The Electro Magnetic Wave...

Looking up J pole antennas for a previous post I came across numerous pictures/representations of the EM wave, as you can see below the M(H) and E fields are in phase. I would have thought that they should be 90 degrees out of phase, i.e. when the Mag is changing most the volts should be maximum etc. it would also give it a direction.

I am probably wrong but, discuss...

[Radio Wrangler]

Instinctively, it seems right that they are in phase. This means that at any point and any time, the impedance of the wave is real, and therefore real power is flowing.

David

[Skywave]

The E-M wave is propagated in 'free space', the intrinsic impedance of which is 377 ohms. Therefore, as with any alternating electrical current flowing in a purely resistive medium, the voltage and current will be in phase. In the E-M wave, the electric field is analogous to the current; the magnetic field to the voltage (or is that vice-versa?) Whatever, for either, the E and the M fields will thus be in phase.
Anyway, FWIW, that's my way of looking at this.

Al.

[G4_Pete]

The phase of the Electric and magnetic fields are commonly published as phase shifted in a circularly polarised wave.

I don't profess to know the maths theory but did study it when building the RadComm "Chopstick Antenna" 70 cms helical antenna. Never worked - could not get it to match despite adjusting the coils and helix length etc. Too many variables and not enough experience at the time.

Pete

ref:C. Richards, “The 10 Turn Chopstick Helical (Mk2) for
OSCAR 10 432 MHz Uplink,” Radio Communication,
Oct 1984, pp 844-845.

[GMB]

In the E-M wave, the electric field is analogous to the current; the magnetic field to the voltage (or is that vice-versa?)

Vice-versa it is. The electric field is voltage and the magnetic derives from current.

[ms660]

That picture looks ok to me, in time phase and space quadrature.

Lawrence.

[G4_Pete]

Is my previous post statement about the E and M waves being phase shifted incorrect or is the circular polarisation wave containing two E fields shifted by 90 Deg but each having an accompanying M field in phase as per the opening diagram?

I am trying to get an understanding of this as I am on 2nd year Open University physics and electromagnetics and Maxwell is coming up next year and it always helps to have a running start. Internet descriptions are not helping it seems!

Pete

time phase and space quadrature

That was the bit I couldn't see, thank you.

[Argus25]

I think the way to understand it is that a transverse EM wave can only have two states of polarization, at 90 degrees to each other and they are at 90 degrees to the propagation. One is occupied by the electric field and the other by the magnetic field, you can regard them as separate. There are no free charge carriers in the medium of space like there are in circuits.

(In circuits where there are charge carriers, the electric field and magnetic field energy of capacitors & inductors interact via current and we are used to them exchanging energy with each other and the usual phase relations of leading and lagging scenarios, but this doesn't apply to an EM wave which is why the magnetic field doesn't have a lead or lag with respect to the electric field)

With the EM wave the electric and magnetic fields are in phase as your diagram shows.

The waves corresponding to E and H are independent and have their own energy.The energy density (energy per unit volume) of the EM wave is therefore partly magnetic and partly electric.

Since the magnetic wave energy and the electric wave energy of the EM wave, are independent (not exchanging energy with each other) the fact they peak at the same time makes sense. On the other hand, in circuits where electric and magnetic energy get exchanged between magnetic fields in inductors and electric fields in capacitors, in the course of some cycle, the voltage and the current (or magnetic field) can never peak at the same time, hence the phase differences as it takes time to transfer energy between the electric and magnetic field.

It is interesting that physical forces at 90 degrees to each other follow the same rule of independence, much like the magnetic force and electric force of the EM wave.

[Radio Wrangler]

Maxwell's equations can describe a magnetic wave travelling along through space and how it induces an electric wave polarised in a plane at right angles to its own plane.

Maxwell's equations can equally well describe an electric wave travelling along through space and how it induces a magnetic wave polarised in a plane at right angles to its own plane.

Both these things happen simultaneously and interact. If the magnetic wave came first, it reduces in amplitude as it induces a magnetic wave accompanying it. The created wave also acts to induce the electric wave, and at a good distance (often takes as >20 wavelengths from the source) the two waves have settled into equilibrium. Each losing some of its own amplitude by inducing into the other and gaining back what it lost by the induction in the reverse direction.

The equilibrium relationship for a pair of mature waves has an electric field of about 377 Volts per Metre for every Ampere per Metre of the magnetic wave.

If we divide the electric field strength by the magnetic field strength to get the ratio of electric to magnetic strengths we get (377 V/m)/(1 A/m)

We can cancel the 'per metre's and get 377 V/A

Now volts per amp is Ohms! 377 Ohms!

Someone called this 'the impedance of free space' which sort of loses the basis of its meaning and has students wondering where to stick the AVO prods to measure it ever since.

Now what if we multiply the two field strengths together?

(377 V/m)*(1 A/m) = 377 V*A/(metre squared)

V*A is Watts

So we have 377 Watts per metre squared

That's the power density in a mature EM wave with 1A/m ad 377 V/m strengths.

But just a minute, these things are at right angles. So we should use vector multiplication. Do that and we still get the 377 Watts/metre squared but the vector product is at right angles to both the electric and magnetic vectors.... Yes! it points along the direction of motion of the waves!

This is brilliant, it gives the power density of the EM signal and even points in the direction it's moving.

It all fits together

Maxwell done good!

David

(The multiplication trick was first done by a chap with the staggeringly appropriate name of John Poynting and Oliver Heaviside discovered it independently so it has to be right )

[Skywave]

Above, David wrote: "Someone called this 'the impedance of free space' which sort of loses the basis of its meaning and has students wondering where to stick the AVO prods to measure it ever since".

Brilliant! Find a chunk of the luminiferous ether (usually simply called the 'ether') and prod away!
No, but seriously, I found your write up succinct - and I also appreciated your resistance to avoid all the heavy maths that is usually appended to this topic. Overall, it took me back to my college days in the 1970s when I was studying communications engineering.

Al.

I always struggled with understanding Maxwells equations, I do like to visualise something, never quite got them to pop out into a picture. I have used them but that is a different thing. It is all much clearer now, 40 years too late (never too late to learn) and I have a picture in my bonce, I think I will have a play with them later and see if I can get the speed of light with understanding and a picture.

[G4_Pete]

From Davids description

"Yes! it points along the direction of motion of the waves!"

Ahh, Thanks a practical example of the vector cross product, not the normal
non intuitive vector pointing out of the paper!

Pete

[G8HQP Dave]

In circular polarisation the E and H fields are still in phase (in time) and perpendicular (in space), but in addition the space polarisation rotates by 360 degrees per wavelength. I like to think of an axial-mode helical antenna as acting a bit like the rifling inside a gun barrel: it both launches the wave and gives it a twist as it goes.

[Argus25]

Originally Posted by Radio Wrangler

The created wave also acts to induce the electric wave, and at a good distance (often takes as >20 wavelengths from the source) the two waves have settled into equilibrium.

This came up recently on another thread about near field vs far field signal strength for pantry TX's. It is one of the reasons why in the near field the magnetic loop TX antenna works so well because of the high level of magnetic field energy in the wave, in the near field, and this gives a very good result with a ferrite rod receiving antenna.

[Radio Wrangler]

Perhaps some people just have a very large pantry...

Most antennae are principally magnetic field radiators, (being long and thin) and the E-field is left to its own devices to grow with distance. And ferrite rod antennae even have an active magnetic core, so they are definitely inclined to favour H-fields.

Some time ago someone had the idea of creating ready-made mature EM waves by having radiators for both the H field and the E field, oriented and phased appropriately. I never got them to say why they thought this would have any advantage over radiating either field alone and letting nature take its course. People seemed to assume that 'it just would'.

David

[G0HZU_JMR]

I think I will have a play with them later and see if I can get the speed of light with understanding and a picture.

If you are interested in stuff like this I'd also recommend making a suite of tiny E and H field probes and do experiments with them on various structures including a dipole antenna, air core solenoids and toroids, transmission lines of various type and also things like stub filters. Lots of experiments! I'd strongly recommend making the probes from 50R semi rigid cable and make them a similar overall length. Ideally, you want lots of rejection of E with the H field probe and vice versa.

If you have access to a fast scope with 50R inputs (eg Tek 2465 or a fast DSO) and a basic UHF sig gen and a decent RF splitter and don't want muddy shoes you can do a few dipole tests indoors up at around 500MHz to 1GHz at low power across a workbench and use a 'short' dipole and use very small E and H probes. Connect the scope external trigger to one arm of the splitter and connect an H field probe to channel 1 and the E field probe to channel 2. Connect the dipole to the other arm of the sig gen splitter. I'd recommend triggering the scope constantly from channel 3 or the external trigger when doing stuff like this. Otherwise the triggering could get frustratingly unreliable as you move the probes. Don't expect perfect results but it should be close enough for study purposes. Swap the probe cables around as a check to make sure they really are phase matched correctly. Also, make sure all scope inputs are set to 50R and you might need to add extra attenuation/isolation in the external trigger path to the scope.

You should be able to test the discrimination performance of your E and H field probes when right up close to the dipole and you can also rotate the H probe to see a 180degree phase shift if it is responding correctly to the H field.

You have to hold the probes in the correct plane to see the best results and try and avoid the 180deg phase flip orientation of the H field probe. I'd also recommend placing ferrite EMC clamps on the probe shafts to limit effects from your hands. But as long as you can get the probes a metre or two from the dipole you should be close enough to far field up at 1GHz. Get up close and you can sniff along the antenna elements and look at the relative phases and amplitude responses on the scope in the near field using the E and H probes. You should see excellent discrimination between E and H field with the relevant probe as you explore the dipole elements and look for peaks and nulls.

If the cables to the scope are phase matched (same length) you should be able to get meaningful phase information in the far field and also when right up close to the antenna elements. With the scope triggered from the sig gen splitter you should also be able to see the phase change on each probe as you move it away from the antenna when in the far field region. You could work out the speed of the wave this way by measuring the distance D you have to move the probe to see a full wavelength change between the probe waveform and the trigger waveform referenced from the sig gen. Then compute speed = Frequency * D. Hopefully it will work out to be the speed of light!

You could repeat the speed test using a different dielectric, eg probe a long piece of terminated 50R microstrip on a big A4 sized PCB and the speed will be much slower. For FR4 material the velocity factor is usually about 0.54. So the E or H field probe will only have to move about half as far to see a wavelength shift on the scope wrt the sig gen trigger waveform because the wave is travelling much slower along the PCB microstrip line. Note that microstrip propagates with a 'quasi' TEM mode rather than pure TEM mode but I don't think this matters much here. I use a microstrip based system like this to calibrate and prove my E and H field probes and also to look for resonances in the probes.