Help needed to design ~1Ghz high power Oscillator. 2

[ElectLect]

I am new to this forum, but was suggested it for my question.

I am trying to design an oscillator capable of generating ~1Ghz fixed frequency signal. It needs to be a very high current capability, but will not be modulated with any signal (ie fixed pitch and amplitude). It doesn't even need to be a perfect sinewave, as long as it has only odd harmonics.

Don't worry about RF contamination, the circuit will drive a completely enclosed system, so no RF will be emitted. Most likely the impedance will also be very low. I am doing this as part of a Post Grad research project into materials property testing, so am not sure who to ask.

Like everything in the (underfunded) research establishment, it has to be developed on a shoestring. Any of you more experianced electronics guys know how I should go about it?

Graham

 

[ElectLect]

Thanks Comcokid. This forum really is every bit as usefull as i had heard! Changed my approach to the problem completely!

Howabout freeware for EM simulation? Are there any limited versions i can get hold of for some initial design studies? If i am going to microwaves I will need to (potentially radically) rethink my initial design. May very well work better, but will take more doing.

 

[BrianR]

High power RF is most efficiently generated with a valve operating hard into class-C. In fact a big triode will oscillate so happily you'll be hard pressed to stop it. An output network which uses a tuned line made of silver plated copper tube, grounded at one end and with the low Z connection being a tapping point close to the ground should be quite efficient.

Reducing the losses in the impedance matching network to feed your low impedance load will be helped by lowering skin effect through using the most conductive metal available such as hard silver plating. Copper plating is OK but not as good however both metals suffer from a rise in skin resistance if surface corrodosion develops, particularly copper as its oxide is a semiconductor. Some kind of surface stabilisation might help but few products are suitable for high power 1GHz use, possibly a thin PTFE coating but there are probably laquers available as well but I don't know which. Use only ceramic insulators and air spaced capacitors to minimise your losses.

The military disposals market does have some sources of fairly powerful 1GHz RF that while not suitable for your final product might be a useful development aid.
Eg, an AN/APX-76 IFF Interrogator generates between 2-8KW pulse power at either 970 or 1030 MHz, I cannot remember which. Similarly ex military TACAN beacons use Klystrons producing many KW of pulse with a high duty cycle at approx 1GHz. The klystron could be reconfigured as a CW amplifier if you need CW.

 

[ElectLect]

Interesting BrianR, some very useful pointers...

Any books I should read to get me up to speed on this sort of design, using valves in particular? Do you have any links, for sites dealing in military disposals market?

 

[BrianR]

Very useful design and construction data is found in the ARRL Handbook. Also (buy as well as - ) their specialty handbooks,

http://www.arrl.org/catalog/?category=VHF/UHF/Microwave&words=

EIMAC Industrial RF heating tubes at

http://www.cpii.com/eimac/index.html

Application note "Care & Feeding of Power Grid Tubes"

http://www.qsl.net/kk5dr/Tubes.html

1296MHz 500W amplifier

http://www.qsl.net/dl4mea/23gi7.htm

 

[ElectLect]

Some really useful info BrianR - Thanks.

The overiding impression I am getting is that I was way overoptomistic to assume I could use transisors, even GaAs MOSFETS. Howabout designing a fixed frequency power resonator? Is there a better method than simply using feedback though a Power amp designed for signals? I'm not overly familiar with valves, so don't know if there is any advantage in running them off --> saturated --> off...
I appreciate that this is what magnetrons really do anyway.

 

[BrianR]

You could use MOSFETS or nearly any other type of transistor to do this job, with enormously varying degerees of difficulty. However the desing is not something a competant engineer could do even with a good understanding of the maths, as there is a lot of practical stuff that only comes with experience. The effects of stray reactance, changes in material properties vs frequency and power, heating effects and the behavior of RF currents on surfaces must be well understood also.

You do not need feedback around an amplifier, (oscillating amplifiers are very unstable with regard to frequency), you could drive an amplifier with a low power oscillator and have fine frequency and power control.

Valves work very efficiently in Class-C transferring huge gulps of power to the resonant load and relying on the flywheel effect of the load to sustain the waveform.

Yours is an industrial heating task which places varying low impedance loads on the RF power source and this further complicaes the design. All in all I would suggest that you should see what you can buy for this job. It will be less costly in the end.

 

[ElectLect]

Thanks BrianR, you really are helping me to think this thing through...

 

[jimvandamme]

A few years ago I bought a 500 watt average power solid state S band power amplifier. It cost $800,000, was 30% efficient, and weighs 120 pounds. I sat it next to our microwave oven which put out 600 watts average, 40% efficient, and cost $90. (I also developed tube transmitters.) Of course the magnetron is not a coherent amplifier, but you don't care in your application.

Another trick we used to do is a resonant ring, where you add power into a loop of waveguide with a circulator, and it just builds up to infinity (minus the losses) instead of dumping into a load at the end of your test cell. It was fun at megawatt levels. This could reduce your power requirements 100X.

BTW IFF systems interrogate at 1030 MHz and respond at 1090 MHz.

Final advice: forget Maxwell, get the ARRL handbook on microwaves.

 

[SolderingGunslinger]

I guess I would go the Maggie route, however if you wish to eschew the lowly magnetron, consider a reflex klystron.

The advantage of the magnetron is that it is a high powered oscillator and only needs a DC supply for the B+ and an AC filament supply. Simplicity of design.

A reflex klystron is also a high power oscillator, but the power supply requirements aren’t as simple as for the magnetron.

Other systems, like IOTs or TWTs require more and more complex external support electronics.

Yes, it all could be done with solid state; however at frequencies over .75 GHz, the efficiencies of high power hollow-state devices begins to exceed those of their solid state brethren.

High power UHF TV transmitters are usually klystron amplifiers driven by solid state exciters.

I remain,

The Old Soldering Gunslinger

 

[Higgler]

Electlect (Graham),
I've used a microwave oven before as a power source. I didn't need the full power so I placed an antenna inside the oven to couple off power at various levels. If you need full power, you would attach a waveguide to the inside of the oven and easily bring out the energy.
The opening of the waveguide is 2.84" WIDE X 1.34" tall for WR284, although you may need a slightly larger size like WR430 (4.3" wide, 2.15" tall) to better match the opening in the oven where the energy exits.

You'd probably get over one kW our your waveguide. Trying to operate it in an "enclosed" area may reflect too much power into the oven and damage it. You must be concerned with that reflection.
kchiggins