Professional Documents
Culture Documents
04 November 2006
In the October 22nd blog, I mentioned how I fear schematic typos. And then, in
the October 31st blog, I made a big one. At least it was a big one in my eyes,
though many will never have spotted it.
And while I bemoaned the e-mail burden with which I daily struggle, I didn't
explain that I had fallen particularly far behind of late because two weeks ago, my
son’s teacher and one of his classmates drowned. A week spent trying to explain to
a three-and-a-half-year-old how something like this could happen and what it
means, along with attending various funeral-related events, put me far behind in
work in general and answering e-mail in particular.
I reasoned that in just a little more time than it would have taken me to answer
the original lost e-mail, I could write a blog entry addressing the same issue. But I
didn't have time to do it complete justice last week. So I decided to write a second
part; and when I gave the original entry a second reading, I spotted the schematic
typo. (October 31’s blog now holds the corrected schematics.)
Notice the parts in red. This capacitor serves as a power supply decoupler,
filtering away much of the 510-volt power supply rail’s noise. So what’s wrong?
Isn’t this what 99% of tube gurus would do? (I was tempted to write 99.9% of tube
gurus, but upon some reflection, I realize that percentage would imply that the
world holds at least 1,000 tube gurus, a number I doubt.) The problem is that if
this decoupling capacitor terminates into ground, the negative power supply rail
noise will be superimposed upon the top output triode’s grid signal. How’s that?
The differential amplifier’s common cathode resistor attaches to the negative
power supply rail and any current variations through this resistor that result from
the resistor seeing a varying voltage potential will be relayed to the plate resistors,
as they all fall on the same current path.
But in the correct schematic below, we see the decoupling capacitor terminated
into the positive-power-supply rail, not ground.
Now the negative power-supply-rail noise will meet and cancel with the positive
power-supply-rail noise, as the cathode and plate resistors values were chosen to
ensure a noise null at the output. If the cathode resistor were larger, then the
output would see positive power-supply-rail noise leaking out; if the resistor were
smaller, then the output would see negative power-supply-rail noise leaking out
instead. (Ideally, we want injected negative power supply rail noise to equal the
positive power-supply-rail noise, so that cancellation is complete.)
How does this trick work? Both the positive and negative power supply rails
deviate from being pure DC sources, carrying the rectification noise residue.
Fortunately, the noise on each rail is nicely related, equal in magnitude, differing
only in being 180 degrees out of phase to each other. (This only holds true when
both rails see the same current load, as they do in this amplifier.)
The complete headphone amplifier is shown above. By the way, speaking of noise
reduction, I just reread something that I had written in the second issue of the
Tube CAD Journal, which is worth repeating every seven years:
Given the choice between low distortion and low noise, choose low
Given the choice between low distortion and low noise, choose low
noise. The difference between .1% distortion and .01% distortion
cannot be heard; a reduction of noise from -60 dB to -80 dB can. If
you want to improve the sound of your tube equipment, lower the
noise. Of course, low noise and low distortion are not mutually
exclusive and it would definitely be better to have both.
Here is the irony of the situation. Most solid-state gear is very quiet,
because of the low distortion race among stereo manufacturers. In
order to please the total harmonic distortion meter, an amplifier must
be both clean and quiet, as the meter cannot distinguish between
distortion and noise. Whereas, the tube audio designer, having no faith
in total harmonic distortion measurements, holding to his belief that
no meter can beat his ears, forgoes the requirement of extremely low
noise operation.
Some tube audio designers are very pessimistic about the noise given
off by their designs. They tell us that the vacuum tube is an inherently
noisy device and that it is simply not possible to match the solid-state
gear in low noise operation. We might believe them, if it were not for
counterexample offered by some of the best tube gear: tube
microphones, preamps, and amplifiers that do not hum along with the
music or hiss like tea kettles. (I built a direct coupled electrostatic
headphone amplifier that is so quiet that a friend used to borrow it to
evaluate the noise of hand picked transistors for an ultra-low noise
preamp he was building. The headphones and amplifier revealed more
information about the amount and quality of the noise than did
looking into an oscilloscope.) There is simply no excuse for one tenth
of a volt of hum from the output of a tube power amplifier.
We are not saying that the tube is just as quiet as the transistor or FET,
as, ultimately, it is not; only that much of what is built with tubes
could be much quieter.
http://glass-ware.stores.yahoo.net/
ET-T3021
POWER-AMPLIFIER PENTODE
MECHANICAL
Mounting Position—Any
Envelope—T-6½, Glass Aikido PCBs for as little as $24
Base—E9-1, Small Button 9-Pin Outline Drawing—EIA 6-4
Maximum Diameter 7/8 Inches http://glass-ware.stores.yahoo.net/
Maximum Over-all Length 3-1/16 Inches
Maximum Seated Height 2-13/16 Inches
Only $12.95
to keep track of your
tube and part collection
TCJ My-Stock DB
MAXIMUM RATINGS
DESIGN-CENTER VALUES
Plate Voltage 250 Volts
Screen Voltage 200 Volts
Plate Dissipation 12 Watts
Screen Dissipation 1.75 Watts
Peak Screen Dissipation 6.0 Watts
DC Cathode Current 100 Milliamperes
Heater-Cathode Voltage
Heater Positive with Respect to Cathode 100 Volts
Heater Negative with Respect to Cathode
DC Component 150 Volts
Total DC and Peak 300 Volts
Grid-Number 1 Circuit Resistance
With Cathode Bias 1.0 Megohms
TERMINAL CONNECTIONS
Pin 1—Internal Connection
Pin 2—Grid Number 1
Pin 3—Cathode and Grid Number 3 (Suppressor)
Pin 4—Heater
Pin 5—Heater
Pin 6—Internal Connection
Pin 7—Plate
Pin 8—Internal Connection
AVERAGE CHARACTERISTICS
Plate Voltage 170 Volts
Screen Voltage 170 Volts
Grid-Number 1 Voltage -12.5 Volts
Plate Resistance, approximate 23000 Ohms
Transconductance 10000 Micromhos
Plate Current 70 Milliamperes
Screen Current 5.0 Milliamperes
CLASS A1 AMPLIFIER
Plate Voltage 170 Volts
Screen Voltage 170 Volts
Grid-Number 1 Voltage -12.5 Volts
Peak AF Grid-Number 1 Voltage 9 9 Volts
Zero-Signal Plate Current 70 Milliamperes
Maximum-Signal Plate Current 70 Milliamperes
Zero-Signal Screen Current 5.0 Milliamperes
Maximum-Signal Screen Current 22 Milliamperes
Load Resistance 2400 Ohms
Total Harmonic Distortion,approximate 10 Percent
Maximum-Signal Power Output 5 6 Watts
In practice, this amplifier may work well enough that these added costs are worth
paying. Remember, the 6N1P can be much quieter than the 12AX7.
-+510V
39k 39k
167+170V
운후
^No
TCJ My-Stock DB helps you know just what you
6B05 100 +165P1 have, what it looks like, where it is, what it will be
W
400v
33k used for, and what it's worth. TCJ My-Stock DB
helps you to keep track of your heap of electronic
1470uF
parts. More details.
150V
630v
200v
TrOuF
250v
Version 2 Improvements
5V OV output OV List all of your parts in one DB.
Add part Images.
One-click web searches for part information.
Vertical and horizontal grids.*
•10k Create reports as PDFs.*
1M 6B05
₴100 Graphs added 2D/3D: pie & bar.*
More powerful DB search.
1470uF
1N4001 Help system added.
200v
10uF Editable drop-down lists for location, projects,
1250v
brands, styles, vendors and more.
16.
163143 *User definable
4AT
,708
163
4AT
,708 Download or CD ROM
www.TubeCAD.com Windows 95/98/Me/NT/2000/XP
For more information, please visit:
Another input tube that I would experiment with is the 12BZ7, which also holds a
mu of 100, but with half the rp of a 12AX7. www.glass-ware.com
//JRB
To purchase , please visit our
Yahoo Store:
http://store.yahoo.com/glass-ware
Only $9.95
to start designing tube-based
crossovers and much more...
Bessel,
Butterworth,
Gaussian,
Linkwitz-Riley.
While the program’s main concern is
active filters, solid-state and tube, it also
does passive filters. In fact, it can be
used to calculate passive crossovers for
use with speakers by entering 8 ohms as
the terminating resistance. Click on the
image below to see the full screen
capture.
Download or CD ROM
Windows 95/98/Me/NT/2000/XP
For more information...
http://store.yahoo.com/glass-ware
< Back www.tubecad.com Copyright © 1999-2006 GlassWare All Rights Reserved Next >