Waveguide "cookbook"

[logbook]

Has anyone seen a sort of waveguide cookbook for the tuning structures that you need to put into T-junctions and magic-Ts to make them work at all? It is evident from my HFSS simulations that untuned structures are no use at all at all frequencies in the pass-band. I am of course running parametric sweeps over the weekend to see what I can come up with from first principles, but some sort of starting point would help.

I am just using standard rectangular waveguide. You would think this was just a standard text book item. (I have the Waveguide Handbook from Marcuvitz, but disappointingly it doesn’t mention tuning a specific structure, it just seems to catalogue lumped impedances for lots of simple cases.)

 

[biff44]

Not likely to, either. Those are all trade secrets from days gone by. These were developed by machine shops that had the knack for fine and detailed machining. After some time, they would get an engineer or two and, mostly empirically, figure out how to improve bandwidth or performance. Sometimes they would take an enormous model of a waveguide, made out of plastic and metal foil, and hack away at it--testing a 1 Ghz or so, until the shapes were right. When it worked, they would then frequency scale it up to 40 GHz or whatever, and do some final tweaks.

 

[logbook]

Thanks Biff. This is what I had heard too, but it was just hearsay.

I have Googled and found a few papers by academics about T-splitters that don't need tuning, but most are IEEE explore abstracts only.

I also found a small piece in "Principles of Microwave Circuits" by Mongomery, Dicke and Purcell, which describes a simple tuning structure for a magic-T; I will have to try this in HFSS to see how good it is.

 

[biff44]

If I were you, I would get onto ebay and buy some Cband magic tees, saw them open on a bandsaw, and use those dimensions as a starting point. I am sure that with modern Emag analysis techniques, one could improve on these 1960's designs.

 

[logbook]

Interesting idea. Thanks.