|Views: 200|Likes: 2

Published by mohsinx111

See more

See less

38 Jan/Feb 2005

11581 Aspendell DrSan Diego, CA 92131

Why Antennas Radiate

By Stuart G. Downs, WY6EE

Antenna theory is a popular subject with hams. We

love to read and talk about it. Now put on your

thinking cap and fasten your safety belt for a

review of the math and science behind it.

1

Notes appear on page 42.

M

astering an understanding of electric and magnetic fields isnot easy. The electric field E,the magnetic field B and the vectormagnetic potential

A

, are abstractmathematical concepts that makepractical presentations difficult. Thosefields, however, have everything to dowith why antennas radiate. Explana-tions for the qualitative and quanti-tative relationships between electric,magnetic and potential fields arepresented. All electromagnetic field equationsare interrelated. Each represents adifferent aspect of the same thing. In-deed, we may derive one from another!Here we shall examine the dc case, andthen move on to RF. First, we owe it toourselves to cover the fundamentals.We shall begin with E and B. As-sume that both fields are constant inmagnitude (uniform) and observable.In our first example, they are producedby a single charged particle moving ata constant velocity

v

in a vacuum.Next, changing magnetic and electricfields produced by time varying an-tenna RF currents are discussed. Inall cases, fields are produced both bystationary charge and charge in mo-tion. All charge we assume to be con-nected through fields. We observe thatthe fields change when charge is inmotion relative to an observer.It is very important to realize thatboth constant and time-varying fieldscause action at a distance. That is tosay, an electron’s field affects otherelectrons some distance away.What is a field? No one reallyknows. Field lines were visualized byMichael Faraday. The idea came fromthe orientation of iron filings that heobserved on top a piece of paper witha magnet (lodestone) placed beneathit. According to the Richard Feynman,

1

a field is a mathematical function weuse to avoid the idea of action at a dis-tance. We can state that a field con-necting charged particles causes themto interact because the field exerts aforce on charged particles.Does this mean that all matter inthe universe is connected throughfields? As one so elegantly put it:

All things by immortal power,

Near or far,

Hiddenly

To each other linked are,

That thou canst not stir a flower

Without troubling of a star....”

Jan/Feb 2005 39

From “Mistress of Vision” by En-glish poet Francis Thompson (1857-1907)

Charge in Motion Gives Rise to a

Magnetic Field

All charged particles produce anelectric field E. The E field can pointinward or outward, depending on thesign of the charge, and is infinite inextent. The field causes action at adistance. See Fig 1. Because the fieldhas both a magnitude and direction,we shall represent it as a vector. Thesame is true of a particle’s velocity, soit is also a vector. Vectors are hence-forth indicated by boldface letters.The only magnetic field associatedwith a stationary charged particle isits spin magnetic dipole moment, butwe shall ignore that for now. No othermagnetic field is produced by station-ary charged particles because the par-ticle and its

E

field are not in motionrelative to the observer.

When a particle with charge

q

moveswith velocity

v

, its

E

field changes: Itbecomes dynamic. The dynamic

E

fieldgives rise to a

B

field as seen by “JoeHam,” a stationary observer, in Fig 2.Joe observes that constant-magnitudeelectric and magnetic fields are presentsimultaneously. The magnitude of

B

depends upon the velocity

v

of the par-ticle. The magnetic field that Joeobserves is:

2

EvB

×=

2

1

c

(Eq 1)where c is the speed of light in m/s,

v

is charge velocity in m/s and

E

is theE field in V/m. Note that if

v

= 0 then

B

= 0 (no magnetic field). If the veloc-ity of the E field were c, then we wouldhave

E

=c

B

, which is what we havewith a freely propagating electromag-netic wave. The cross product, desig-nated by

X

, between the velocity vector

v

and the

E

vector indicatesthat particle velocity is perpendicularto both the E and B fields. The B fieldin Eq 1 is bound to the moving chargedparticle, as observed by Joe. The B fieldcomes from a relativistic transforma-tion of the E field involving the ratio

v

/c. Einstein introduced the world torelativity 100 years ago, in 1905!The

E

field multiplied by

v

/c² thattransforms to the become the B fieldin Eq 1 is:

2

cvˆ )sin( )( rq

/

=−−=

β θ β β πε

rE

232220

114

(Eq 2)where

ε

0

is the permitivity of a vacuum,

r

is the distance from theparticle’s line of travel,

θ

is the anglebetween the

E

field and the particle’sdirection of travel, and

r

ˆ

is a vectorof unit length pointing in the directionfrom the particle to the place where

E

is evaluated. The magnitude of

B

(seeNote 2) is found by substituting Eq 2into Eq 1:

2322220

114

/

)sin( sin )( cvrq B

θ β θ β πε

−−=

(Eq 3)Therefore, for a moving chargedparticle in free space with no externalinfluences, a portion of its E field getstransformed, becomes dynamic, andappears as the B field multiplied bythe coefficient v/c². This means that if particle velocity increases, the magni-tude of the B field increases propor-tionally. An observer sees both the Bfield and the E field at the same timewith charge in motion. Note that the

B

vector is perpendicular to the

E

vec-tor, to the velocity vector v. Einsteinrecognized this very fact by thinkingabout one of Maxwell’s equations andthe result was relativity. The pictureassumes that

θ

in Eq 3 is 90°.To summarize, the magnitude of theresulting magnetic field depends uponthe velocity of the charge and theamount of charge. This means that theB field really is the relativisticallytransformed E field! The two fieldsmust always change together and theydo. If the B field source is the E field,can there be such a thing as a B fieldby itself without an E field? The an-swer may surprise you.

Charge Moving at Constant

Velocity Produces a Constant

Magnetic Field

Charge flow in a wire is obviously very much different from that of anisolated charge moving in free space.However, exactly what is different andwhy? Assume that there are no un-matched charges in our wire. Thatmeans the number of electrons (nega-hed c

Fig 2—The B and E fields of an electron moving with a constant velocity

v

.

Fig 3—Magnetic field lines around a wire with dc current.

Fig 1—An isolated charged particle with

its E-field lines. E fields cause action at a

distance. The effect is very small for small

charges a great distance apart.

40 Jan/Feb 2005

tive charges) equals the number of protons (positive charges) in the wire.Experiments have confirmed that cur-rent flow in such a wire produces onlyan observable magnetic B field at adistance

r

from the wire. The electrons’radial E fields are not observed at all.If Joe Ham were the observer, hewould say there were no E fieldspresent. The drift velocity of the mov-ing charge is constant, so the B fieldsis just a function of the number of charges in motion and their velocity.See Fig 3.Constant DC current in an infi-nitely long wire produces a uniformmagnetic field along the wire’s entireaxial length. The magnitude of

B

ev-erywhere along the axis at a distance

r

from the wire is the same. The mag-netic field lines extend to great dis-tances and they always close onthemselves, unlike electric field lines,which always terminate at a charge.Joe might ask the question, “If theB field came from the E field, then howis there now only a B field observable?”Perhaps we should state that there isno

net

E field present. Now I wouldlike to propose another question: If there is no net E field present, is therean E-field energy density present?

The Wire’s Unobserved Net E Field

Let’s say that there is a moving ra-dial E field that accompanies chargeflow in a wire and it cannot be ob-served for some reason. We know thatprotons are fixed in the wire’s metal-lic lattice and that the electrons arethe charges that actually move. Theproton static radial E field is uniformand produces no B field because thereis no relative motion with respect tothe observer. The electrostatic E fieldfrom all the wire’s protons and elec-trons cancel and we are just left withthe B field. The E fields of the protonsand electrons, of the same magnitudebut of opposite directions, must be zeroor very close to it for Joe not to ob-serve them.Is there a basis for this in physics?The answer is yes. The big questionis: How do we know this? The energydensity of each field must be present.If this were not the case, then the lawof superposition would be violated.

Superposition of Like Field

Quantities

We know from the principle of fieldsuperposition that the net vector fieldproduced by two separate vector fieldsof the same kind (either E or B), atthe same time and place, is the vectorsum of those individual fields. Thatmay help additionally to explain whyno net E field is observed when cur-rent flows in a wire. The E field from afixed proton must exactly cancel thenonrelativistically transformed E fieldfrom the moving electrons.The superposition principle alsoworks for magnetic fields according toFeynman (See Note 2). That is to say,individual small amounts of magneticfield strength

d

B combine to producethe macroscopic strength that we ob-serve. Is it possible to test for energydensity where the net E field is zeroand is not observable?

Gauss’s Law

Let’s use superposition and the con-cept of the closed surface to prove whatwe discussed earlier and what we sawin Fig 4. To do it, we extend the dis-cussion to another term called flux.Flux is the electric field strength as-sociated with each unit area througha surface. Feynman

3

relates flux andcharge in the following way: “The E-field flux

Φ

through any closed surfaceis equal to the net charge inside thatsurface divided by the permittivity of free space.” That is Gauss’s law. Theclosed surface could be a sphere con-taining some charge, or any othershape so long as it fully encloses thecharge. In mathematical terms:

0

ε Φ

totaltotal

q

=

(Eq 4)Thus the net E-field flux passingthrough the surface enclosing our wire,which has a net charge of zero, is zero.It is interesting to note that if weused superposition to determine en-ergy density (energy/volume) by sum-ming the E-field energies of protonsand electrons, the energy densities donot cancel, they add! The reason is thatin the calculation of energy density, theE-field is squared. This yields two posi-tive numbers so there can be no netdensity cancellation. The proof thatfield energy density is present is be-yond the scope of this paper. I suspectthat it has something to do with grav-ity. Didn’t Einstein show us that mat-ter and fields were both essentiallyforms of energy?

The Biot-Savart Law

It turns out that there is a law thatmay be derived from Eq 3 that sup-ports the idea that all B fields arisefrom changing E fields. Everythingabout it is self-consistent. It is calledthe Biot-Savart law.The Biot-Savart law allows deter-mination of B a distance r away froma wire for a given dc current. To per-form an actual calculation, one wouldbreak the wire into infinitesimallysmall segments

dl

and integrate alongthe entire length of the wire all theinfinitesimally small

dB

contributionsthey produced at some distance

r

fromthe wire. The sum would be the mag-netic field strength B at that distance.To derive this law, we begin withEq 3 and assume that the magnitudeof B is changing with time. In the no-tation of calculus, the rate of changewould be

dB/dt

, where

t

is time. Tak-ing the time derivative of the magneticfield, we get:

−−=

2322220

114

/

)sin( sin )( cvrqdtddtdB

θ β θ β πε

(Eq 5)

Fig 4—Electron motion in a wire and particle E fields.