According to the textbook I have, WC329 at 8.36cm diameter is recommended for 2.42GHz to 3.31GHz using the TE11 mode.

Circular waveguide can be a problem compared to rectangular.

What type of coatings? At that low frequency, you could probably coat it with many things and never notice the effect.

kch.

Depending on the power loss requirements or Q-factor, silver or gold plating is sometime required, good surface finish is also important.

Hello
Thank you for your replies. Which text book you refer logbook. I want to heat a reactor using microwaves.
Regards

Microwave Engineering by David M. Pozar will have what you're looking for.

Hi
Thank you bearing01 will buy one right now. The magnetron is fixed to a rectangular waveguide in a specific geometry to avoid back coupling i guess. How should the magnetron be fitted in a circular waveguide to avoid back coupling.
Regards

I'm puzzled about the low resistance coating for this application. I thought the skin effect was only a problem on wires. Cavities do not suffer from the skin effect, if anything the current would be pushed to the outside of the cavity conductor.

Is this right, or am I missing something?

Mart

Extassi,
    sorry for the delay in answering. I missed your question.
The book referred to was "Microwave Circuit Analysis and Amplifier Design" by Samuel Liao (1987).

GraviMan,
    cavities do suffer from the skin effect. The waves are inside the cavity/guide and are constrained by the guide walls. Currents flow in the guide walls and therefore dissipate heat. I seem to recall cavities being plated with silver to increase the conductivity in the inner surface layer, increasing the Q of the cavity. Note it is the inner surface which is in contact with the wave. The skin depth is so small that the wave does not penetrate the conductor and get to the outer surface.

Extassi, it is called a rectangular to circular waveguide transition. These are something that are very mode dependent, of which I doubt that you have a strong grasp on of, as yet.

"cavities do suffer from the skin effect. The waves are inside the cavity/guide and are constrained by the guide walls...The skin depth is so small that the wave does not penetrate the conductor and get to the outer surface."

Thanks for helping my mechano brain, Logbook! If there was an identical wave approaching, say a flat sheet of copper from the other side, would the skin effect still apply? Basically the em field would no longer have a gradient across the sheet.

For the cavity I was thinking about it being circular cross section. I take it that if it was actually carrying voltage a signal, that current would be forced to the outside? Is there a case for a weak signal in the waveguide, matching at 180', the em field being carried? This would also avoid any radiation, as well as expensive metals!

Mart

The best coating for waveguides is silver plate, with a rhodium flash on top (to keep the underlying silver from oxidizing).  The silver should be thick (200 microinches or more), and the rhodium should be thin (this is from memory, but 10-30 microinches maybe).

"The silver should be thick (200 microinches or more)..."

That's about 5um in metrispeak. What's the best way to get this done? Any idea how much this would cost, say per cm^2 (or inch^2)? What sort of field strength can this handle (V/m or teslas)?

Mart

I would imagine the coating is good for a couple tens of killowatts cw in a standard waveguide.  What you have to address is that, in a resonator, there can be very large standing waves.  Wherever the voltage standing wave is a minimum, the current will be at a maximum.  You can figure out how much resistance is in the surface layer, and how much heat that will produce, and decide if that is acceptable.  

The depth of penetration into the silver coating will be:
D [meters] = 0.0642 / (f)^1/2  = 1.3 micron or 51 microinches, hence the 200 microinch recommendation.

Conductivity of silver is 6.17X10^7 mho/meter.

"Resistance of a conductor with exponential decrease in current density is exactly the same as though current were uniformly distributed over a depth D"

If it is helpful, the surface resistivity in ohms per square is Rs = 2.52 X 10^-7 * (f)^1/2

OF COURSE, if there are any discontinuities in your resonator, like where a lid meets the cavity body, you had better be exciting a resonator mode that does not put a current maximum there.

The best way to get it done is the standard way: take your machined resonator to a good plating house.  I think you can directly plate silver over copper, but ask them.  Also, becarefull of plating berrilium or other alloys of copper as sometimes the cleaning solvents will attack the copper after plating.  Cost is pretty reasonable, as I recall.  The cost of the silver is negligible.

Beware if the plating wants to first plate with some lossy metal, like nickel!

Thanks Biff44, for some really practical advise. I would imagine that there is not much point making the silver coating any deeper than the skin effect depth.

Mart

Well, you need it several skin depths deep, as a rule of thumb, because the electric fields decay exponentially.  At one skin depth, you still have a fair portion of the field left.  After 4 skin depths the fields are mostly gone.

How does this change if the field is coming from both sides at the same time (ie constant electric field strength)? If a circular waveguide had a voltage directly injected into it would this help with the skin effect?

The concept is that the voltage would induce a current in the outside circumference, while the em wave induced the current in the inside. By chosing the signal voltage the wave would still be totally reflected along the wave (ie no leakage), but the entire thickness of (say) copper would pass the induced current. This would minimise resistance losses...

Mart

I am having a hard time picturing what you mean.  If by circular waveguide you mean a long "pipe", and you want energy on the inside of the pipe, and energy on the outside of the pipe, THEN if the pipe walls are more than around 8+ skin depths thick there will be NO interaction of the fields whatsoever.  So energy added on the outside of the pipe will not supplant energy lost on the inside of the pipe.  

There are "active resonators" being studied for improved Q.  They involve an active semiconductor device making up for resonator losses.  I am not sure how practical they are, but there are some papers around.  There is a term "vacuum port" used sometimes in RF space communication where you can use reactive energy to improve a receiver's noise figure over what should normally be considered to be the theoretical limit.  

Give me a better idea what you are interested in doing, maybe I can help.

Yeah, wacky stuff like this:
"The main goal of our research is the achievement of substantial squeezing of the vacuum fluctuations associated with a soliton. If the soliton is injected into the probe port of a Mach-Zehnder interferometer and the squeezed vacuum is injected into the vacuum port, sub-shot noise measurements of the phase difference between the two arms of the interferometer can be performed."

http://www.stormingmedia.us/10/1034/A103473.html

So if you want to squeeze some solitons, don't squeeze too hard?  This stuff makes my brain ache.  Solitons, BTW, are waves that never decay because they keep reforming themselves.  Sounds a little like what you are trying to do?  Solitons were discovered in nature, from satellite pictures of the oceans.  People realized that every once in a while, a wave forms that goes most of the way around the world.  Cool huh?  You can make a broadband microwave frequency multiplier with solitons, but that is a different story.

Wowsers! Shameful to admit my brayne now aches too. I have read about the use of Schrodinger wave eqn to predict higher than normal ocean solitons. But, this is not quite what I was getting at...

All I am interested in is using the full copper thickness of the waveguide. If the waveguide is circular section it could be thought of as a large hollow wire. If a voltage signal is passed through the waveguide/wire then the skin effect says the current will be concentrated at the outside. If an em wave is passed inside the waveguide/wire then the current is concentrated at the inside. If the two are 180' out of phase, and have the correct levels, the two effects should cancel. The practical upshot is that the current will be evenly distributed through the entire wire/waveguide thickness. Is this wrong?

This is also the crux of my query about how the back-emf current is distributed in a plain copper sheet, subject to the same em wave from both sides simultaneously. Again since there is no actual voltage field across the copper, I imagine that the skin effect is cancelled so the current flows evenly across the entire depth. Ok in this case ther will be a practical upper thickness limit, but i'm really attempting to understand the skin effect mechanism for a flat sheet.

I will be studying Maxwells eqn's (vect calc - how hard can it be) with the Open university. In the interim i am really trying to gain an intuitive understanding of the cause of the skin effect...

Mart

Read up on skin effect.  I think you are confusing low frequency electronics with microwaves.  They are quite different.  A circular waveguide is in no way like a "wire".  

Sorry Biff44, perhaps I'm just being thick here...

I spent 2 hours on the web looking for an explanation of why a plain sheet of copper should suffer the skin effect. Nothing. There were plenty of examples explaining how the induced current in a (solid) conductor forces the main current to the surface. I also found a very neat applet which i thought i would share:

http://www.falstad.com/emwave2/directions.html

I really just don't understand why a plane em-wave should induce secondary currents that force the main current to the side from which the em-wave comes. In free space i can understand that the spacial divergence would cause the field to be stronger on the wavefront side, hence skin effect. I just don't see why it affects waveguides, when the wavefront is planar - isn't it?

Mart

Look into a mirror. There is a reflection. This is due to opposite currents, induced into the metal, that re-reflect the EM field right back at you. Well,  however small a time these currents exist in the metal, most stick near the surface, yet the finite Er (permittivity) of the metal requires maxwellian-wise to drop off with an exponentially decaying field strength. Find a perfect metal based mirror, then we will talk., Solitons for all dude!

Thanks GOTWW. So it's an exponential decay of the opposing currents. Just took me a while to gain intuitive understanding...

Actually there are some nice demos in that applet, that nicely demonstrate the skin effect in lossy conductors.

Thanks all for your patience. I have only limited practical experience in this particular field to fall back on. Can't wait to get started on Maxwell's equations!

Mart