what we are witnessing. A balance between interia and an attractive force.

Cheers

Harvey

Harvey- Jan 17th 2007


Al has been absolutely swamped trying to stay on top of things and help the replicators as
much as possible. I did see a post regarding a dual trace scope but didn't follow up on it.
I'm sure he's spending his time trying to pin down the cause of this 'artifact'. Theories,
Hypotheses and Ideas abound but ultimately they all boil down to two simple necessities:
1. Reproduction 2.Geographic independance. If you want to see how the unit is made
check here:
http://www.ospmm.com/whipmag/

I've reproduced the waveforms Al posted using my Mock Up, and many have achieved
AGW sync. Some are attempting exact replicas (something I recommend both AL and
OC should do to pin down tolerance anomolies). Others have created wide variants based
on principles alone. As of yet none have acheived sustained operation or acceleration as
AL has and everyone involved is anxious to identify the cause.

kdelaere- Jan 17th 2007


From http://www.mceproducts.com/knowledge-base/article/article-dtl.asp?id=90

Magnetic field effects

A magnet can be partially demagnetized based on the magnetic load placed upon it. This
effect is normally studied by looking at the second quadrant of the hysteresis curve of the
magnetic material, also referred to as the demagnetization curve. Typically, this is the
only section of the hysteresis curve that is reported for a hard magnetic material. This
curve shows the response of the magnets delivered flux into the space around it (B) to the
demagnetizing force (H) imposed on the magnet. To assess the performance of the
magnet material in a given situation, the user should calculate the ratio of B/H (taking
care to ensure the conformity of units). Once this ratio is calculated a line is
superimposed on the demagnetization curve, as below for a B/H ratio of 0.8 (note that the
illustration shows four demagnetization curves to reflect different operating
temperatures):

http://www.mceproducts.com/_img/intranet/DemagGraph.gif

The curve of interest that we are comparing our blue line to is the diagonal red line, the
demagnetization curve. Notice that each of the diagonal lines has a bend in it,
colloquially referred to as the “knee” in the curve. If the blue line, which shows the
working condition of the magnet crosses above the knee in the curve, then the magnet is
operating in its safe linear region and should perform as expected. If the magnet’s load
line is below the knee on the curve, the magnet will become demagnetized and damaged.
It is also important to notice that the knee gets higher with increasing temperatures,
which reflects the material’s increasing vulnerability to demagnetization at higher T.

Craigy- Jan 17th 2007


The operation of this motor is reliant on the the stator being able to vary its speed. The
acceleration, and deceleration must be of a controlled nature so that the overall stator
interaction is asymmetrical and we get a gain. If the stator and rotor were locked or
geared together you would not get a gain. If you could gear it together but with a 30
degree null zone on the gearing this would allow the decelerations and accelerations so
vital to get through the rotor field with a gain. In Al´s motor it appears he had the balance
just right, and the inertia was in just the right amounts.

Therefore if the stator is too light, the stator will go too fast in the accelerations and
slam into the wall of repulsion, if the stator is too heavy it won´t react fast enough and
it will slam into the wall of attraction. The trick is getting a dribble of one or the other.

In this femm graphical representation of the interacting fields we need to stay in the blue
zones to escape with energy
this was done

rX- Jan 17th 2007


This morning I noticed that some of Al's rotor magnets did not seem to be centered in
their slots when looking at 7Ax8_1.JPG.

Then I noticed at overunity.com that starcruiser said this.

"guys,

I was looking at the stills and videos of Al's unit again and it seems to me that some of
the rotor magnets are not centered, or so it seems. can some one take a look and confirm
this? I was wondering if this "imbalance" of the magnets may be a key."

Can someone take a look at this? It sure seems to be off to me.


But maybe it is an optical delusion.
Harvey- Jan 17th 2007

vibrator:Yes, well spotted. It could just be the camera angle, but i


loaded that image into mspaint, turned it into a monochrome bitmap,
zoomed in and drew white lines parallel to the ends of each magnet.
When the image is superimposed over itself and rotated 180deg, the
magnets on two, opposite sides seem out of kilter, maybe 10-20% of
their length...?

ETA: i tried it on my rig, suggest others do likewise as i ain't taking


data, but wind down seems longer.... may be imagining it tho...

I would take ratios for this analysis: Octogon Corners:Magnet spacing. Then
compare the ratios. That way you account for the parallax.

If we have a differential, then you are changing the field density and there must be
an exchange of inertia to compensate. If there is resonance such that drag is
reduced or cancelled by the exchange (Magnetic Reactance = 0) then this may be
a key.

The natural tendency of the device is to conserve momentum. Altering the 'length'
of the field changes its density. When the rotor encounters a density out of sync
with its inertia it will go through an adjustment speeding up or slowing down as
necessary. When these adjustments become resonant its possible to reduce the
magnetic impedance.

You may be on to one of the clues there.

Harvey- Jan 18th 2007


http://www.youtube.com/watch?v=aJBRmWTcdEc

This is the continuation of Vid 1 and shows the manner in which I identified the pole
faces of the stator. I rewound the tape; so the beginning of this one overlaps the end of
Vid 1. It's a bit boring, but it does provide the necessary labels to reduce the associated
confusion.

In addtion, I watched the entire sequence again and using pause at the critical points was
able to see clearly each pole's association. There are other tell tail markings, a glue blob
and a missing fan blade that help ID the orientation as well.

So, from the two AGW strobe recordings I have made thus far I have learned that
excessive mass (or magnetism) seems to create erratic shears. I will be testing this
further by increasing the mass but keeping the lower magnet strength.
p.s. I have found my capture card but I'm not sure the drivers are still available. If I can
get that working the quality of the videos should improve. At that point I will take an
overhead shot and set things up a little more professionally.

Until then, if we can get other testers and replicators to post their AGW stroboscopic
shots it would be helpful.

Al- Jan 19th 2007

riterX:

MeggerMan:
@Al, one question/test for you, if you stack two of the
rotor magnets in a plastic tube in repulsion, what is the
gap that seperates them?
I know the gap from a N42 1/4"x1/2" rod.

Regards
Rob

I would really like to know the answer to that one also.

Good test.
Almost exactly 3 cm, maybe a rch more, like 3.05 or so.
Of course, density is unknown.

MeggerMan- Jan 19th 2007


Hi Al,
Thanks ... 30.5mm sounds very close to the results for a N42:
http://i100.photobucket.com/albums/m25/kingrs/DSCN5262.jpg

31.5mm by bouncing the magnets and looking at the average.


The tube is Kapton, so its difficult to see through.

Perhaps its N40?


[edit] by adding another 1/4" x 1/4" rod (total = 1/4" dia x 3/4") to the bottom magnet
increases this to about 33mm [/edit]

Regards
Rob
Al- Jan 19th 2007
@Rob--maybe, I just don't know. I don't think so though, they certainly don't seem as
strong as the 834DIA stators.
@CLaNZeR--I will email you and you can send me the login. Thanks--it's just the dual-
trace scope readings that Harvey wanted to see.

vibrator- Jan 19th 2007


i'll correct you there again - it was OC's reconnection theory i was looking at. i suggested
magnetic islands might shear off two reconnecting fields creating an asymmetry. OC then
found a paper suggesting the same thing,
eta: incidently i stated if anything i thought Sv was the culprit - you and OC instead
claimed you'd ruled it out. I knew however that you couldn't have, because i know
something you don't know which makes that claim impossible.

CLaNZeR- Jan 19th 2007 from OU


The main problem has been that it is very hit and miss to get the latching as others
replicating will tell you. One minute it will work like a dream and the next it will be a
bitch to get going for a long time. When I get to that stage I remove my stator magnets
for 30 mins away from the rotor and pop them back in for testing. Somehow it seems the
stators are getting degraded over time or something weird going on.

vipond50- Jan 19th 2007 from OU


To All Replicators and Group
Today after reconfiguration of the Stator magnets to a dual bearings I had a
approximately. "one minute" self operation of the device and maintained a constant RPM
once Stator Sync occurred. Note: What I mean by " Self Operation' is NO outside added
input to the device. I added the second bearing to the stator due to instability to each
arrangement. The instability induced friction and consequently drug down the rotor
speed. Also with the device i am able at any time during spin up to latch any of the three
Stators at any given time. In addition this allows me to have any combination of the
stators running in synchronization.
My next step is to spend the necessary funds to Source totally all Non magnetic parts i.e.
Bearings, Fasteners, Rotor center support shaft, etc. to construct a higher quality device.
I also understand that this is a variant of the original concept, but basic Engineering
dictates that this is the next step.

Happy testing and may Your Rotors Rev up

fritz- Jan 19th 2007 from OU


I think there are 3 clear optimization steps:

1) Finetune the position of the rotor magnets


(I don´t think that all this magnets have exactly the same strength)
or match the magnets out of a bigger batch with a
magnetometer - or use a reference stator with a newtonmeter
attached.
This first goal is to get a symmetrical, smooth rotorfield.

2) Finetune the position of the stator magnets


I think that the stator magnets have slightly different strength.
If you finetune the distance of each statormagnet to
the already adjusted rotor using newtonmeter or measure the
resulting field - you can compensate differents in stator magnets
and achieve symmetrical statorfields.

3) would be to balance rotor and stators in the conventional way.

Even slightly unsymmetry and slightly different strength in the used


material can have a huge difference in the resulting force.
Measuring the force and the field is the key to operation -
then it would make sense to alter the tilt/distance/force of the stator
magnets.

modervador- Jan 19th 2007


Dirtfarmer:Who can help interpret this?
http://yirkha.fud.cz/steorn/alsetalokin/Dual%20Trace/

Just remember that each voltage peak represents the highest rate of change of
magnetic flux through the coil, not the highest flux. So for example a positive
peak might represent a rapid switch from N to S, and a negative peak is a rapid
switch from S to N.

That said, the symmetry_t#.jpg scope traces are not surprising. The upper and
lower traces show basically the same thing, displaced in time in time because of
the angular difference between the coils. The coils are apparently hooked up in
opposite polarities, consistent with being hooked in series and grounded at the
center tap.

The readings from the two coils in dual_1.JPG are likely different because the null
of the stator field lags a few degrees behind the null of the rotor field, so the coils
do not see the same flux changes vs rotor angle. It would be interesting to see if
the upper and lower traces reverse if the rotation is reversed, or how they change
according to RPM.
Dirtfarmer- Jan 19th 2007
modervador,
A few questions,

1. Is it likely that the difference in magnitude, (presuming 20mv/div) between A and B is


purely related to coil air gap? (Not precise placement)
2. We had seen a 6 peaked form( +&- )in the symmetry 2 pic.. Why is it that we still see a
fundamentally 6 peaked saw in "dual"?
Is this unexpected?
3. Can we say that Trace B (top trace) is more energetic than Trace A in "dual"?, Or is
there not enought information...
If so, would this not be unusual?
It would seem to imply an imbalance between the two points... (and hence a potential... )
(I understand that potentials as that always exist in a waveform, but here it "seems"
constant... and only divided by equi-distance(big presumption) on an arc..

modervador- Jan 19th 2007


1. in the case of the symmetry traces, yes. In the dual_1.jpg, the different peak height is
dominated by the different rate of flux change.

2. The 6 peaks per period are probably because of the same changes in rotor flux in
symmetry_t2 and dual_1, however in dual_1 the stator flux superimposes on the rotor
flux to alter the shape.

3. I'm unsure of the energetics of this. The flux through the coil is proportional to the
integral of the voltage, but the integrated force of attraction between rotor and stator on
the "upspin" and "downspin" regions is not necessarily the same as what could be
discerned from what the coil measures at this position.

sveinutne- Jan 19th 2007


This might be very good news.
If my latest FEMM simulation is for real, it might give us the information we need.

It shows how we can get a positive torque for both the rotor and the stator.
Before I get too carried away here, we need to wait with the champagne till this has been
triple checked.

It shows that the stator needs to deviate from a constant speed to get a gain. In this
simulation this is taken to the extreme where the stator almost stops at the close passing
of the rotor magnet, and gets maximum acceleration. As speed picks up the momentum of
inertia increase and the torque from the magnets will give less change in speed. This will
give less extra torque, and the power gain will be less.

So with a stator with high mass or momentum, it will be more easy to get into AGW
locking, but the inertia will soon be too high for further increase in power. With a lighter
stator it might be difficult to start, but at higher speed it will still fluctuate and give better
torque.

Personally I thought my stator was too light, so I put lead inside it. It helped in getting a
more easy AGW lock, but I do not see the OU any more. The first day before I added the
lead, I had a gleam of what Al might have felt when his rotor started to spin by itself.
So maybe I will remove the lead and see if I still can get an AGW lock.

omnibus- Jan 19th 2007 from OU


I'd like to reemphasize the points @fritz made and my reply
there:http://www.overunity.com/index.php/topic,3871.msg71709.html#msg71709 Proper
form of the magnetic fields is of crucial importance in all these motors. Very hard to
achieve. I'm looking now at one of my ostensibly (on the face of it) exact replicas of
@alsetalokin's motor. Supposedly, when slowly turning the rotor the pattern one observes
when the rotor is let go form a maximum to maximum (one feels these maximums when
turning the rotor slowly by hand) should be of the same form repeated four times. Not so
in my case. I'm observing the stator equally outstanding from the other two. If I start with
a rotor magnet facing that stator the rotor bypasses the next magnet and bumps into the
barrier formed by the third one. If I, however, continue and let go the rotor, further
clockwise, from that barrier on the rotor stops at the fourth while it should bypass it if the
pattern is to be preserved and so on. There's only one other place where the initially
described pattern is observed, although slightly different (intermittent accelerations
appear different). Now, that may be a significant problem. I've ordered more of these
cylindrical magnets and will try to find matching ones to see whether this picture will
improve. Of course, the stator magnets should be matching too. Wonder if someone else
would be willing to check his rotor in this fashion.

Al- Jan 19th 2007


Oh, and I believe that a couple of folks have shown that the stator does accelerate and
decelerate on each cycle, like RB predicted and as you have also postulated.

Meggerman- Jan 19th 2007


Hi alsetalokin ,
Is it possible that your stator is at a slight angle to the rotor in one of the two planes?

When I was testing last night with a motor driven stator magnet in AGW rotation the
stator magnet wobbled a lot.

It could be that drill holes for the stator/holder (bearing side or magnet side) were off
angle which may explain why some stators work and others don't even though the
magnets in them are identical.

Could you check this for me as I have an idea this could be the vital missing link.

Another test you could try is using a couple of fibre washers cut in half to mount the
stator diliberately at an angle to the rotor.
Maybe this will enable the sync to work better, or to allow stators that did not work
before to work now.

Regards
Rob

Harvey- Jan 19th 2007


I don't know Thomas or where he gets his information from but someone has given him
faulty data. Check the Discussion "Whipmag for Dummies" and you will find the
timeline of this opensource project and how it has been exaggerated beyond the simple
statement that it spins and we don't know why. The only way to solve the puzzle is by
replication and testing. That is currently being done and new data is added daily.
Mathematical models and simulations are proving helpful as well. You can search for
posts by the participants and gather more info. If there is anything negative and you can
tell it was not meant as a joke, just discard it as the worthless dribble it is. The internet is
full of junk posts and I trust your smart enough to tell the difference If you find any posts
with compelling evidence that proves otherwise, please share it with us. All of us are
imperfect and we make mistakes and most in here are willing to have them corrected if
brought to our attention.

modervador- Jan 19th 2007


The "noise" on these traces is about 80 Hz.

Dirtfarmer- Jan 20th 2007


We have this information:
Dual trace readings with new sensor coil
Unit rotation rate ~1250 rotor, 5000 stator

5 ms/div horiz, 20 mV/div vert

A on bottom, B on top, coil A upspin, B downspin.

Dual_1:is the trace from active zone


postion_1: coil position
pos1_flash: same as above but flash frozen
sensor_pair: the coils
symmetry (various): showing coil symmetry in various positions

Harvey- Jan 20th 2007


Stator Mass:

I broke some fins on my mockup stator so I dremeled them all off and put a washer
down. The extra mass of the washer was very counter productive. Not only does it
not sync AGW and barely syncs GW it puts a serious load on the rotor sync'd or not.
.
Hence my post elsewhere regarding the critical nature of both the mass and M value
We need the Math!

hdeasy- Jan 20th 2007


Hi Harvey,
Sorry for the misunderstanding - my plots only cover about 60 deg in Rotor rotation
angle. Normally I plot against time, a range of about 0.005 seconds, but as the rotor is at
about 200 RPM, that means about 0.3 seconds per rev, and so 0.005 sec is about 1/6 rev -
OK?. Just to spell it out ever plainer, here is the torque again but plotted against stator
angle:
http://www.geocities.com/deasyart/steorn/torque.GIF

@MrEntropy

Yes, at BR's suggestion I first commented out all lag /Sv / viscosity and then even BH
movement (magnetisation) as BR thought my gain was due to that. To my surprise even
after switching all that off I still got the same gain. As you can see by my posted graphs,
the gain now seems to be due to the stator N being nearer to the rotor N at the positive
torque phase than at the negative torque phase. THis makes the push stronger than the
pull and is an effective lag due to motion.
Tomorrow I will implement the mod I worked out after BR queried my 'fudge' factor on
the stator torque. I don't think it will make a difference, but watch this space!

As for that Nov 6th comment - that was back when BR queried my results for the static
stator case - he had sugggested putting in 1/R and not 1/R**2 for the force to see if gain
went away - it didn't, so I re-instated the physically more realistic inverse square law.

Hopefully good news tomorrow on that stator torque...


Oh and MrEntropy, the corrected stator torque is:

Ttops = 0.1*Qbig*Qsmall*( (1.0/(x32+y32)) * (x3*(YS1-YC) - y3*XS1)/Dout1 - (1.0/


(x12+y12)) * (x1*(YS1-YC) - y1*XS1)/Din1 )
Tbots = 0.1*Qbig*Qsmall*( (1.0/(x22+y22)) * (x2*(YS1-YC) - y2*XS1)/Din2 - (1.0/
(x42+y42)) * (x4*(YS1-YC) - y4*XS1)/Dout2 )

Then replace

Ttots = Ttops - Tbots

I'm going to implement that tomorrow, along with the 5g. Maybe you can lug into your C
++ and beat me to it?

CLaNZeR- Jan 20th 2007


Now it is making sense why the weaker stator magnets I have been playing with latch
better than the stronger, correct ones.
Weaker ones were 10g and correct ones were 5g as this picture shows:

< http://www.overunity.org.uk/ocpm/weight.jpg >

I added some blue tack and took the Correct Stators to 10g and latches very easly now as
shown in this video.

< http://www.overunity.org.uk/ocpm/CLaNZeRSaddingweight.wmv >

On this chart I moved the stator magnet closer as you can see by the sharper decline in
accelaration.
This one latched at around 350rpm .

I need to video the wheel and Tacho at the same time as logging so we can tie this all in
to when it latched etc etc. Will see if I can get one of the kidz to assist LOL

< http://www.overunity.org.uk/ocpm/OCMPMMtest3.xls >


hdeasy- Jan 20th 2007
Ah yes, I see it better here. Round 350, the decleration is held up for a while and then the
spin-down takes over again. So - maybe by slogging through the gap distance and blue-
tack amount you can find the optimal combination for which the bump explodes upwards.

Or is it the straight section between 8 and 13 time units which is the synch period?

By the way, Svein said this a few hours ago in the Petition thread:

"So with a stator with high mass or momentum, it will be more easy to get into AGW
locking, but the inertia will soon be too high for further increase in power. With a lighter
stator it might be difficult to start, but at higher speed it will still fluctuate and give better
torque.

Personally I thought my stator was too light, so I put lead inside it. It helped in getting a
more easy AGW lock, but I do not see the OU any more. The first day before I added the
lead, I had a gleam of what Al might have felt when his rotor started to spin by itself.

So maybe I will remove the lead and see if I still can get an AGW lock."

CLaNZeR- Jan 20th 2007


One more as grub not on table yet LOL

This time 0.05 sampling rate and also Dummy Run with no Stator to compare.

< http://www.overunity.org.uk/ocpm/OCMPMMtest4.xls >

Turbinator- Jan 20th 2007


I've a couple of ideas that will be implemented seperately; one is to mount the stator
magnet on a spring loaded pivot arm such that the spring tension holds it at a mid point
and allows movement in both directions; toward and away from the rotor. Secondly we
discussed a pulley system, since the stator has a fast and slow rotation through 360°
machine an eliptical pulley in order to copy the uncoupled motion.... this could be worked
in with the pivoting arm simply using a tension wheel, but one thing at a time.

fritz- Jan 20th 2007 from OU


I have some experience building magnet sensors using Philips/NXP:
KMZ10A +/- 0.5kA/m; = +/- 625uT
KMZ10B +/- 2kA/m; = +/- 2.5mT
This stuff is probably way to sensitive - but using it in a metal core as shunt
can give nice sensors.
If somebody is interested, pls. contact me.

Anyway, using a spring gauge to check attract/repell forces is a very easy


way to match the magnets.
Another idea would be a setup to degrade magnets with a strong m field,
checking the field after every degrade cycle to achieve exactly matched magnets.

blueletter- Jan 20th 2007


Korkskrew,

If it makes any difference, I believe the photo you referenced was a flash photo taken
while the unit was in motion.

korkskrew- Jan 20th 2007


Wow! If true, that would be really interesting. It would be completely at odds with my
theory. More than one person has suggested that my prediction is about 180 degrees from
observations.

/- Jan 20th 2007


Yes, that photo was taken with flash. Here are all three flash-frozen positions we know so
far (or rather knew few days ago, there was one more in the most recent photos from Al).
As a matter of fact, I made my early simulations 180° off too, I hadn't predicted it would
be this way...

Al- Jan 20th 2007


"Test1:
With the rotor removed from the baseplate, using a single rotor magnet, measure how far
away from stator you need to be to be able the move the stator magnet using repulsion
force. With a heavily magnetic bearing I expect this distance to be very small because of
the large amount of friction.
From the ease with which you seem to be able to spin the stators I suspect you have some
very non-magnetic bearings."

The bearings are very ferromagnetic, but they seem not to lock up as I would have
expected, maybe due to the diametric magnetization of the outer race and the free-
spinning of the balls and the inner race.
The answer to the test question is about 10.5 cm.

"Test2:
Using a piece of plastic tube slide the magnet and bearing onto it and measure the
maximum gap before the bearing falls off from its own weight overcoming the attraction
force."

I don't want to do this test right now, as the only bearings I have of this type are mounted
with magnets, and I don't want to disassemble any of them, for what should be obvious
reasons.

"Test3:
Using a tube that the rotor magnet can fit in, measure the repulsion distance between like
pole of the stator ring magnet. Setup should look like and exclamation mark!"

The answer here seems to be, magnet in holder with bearing, about 3.5 or 3.7 cm. A bare
magnet gives roughly the same, maybe, strangely, a few millimeters less. I don't really
have a good tube here, the tube I'm using might have some drag.

Harvey- Jan 20th 2007


These pictures are quite consistent with my strobes. The difference is my device runs
(powered) CCW and AL's operates CW. The magnets are in repulsion as they pass
equatorially.

@ Clanzer

It is really good to view the real life dynamic data you have provided.

The data clearly shows rotor loading esp. after sync as the KE of the rotor is transferred
to the stator.

Now for the hard part. We need you to plot data for different speeds without changing
any other parameters. I am hoping to see some flat lining with the rotor at speeds between
1000 and 2200
-----
Magnets themselves can have different domain orientaions from the physical shape but in
Al's case he very carefully (As I Recall)mapped the equator of each and determined them
to be accurate.

However, as Slash noted above, moving the magnets in the chord can produce variations
in the field shapes. For instance, pushing all the marked ends (we'll arbitrarially call this
north) together such that the south ends are much farther apart would produce an odd
scallop effect with sharp N fields and rounded S fields. And the dwell time N would be
small while for S would be large. So we trade density for dwell. This would cause the
stator to have to speed up and slow down for each rotation necessitating a resonance with
inertia to function at all.

andromeda- Jan 20th 2007


Grimer,

With all due respect old boy, the needle on your meter is pushing redline.

Maybe best to get some sleep and make it a new day?

/- Jan 20th 2007


Franx:

Harvey:However, as Slash noted above, moving the


magnets in the chord can produce variations in the field
shapes.

I read that but didn't understand what was meant by "moving the
magnet in the chord".
Is this some analogy to the chord of an airfoil ?

I meant it like this - illustration.


If the magnet is moved, but only around the circle, the field outside is still
symmetrical (1a & 1b). If it's moved on the chord, the field as perceived from the
outside changes it's shape and it's not perpendicular to the tangent (2a & 2b).

vipond50- Jan 20th 2007 from OU


I did a count by hand a found that the stators rotate four(4) rev's/ one(1) rotation of the
rotor. Has there been a standard for determining the magnet gauss? Read about various
methods, but no standard that everyone can follow to get accurate results when
compared.

Need to develop a procedure for this


omnibus- Jan 20th 2007 from OU
Please try the following--adjust by hand the rotor at one of the larger maximums and let it
go. It should skip the next rotor magnet and will bump into the maximum formed by the
third. That's the pattern that should be repeated. See if it is indeed repeated four times per
full turn.

As for the procedure of measuring just place the probe on the face of the magnet pole and
compare the readings with what you'd get by placing the same probe on the same place
on the face of a pole of another magnet. That's very rough estimate but will give you
some idea.

Bruce_TPU- Jan 24th 2007 from OU


N35 Separation:
28.5 mm = N35 magnets

Al's Separation:
30.5 mm = ?

N42 Separation:
34 mm = N42

Omnibus Jan. 24, 2008


@btentzer,

Schematically it should be like this:

Measure the separation in this way first

S
|
|
|
N
separation to be measured
N
|
|
|
S

and then this way

N
|
|
|
S
separation to be measured
S
|
|
|
N

blue_energy Jan. 24, 2008


With my magnets which are supposedly N42 the first measurement gives something on
the order of 27mm while the second measurement gives 33mm for the same pair of
magnets.
Well - so much for the 'stator deceleration at critical points' theory. The rotor and stator
look as if they are connected with a gear or a belt. And as I pause the video and look at
the relationships between the magnets - I don't understand what would cause the rotor to
accelerate. At that stator speed of over 4,200 rpm the other two stators must be stopped.
Presuming that the other two stators break even in push and pull, what is there about this
one relationship that would cause the rotor to accelerate if not for the dampers? Actually,
I would expect the entire system to slow down - not speed up. Try freeze framing and
ask yourself what magnetic forces are currently in play. Do you see any way for the rotor
to do any better than break even? I'm consused.
gdr Jan. 24, 2008
I guess the dampers do help, but only after the unit already gets the effect ('the effect'
being acceleration of the rotor). Al said that the dampers are there because the
rotor/stator speeds up too much and then loses sync (we should have such a problem ).
My thought is that adjusting the friction in the stators may help to get the effect to start
with. I actually think some of the replicators are too good at what they do, and might
have trouble getting things working because their stators spin too freely.

PolyMatrix Jan. 24, 2008


Looks to me like the stator magnets are just above the rod magnets, which would be in
keeping with some the conclusions I had while testing out some ideas with magnets. The
stator magnet in my theory has its bottom edge just above the top edge of the rod magnet.

Edit Also concerned that whether the rod magnets are precisely cantered. It would only
take 0.25mm to be significant.

robbie47 Jan. 24, 2008


@Omni,
I thought about your remarks a little longer.
At the final speed there must be a driving force that exactly compensates the losses (e.g.
friction, air draft)
This makes the amount of losses extreemly critical as well and therefore the quality of the
applied bearings.

At lower speed there must be a driving force higher than the losses at lower speeds, at
least in Al's first video.
Also at lower speeds the avaraging effect of the stator will be less. There is most likely a
wobbling phase effect.
This effect will still be there at high speed but less, due to the higher flywheel effect.
Mmmmmmm, need to think a bit longer on this still....

MeggerMan Jan. 24, 2008


Repulsion results:

N42 1/4"OD x 1/2" rod vertical repulsion = 31mm


N38 6mmOD x 12mm (6 of 6mm OD x 2mm discs) vertical repulsion = 28mm
N42 Ring 1/2" x 3/16" x 1/4 repulsion to N42 1/4"OD x 1/2" rod = 38mm
Tried both orientations with all the above and I could not see any difference.

Al's Results (from memory):


N?? 1/4"OD x 1/2" rod vertical repulsion = 30.5mm
N42 Ring 1/2" x 3/16" x 1/4 (with bearing) repulsion to N?? 1/4"OD x 1/2" rod = 37mm

Hugh Jan. 24, 2008


Sean - I had some time after all and could make the following comparisons to show
how sensitive the gain is to parameters: well, okay, some of the changes to go to Al's
setup were a bit large:
I will now try to see how far I can 'bend' my old sim to Al's setup whilst retaining a gain.
My 'old'setup
Strength of rotor magnet/ Teslas ----------------------- 0.05
Strength of Stator magnet/ Teslas ----------------------- 0.20
Viscous lag time / Sec ----------------------- 0.001
Radius to outer rotor pole / m ----------------------- 0.100
Radius to inner rotor pole / m ----------------------- 0.085
Radius of Stator magnet/ m ----------------------- 0.025
Initial Stator angle / Deg ----------------------- 20.0
Mass of rotor / Kg ----------------------- 0.200
Mass of stator / Kg ------------------------ 0.0025

*KE Gain / Joules after 6.5 sec 0.018*

-------- Al's setup -----------------


Strength of rotor magnet/ Teslas ----------------------- 0.10
Strength of Stator magnet/ Teslas ----------------------- 0.20
Viscous lag time / Sec ----------------------- 0.001
Radius to outer rotor pole / m ----------------------- 0.072
Radius to inner rotor pole / m ----------------------- 0.052
Radius of Stator magnet/ m ----------------------- 0.025
Initial Stator angle / Deg ----------------------- 20.0
Mass of rotor / Kg ----------------------- 0.254
Mass of stator / Kg ----------------------- 0.010

Al's setupKE Gain: Negative for inital stator angles 30, 40, 50, 120, 210, 300, 330 deg.
(though for some of these the rotor spun down, giving peaks of gain of about 0.01 J
briefly (just PE -> KE -> PE peaks)
before the gain is converted into a loss of hundreds of Joules!)
The mass of the stator was the first parameter I tweaked to go from my 'old' sim to the
OCAL one. In the old sim, it was rather low at 0.0025 kg. A figure quoted for OCAL has
been 10g (5g for the magnet and 5g for support, holders ec.). It seemed unrealistic to
have anything much lighter, so I set it to 10g.
However, with this heavier stator the gain disappeared!!
Thus it seems the gain is very sensitive to weight of stator. It would be good to get it
down as low as possible. Since you had to zonk it up with blue-tack, you might have had
a light stator to start with: I would take off the blue-tack if I were you.
In order to balance the heavy stator, I thought to increase the rotor weight. At 400g, the
gain was there again ! But it was only 0.0015J - less than a tenth its previous value. If I
increased to 800g I got 0.004J and there was little to be gained by increasing rotor weight
further. So I adapted a weight of 800g to be consistent with stator weight 10g.
Next magnet strength was considered: my old sim had stator stregth 0.2T - which I
believe is what OCAL had. But my rotor strength was only 0.05 Teslas. I now increased
it to 0.10 Teslas, which I believe was closer to OCAL's value. This had a positive effect
on the gain, increasing it to 0.017 J : almost what we were getting for the old sim.
By the way - I always have start speed of rotor about 190 RPM and set stator to be
always in 4:1 AGW synch..

Omnibus Jan. 24, 2008


@MeggerMan,
I have the results for the vertical repulsion of the batch of 12 new N42 1/4"OD x 1/2" rod
magnets. The separation is between 31mm and 31.5mm both ways. Same as yours. I don't
have these magnets with me right now. Hope to have them soon here so that I can
measure this myself as well as the kgauss on both sides and also try them in my rig.
The measurement of N42 Ring 1/2" x 3/16" x 1/4 (with bearing) repulsion to N?? 1/4"OD
x 1/2" rod has to be discussed more so that our results match those of @alsetalokin.
That's an even more important measurement than the above.
Forgot to mention--all rotor magnets are 3g each.

mudwump Jan. 24, 2008


This is in reference to the strobe video. One thing that I noticed is the the magnets are not
symmetrically placed around the rotor. It appears that there about a 2mm jump in the
various placements. This might be important for an asymmetrical force for acceleration.
evil-doer Jan. 24, 2008
the strobe is NOT in sync with the rotor per full rotation, you are seeing 4 different
magnets in each location. the same as you see 4 positions of the tape. this is why you can
see the magnets are not positioned equally in one of the slots. the question is.. does this
help or hamper the device? if they were more even would it go faster or longer, or would
it not work at all?
RB Jan. 25, 2008
I thought that at first, but look at the relective strip on the rotor, it only shows up in 4
positions, which would make it 1155 RPM.
btentzer Jan. 25, 2008
Yep, and it is not a magnet. Blue_ hit on it some time ago. It is a golf tee. It is non
magnetic and make it easy to move the damper around for testing. Look at the center
dimple. A golf tee no doubt, and no joke! LOL
@ Rob
Strange that your N38's are 28 mm separation and my N35's are 28.5 mm (tested both
ways, no difference)
@ Clanzer
It will be interesting to see your wind down test results with the identical alloy that Al is
using, compared to the aluminum.
Also, if you could please ask AL to measure the separation from the black ends of the
rotor magnets and the separation between the silver ends. Jason can figure it out, but it
would be far more accurate to have Al simply measure it for us. This could also, make a
huge difference between a working replication, and one that does not. Thanks!!
@ Omni
Still no success finding a magnet with 30.5 mm separation. We need to keep looking
Jdo300 Jan. 25, 2008
@All,
After seeing Al's third video with the strobed images, I am very relieved to see that the
stator magnet is rotating at a constant velocity. I have read about the many proposals to
artificially reproduce the acceleration using motors. I've given some thought to this
myself and it seems to me that we should look into actually locking the stator and rotor in
a 4:1 gear/pully ratio. In this way, we can simply var the phasing of the system and then
use a motor to spin the stator to see how that affects the power consumed by the motor.
(But again, this is assuming that the stator velocity is constant as the video strongly
suggests).
Also, now that I can analyze the stator magnet position with respect to the rotor. I can
almost do some simulations to show the torques on the stator and rotor. But before we go
there, we need to know if the rotor magnets are, in fact, off center in their slots in the
rotor. I'll wait to see if Al comments on this but, if not, it would be a small matter to
simply add that as a parameter in the simulation to get some preliminary results.
blue_energy Jan. 25, 2008
I took the video apart. As it turns out, the rotor is striped radially in quadrants - except
that the fourth quadrant doesn't have the stripe. This allows us to positively identify the
location of the rotor in each frame. The first frame or two happen before the strobe turns
on - so I skipped them. Then, the next 4 frames happen with the strobe turned on and I
captured them. Notice that the stator seems not to move almost at all. But, in reality, we
see that the stator has moved ~360 degrees (follow the location of the missing stripe in
the rotor in the four frames - showing that it has moved 90 degrees in between each
frame). Now I'm not as sure about the reliability of my earlier statements regarding the
lack of speed changes in the stator. Can we really tell whether the stator slows down or
not if each frame represents slightly more than 360 degrees of stator motion? To tell the
truth, I'm not sure what it all means anymore...

evil-doer Jan. 25, 2008


well.. i think youre wrong. in the time the camera has taken one picture frame the strobe
has flashed 3 times. thats why you see 3 strips, its all the same strip, in 3 places.
this can easily be seen by the identical writing at the end of all 3

geodan Jan. 25, 2008


that would confirm 4:1, every time that the rotor moves 1/4 turn the stator has gone
around 360...

blue_energy Jan. 25, 2008


So - you think the rotor is moving 270 degrees between each frame? That would make
sense too. That would also mean that the stator is revolving 1080 degrees between each
frame.

Omnibus Jan. 25, 2008


When I count frame by frame 1s is the 25th frame.
As far as I can see visible full turn of the stator (restoring the initial black-white position)
takes about 120 frames, that is 4.8s.

evil-doer Jan. 25, 2008


the video says the strobe is set to 4620 rpm (im guessing flashes per minute)
/60 for 77 per second
/25fps for 3.08 flashes per frame. so it makes sense we see 3 strips
because each strobe is 90 degrees more than the last im guessing the device is spinning at
4620x1.25=5775rpm

ya, im just guessing

edit: ok i take that back, its probably slower than 4620, so youve gotta divide by
something. im too tired to try to do math or think

Omnibus Jan. 25, 2008


@evil-doer,
By 'device' you mean the rotor (speaking of strips), correct? That's too high, isn't it? The
stator then should rotate four times faster. Shouldn't it be 0.75 x 4620 = 3465rpm because
we see three strips per frame (not always, though, some frames show 2 strips only,
sometimes the follow up frame being the same 2-strip pic). That seems high too
considering what the stator rpm should be then.

mikeytown2 Jan. 25, 2008


CLaNZeR Thanks for putting up Alsetalokin's Third Video! If you could somehow get
me the raw Interlaced PAL video that would be helpful. It appears that the deinterlacer
used blends the fields. I can use a motion compensated bob deinterlacer which will give
us 50 fps video. Here is your video at 5 fps forwards then reversed, then zoomed. Sorta
took care of the flicker as well.
http://www.youtube.com/watch?v=7V8y7qfO9gw
Also if you guys are taking measurements from pictures, because of lens distortion, it
won't be exact. If the ORGINAL jpg still has the EXIF data you can use this to
undistort the image
http://www.brothersoft.com/gml-undistorter-download-42391.html
Otherwise use this GIMP plugin
http://wiki.panotools.org/Gimp_wideangle_plugin
GIMP
http://gimp-win.sourceforge.net/stable.html
WhiteLite Jan. 26, 2008 FizzX
Fantastic video! Thanks for doing this Al.
I guess it confirms that the stator is in attraction when between rotor magnets and in
repulsion when next to one. It seems slightly off though, as in the stator magnets poles
are parallel to the rotor just slightly after one of the rotor magnets has passed. This may
be down to the black colouring of the stators pole not being dead on although I think Al
said he used magnetic film to determine where to colour the pole. Might be significant?

Al Jan. 26, 2008 FizzX


Thank you for bringing this up, WL. In the strobe video I am using a different stator
magnet/bearing pair than in the first video--the other of the two good ones I have. I just
re-checked the marking with the viewing film, and you are right, I marked this one a
little off. Less than 5 degrees, but a perceptible amount. If you are looking down at the
top of the magnet, the true pole positions are rotated a couple of degrees clockwise
from the markings on the magnet. Sorry! I hope a description is good enough, it may be
some time before I can post a corrected strobe video.

Al Jan. 26, 2008 FizzX


@OC: Of course not--at least not yet! Apparently 7.5 hours is no big deal and doesn't
qualify for the "sufficiently long time" rule. No need to go further until it can be shown
that either A) it will run longer than any onboard power source will allow or B) a dummy
unit containing an onboard power source can be allowed to run longer than the unit.

And in answer to dear Babcat's famous question: Leaving aside aber0der's point below (I
believe the Bible because the Bible tells me to, and so forth) consider the following
points: Either the unit runs of itself ( free energy, OU, cohering the zero point, whatever),
or it doesn't (motor and batteries, the RB effect, cosmic ray inertial transflux, whatever).
And either I am lying about it, or I am not. So there are 4 distinct possibilities. 1) Free
energy device, and I am telling it just like it is (the answer to our prayers). 2) Free energy
device, and I will lie about it (to get everyone off my back and keep the profit for
myself). 3) No free energy device, and I am telling it just like it is. (The real case, by the
way.) 4) No free energy device, and I am lying about it (hoax of some kind).
Now, obviously, my motivation to lie is strongest in Case 2, if there was a real free
energy device like the "WhipMag". If I did have a functioning permanent magnet motor,
it would be to my great advantage, at this point, to lie about it and to say that it was fake.
Right?
So OK, Babcat, just for you: It's got a motor and a battery in it. Somewhere. A
transparent motor and a glass-battery.
(but actually Case 3 is the real deal).

OC Jan. 26, 2008 FizzX


That wobble may also cause some binding in the bearings which will tend to retard stator
rotation when attraction/repulsion is greatest. This might actually help maintain the
correct orientation to leverage more force.

Meggerman Jan. 26, 2008 FizzX


Hi Al,
I have test that I would like to know the answer to, so that we can find a matching bearing
for the stator.
TEST:
With the stator mounted on the baseplate, hold a metal rod or ruler say 20cm long against
the side the the stator and check the play distance at the top of the 20cm rod/ruler.
This will give us a guage as to how loose the stator bearing needs to be to get the right
amount of wobble. You said some stator sets would not work at all, other ok and some
very good, so I wonder if it has nothing to do with the magnet in them but more to do
with the wobbliness (if there is such a word) of the bearing.

Harvey Jan. 27, 2008 FizzX


@Al Edit (And all replicators )
Did we ever get a gauss meter?
Crucial data needed:
1. Static readings (away from stator) of of each pole set at leading equator, at max
reading(near center between each rotor magnet), and at trailing equator.
2. Same as above in leading zone and trailing zone at stator.
3. Dynamic readings (averages) in leading zone and trailing zone.
I have been mentally designing a piggy-back gauss sensor to ride on the rotor for some
relative tests. Quite simply it would consist of a two hall effect sensors mounted 180
degrees out on a balanced arm that allows full adj. through all zones. (visualize a flat
acrylic bar with hole drilled center resting on the rotor with sensors out at the ends). At
the center above the spindle shaft is a mass balanced module containing a small battery,
opamps, vcos and pwm circuits. The pwm would output to an IR led. The pulse width is
indicative of the density readings. Each sensor will have its own frequency range. A
linear approach would be to vary the brightness relative to the field densisty and would
be simpler to construct.
Of course the test depends on the added mass not preventing the device from functioning.
OC Jan. 27, 2008 FizzX
If you review my initial design, you will see that I required a latching mechanism to
prevent the stator from rotating away too quickly when in repulsion with the passing rotor
magnet. The purpose was to allow the repulsive forces to be applied to the rotor and not
merely spin the stator around.
If the stator bearings wobble and bind and retard stator rotation when a strong repulsive
force is felt, it will effectively perform the same function as my latches. It will retard
stator rotation enough to allow some of the repulsive force to be applied to the rotor
magnet that has just passed.
Wobbling may affect the attractive relationship as well. In this case it may actually be
detrimental. It might be a good idea to figure out how to retard stator rotation when we
want it (in repulsion) and how to allow free rotation for the rest of the time.

Harvey Jan. 27, 2008 FizzX


Since the repulsion occurs during the passing of equators the greatest repulsive forces are
perpendicular to the rotation and thus have little effect on the torque other than bearing
loading. There is a narrow region here where the net result can be negative torque on both
rotor and stator.
If the stator were mounted in an oval slide that allows movement perpendicular to the
rotation and it were spring loaded, then a friction pad could be placed on the back side of
the stator opposite the rotor that would delay the rotor slightly while in contact. However,
at the speeds we are working with the mechanical resonance of such a mechanism would
be crucial and it is doubtful gains would offset the losses.
I believe a better approach would be to draw the equatorial flux away from the stator with
some carefully designed shunts or custom magnet shapes.

niente edited Jan. 26, 2008


Here's a simple analysis (by using FEMM) of the torque of the rotor covering 45 degrees
of the motion (then it repeats itself because of the symmetry).
When the fulcrum of the rotor and of the stator are connected with a line, and one of the
magnets on the rotor is perpendicular to this line, the stator position shows a difference of
about 25 degrees in respect of that line (as in the third Alsetalokin video). This fact is
included in the simulation (visible here:
http://www.metacafe.com/watch/1060658/oc_mpmm_magnetic_simulation/).
The simulation data and graph are included in the attached Excel document, and they
show a total positive torque (it should be 0 if there were no self-rotation!).
Sorry for any mistake I did!
[EDIT] I made a little mistake in the Excel file, now the total number of samples is 449
(not 451) and the total torque is correct (not 222 but 11467!!) - I typed a ";" instead of a
":" in the sum formula [/EDIT]

Jdo300 Jan. 26, 2008


Hi Niente,
There is a problem with your simulation setup. If you look at the square boundary around
your rotor disk, you will notice that all of the field lines intersect the boundary at 90
degrees. This shows one of two things. First, the boundary is way to close to the problem
region. This distorts the fields and will produce inaccurate results. Secondly, for magnetic
simulations, it is best to use a circular boundary and boundary conditions that
approximate an infinite problem region to get the most accurate results.
For my simulations, I wrote a simple LUA script that creates a circular boundary of the
desired radius and automatically calculates and applies the correct boundary condition for
the problem. It also adds a block label set to "Air" for the problem region. Though you
will have to go in and set the mesh density to what you want. I have attached the script
file below.
Also. I have made a simulation of the rotor disk (to scale) with a single stator magnet for
everyone to play around with. I wrote an accompanying script file that will rotate both the
rotor and stator magnet either clockwise or counter-clockwise depending on the settings
in the script. You can set the initial starting angle of the stator and can also change the
gear ratio (at the moment it is set to 4:1). It will also print out a data file with torque
values at each position of the wheel. You can just copy and paste into Excel and make a
graph from there for further analysis. The only thing is I don't set the initial position of
the rotor so if the simulation gets stopped for any reason, you'll have to manually select
the rotor magnets and rotate them back to the starting position you want. This script file
is also included with the simulation file and the boundary setting script. I tried to make it
easy to understand but If anyone has any questions don't hesitate to ask.

MrEntropy- Jan 27th 2007 from OU


How many replicators have been careful to use ferromagnetic bearings and to put them
right up close to the stator magnet like Al did? I've been thinking recently that the
magnetization of the stator bearing, and especially its saturation, may be very important
to the effect, since all the other ferromagnetic materials involved are very hard.

Omnibus- Jan 27th 2007 from OU


@All,
Let’s compare some notes.
My rotor weighs 303g.
That is HDPE piece + the two bearings + the eight N42 1/4"OD x 1/2" rod magnets.
My stators weigh 10g each.
That is the Delrin holder + the bearing + N42 1/2" x 3/16" x ¼ ring magnet
Magnetic induction of the eight N42 1/4"OD x 1/2" rod magnets I’ve worked with is as
follows:
kgauss
S N
4.56 -4.49
4.61 -4.37
4.71 -4.40
4.50 -4.45
4.60 -4.50
4.57 -4.48
4.58 -4.42
4.51 -4.42

x(av) = 4.58 x(av) = -4.44


s(n-1) = 0.0655 s(n-1) = 0.0464

Magnetic induction of the twelve virgin N42 1/4"OD x 1/2" rod magnets (just arrived) is
as follows:
kgauss
S N
4.42 -4.48
4.57 -4.30
4.56 -4.37
4.56 -4.43
4.48 -4.28
4.62 -4.43
4.58 -4.51
4.66 -4.77
4.60 -4.57
4.58 -4.38
4.62 -4.47
4.61 -4.50

x(av) = 4.57 x(av) = -4.43


s(n-1) = 0.0672 s(n-1) = 0.0864

The vertical separation of couples of N42 1/4"OD x 1/2" rods placed in repulsion both SS
and NN is approximately 31mm. Needs a more snug clear tube to avoid the tilting,
although quite small, of the magnet couple, to improve accuracy of measurement.
One thing that seems evident is that there is no measurable degradation of the used neo
magnets compared to the virgin ones. In both sets studied, however, there’s a difference
in the absolute value of the average magnetic induction measured for the S compared to
the N pole for the same precision (standard deviation), that precision slightly varying for
the N pole. Someone curious may do some further tests such as student t-test to see how
significant that is but that doesn’t seem to be necessary.
My gaussmeter is F.W.Bell Model 4048
Dynz- Jan 27th 2007
Man, this is a great video(Note: the strobe video, Al's third)... I have been drawing out
magnetic field lines at different stages of rotation, and it seems as though the dampers
would have a noticeable effect on the spin of the stator... Their position is such that the
one on the top would create an induced clockwise field at the same time as the one on the
bottom would create an induced counterclockwise field. (I am assuming the dampers are
just chunks of aluminum)

The way the rotations are syncing, these induced fields would help pull/push the stator
around AGW. This could explain the anecdotal evidence we have been hearing about
dampers making sync easier.

If this is true, the position of the dampers will matter. You want them close enough to the
stator to influence its spin, but far enough apart that they are moving through opposite
magnetic fields (from the rotor) at the same time.
Also, this would help explain why there seems to be a necessary rotational velocity
before we can achieve AGW lock. Up until a certain point, the oscillating field induced in
the damper would be too weak to do much to the stator... It is a bit of a weird situation,
because the induced fields are going to vary with the speed, while the PM fields are going
to be relatively constant.
Background: The fact that AGW lock does not occur until a certain rotational velocity is
met is a huge mystery to me. As far as I can think it through, the only material*
characteristics that would be changing with rotational velocity are the strength of any
induced magnetic fields/electric currents, and "magnetic lag" (which I am not convinced
about in general, and certainly wouldn't seem to have an effect here regardless.)

* - frictional coefficients will change too, but it doesn't seem like this would account for
the behavior.

My point is that the same resonance should occur at any speed, excluding induced current
and/or some sort of "magnetic lag" effect.

- The ratio of rotary velocity between the rotor and stator would be the same at any speed.
- Inertia is dependent on mass. The mass is constant and so the inertia will not change.
(the momentum changes, but because the mass is constant, this just the same as saying
the velocity changes, which shouldn't matter for setting up the "resonance" as long as the
ratios are constant)
- Magnetic effects may change at different speeds, however the changes should be either
as a result of eddy currents (induced electrical currents) or "magnetic lag"
--- Eddy current strength will change with speed, and could give one explanation as to
why the AGW lock only occurs at certain RPM.
--- "Magnetic Lag" (the time it takes for magnetic domains to align in a chunk of
ferromagnetic material) would also be something that gave a variance with speed,
however I can't see how it would make a difference here - The aluminum dampers are not
ferromagnetic, and the magnets themselves are not going to be doing a lot of domain
flipping.

Therefore the fact that the AGW lock "resonance" only occurs at certain speeds must be
due to either the effect of eddy currents (most likely), or some sort of "lag" effect (very
unlikely given the materials, but might explain al's observation that some stators worked
better than others in achieving the effect. EDIT: actually I guess it could be explained by
eddy currents and different conductivity of magnets as well)

Inertia is simply the principle that velocity will not change without a force acting on the
object. The resistance to change in velocity (one possible meaning of inertia of an object),
varies directly with the mass, and so will not change with speed. (http://en.wikipedia.org/
wiki/Inertia)

If you are using "inertia" to mean momentum, momentum varies directly with speed in a
fixed mass system, and so the ratio of momentum will stay constant. Mass (and therefore
all talk of "relative densities") completely drops out of the equation when we are talking
about change and ratios.

If the mass is constant (which it is) it makes NO difference when we are talking about
changes of momentum.

Franx- Jan 27th 2007


The difference between GW and AGW lock is that the rotor and stator at their closest
point are in attraction in GW but are in repulsion in AGW lock.

This state of repulsion is highly unstable and can't be sustained while the rotor is stopped
(The stator will simple flip 180 degrees so that the magnets are in attraction).
Similarly it can't be sustained at a very slow speed for the same reason.
There are three factors which can prevent this flipping action. All three are inter related
and inter dependent.

1) The speed: At higher speed the stator doesn't have time to flip

2)The mass/inertia of the stator: This determines the time it takes for the stator to change
speed. If this is high enough it will prevent the stator form flipping even a lower speeds.

3)The strength of the magnets: The stronger the magnets the more unstable the repulsion
state will be. This will make the stator more likely to flip for any given mass or speed.

So to get good AGW lock these three factors must be balanced.

PS you could add to that the air gap though this will have a similar effect to (3). The
smaller the air gap the stronger the magnetic reaction.

Harvey- Jan 27th 2007


Rotational Inertia: I=Σmi ri²
Angular Velocity = ω = θ/t
Momentum: L = Iω

Aluminum is paramagnetic and subject to domain propagation as well. These interactions


occur near c and thus 'magnetic lags' are really not a contributor to these events. The
dampers help regulate the speed of the rotor preventing rapid accel/decel without adding
mass to it.

Since L is dependant on ω the speed is an integral part of the function. Additionally,


during acceleration the I is an integral part of the stators ability to respond to the rate of
increase. If I is too great the flux will shear and the stator will drop sync. (see my strobe
stills for examples of this). If I is too low it will adversely affect L preventing effective
repulsive shear which is needed for smooth operation.

The magnetic ratio of two N35 magnets to one N42 magnet at the fixed distance places
the density reactions in a specific space between them. These are critical in as much as
the turning through these zones must facilitate the observed 4:1 turning ratio.

Because both of the previous two points are valid, the relationship between them must be
conserved. Consider the use of an N50 on the stator and N42's on the rotor. The ratio is
conserved but the repulsion zone increases in width resulting in a braking of the stator
because of its inertial limits. The speed would have to increase in order to accomodate the
needed L in this case.

As far as relative densities we need to qualify this as either mass or magnetic. If we are
discussing mass, this is constant at these speeds and thus drops out. If we are discussing
field density then the matter is very dynamic. Because the rotor magnets are set in
opposition there is a continuous exertion of each on the other to reorganize the weiss
domains and over time the overall strength of the system would degrade. However, when
interacting with the stator this effect is somewhat neutralized as the stator interacts in an
attractive mode for the majority of the cycle. The sum total of field densites for any three
magnets in interaction should not change. However a differential between the
approaching fields and the retreating fields is highly possible. This would indicate a
higher 'relative density' in the approach zone and a lower 'relative density' in the retreat
zone. And this differential is where I believe we are recieving our gains. It would be a
form of doppler compression of a gradient field.

hdeasy- Jan 27th 2007


Don't forget that the ease of locking on when you increase stator weight is bought with
a high price - i.e. you then are less likely to get OU. Lighter stator, by contrast, is hard to
synch but gives better OU. I just confirmed this in my sim. By the way my earlier sim
was a bit optimistic - motion alone was not enough to get OU: that was due to an
outstanding bug in stator torque. But I'm much happier with the sim now: BR agrees with
most of the results now. And we agree now that BH movement alone without lag may
be enough to give OU.