waveguide power combining, is it possible?

[PG4I]

I am looking into combining two 1 kW waveguide magnetrons so the total output power is 2 kW.

I have tested a WR340 waveguide splitter to see if I can use it as a combiner. A test with the network analyzer shows good performance as a splitter but a 6 dB loss as a combiner. Coupling to the waveguide is done with coax to waveguide transitions. VSWR at the input ports is 2.6 which is already a 20% loss.

Next, a WR284 magic Tee was tested. Great port isolation, but the power at the sum (sigma) port is 3dB below the expected level.

Is it at all possible to do waveguide power combining?

Thanks

 

[biff44]

a magic tee is a VOLTAGE vector combination device. in other words, if you have a signal at port 1, and another signal at port 2 that is the SAME AMPLITUDE AND PHASE....then almost all the power ends up at the SUM output port, and little of the power ends up at the ISOLATION port.

but you said you are using magnetrons? Are the two magnetron outputs at exactly the same phase? to get them to be the same phase, they have to be PHASE LOCKED somehow. If they are just free running, then they are non coherent, and will combined very differently.

 

[PG4I]

Yeah I know what you mean... I am trying to have it working at low power and work from there. One signal source is a network analyzer which i have set to a 1 Hz span. The other is a signal generator which has a 1 Hz resolution. I guess even this setup won't work, unless you are lucky (because phase dependency is critical)?

Would there be any solution available when the 2 signal source differ in frequency?

 

[biff44]

if the two signal sources are different frequencies, they will not be phase coherent and should incur a 3 dB combining loss. I keep meaning to do some optical wg combining experiments to see if there IS some way to spatially combine two incoherent signals with 0 dB loss....but I never get around to it. In my mind there SHOULD be a way of doing it, but so far everything I try is 3dB loss.

 

[PG4I]

That's what I am seeing with the Magic Tee: a 3 dB loss when I move away from the coherent frequency! Thanks so far for your comments.

Regards,
Joop

 

[vfone]

Here is what they say about this task 70 years ago:

https://dspace.mit.edu/bitstream/id/5113/RLE-TR-014-14261743.pdf/