High power terminations

[molloy]

I should design a wilkinson combiner to combine 3 amplifiers (1.5KW output) in UHF band (up to 860MHz) and the problems are the high power terminations (strip or microstrip technology) i have to design. Anyone has papers or advices?

thanks

molloy

 

[biff44]

They make chip resistors that are printed onto BeO substrates. You either solder or bolt them down to a metal heat conductor. Watch the Efield arcing.

 

[molloy]

Terminations available are up to 250W (500W have not good S11) and anyway i have to manage even the failure of the amplifier. So the terminations should dissipate higher power.

 

[sergio mariotti]

I think the way may be to increase heat exchange area. I.e. if you place a power 3 dB attenuator, followed by a power load, you'll divide by 2 the dissipated power of each component. In this manner, because the area is wider, it's easier dissipate heat.

 

[biff44]

Well, you do have your work cut out for you. It does look like resistors above 250 W are only suggested for use to 500 MHz.

There is some design work to be done, based on balancing the power dissipated in resistors to not exceed the 250 W rating.

I would first off try to use 2 or maybe 3 resistors in series (if I could get away with the capacitance to ground), in every location. Right off the bat that would get me either 500W or 750W power handling. Of course, you need to mount the resistors to a high thermal conductor (copper bar, etc).

I would then think about a multisection wilkinson combiner. I would fiddle around with the transmission line impedances and the resistance values to get the power evenly balanced into the resistors when there is a fault on the output or input. By balanced, I mean getting the resistor closest to the output to handle 1/2 of the fault power, and the resistor in the middle of the combiner to handle the other 1/2.

For a 3way combiner, how are you going to configure the resistors, in a "Y"?

I would also carefully consider the worst case operating conditions. In the event of a fault, do the power amplifiers shut down automatically? If so, then the power resistors can handle more power than they are rated for for the brief period of time the amplifiers are running.

Also, of course, you realize that there is no power disipated in the resistors unless there is a bad VSWR somewhere. Normally there will be little heat in the resistors, contributing to their life.

 

[molloy]

We have the technology to dissipate so high power, so that is no problem.
Anyway it is possible to use balun in such a manner that the power i have to dissipate in 1 resistor will be dissipate in 2 terminations for instance. Plus this is a way to use 50ohm termination for every resistor (that in a wilkinson are different from 50ohm).
Anyone has papers or advice about this?
thanks

molloy