waveguide attenuator
is there somebody, who know, how can I design freguency independent waveguide attenuator ? I know, how can I design freguency dependent wav. att. only :) - its thin resitive film on glass, but independent its not soi easy. Have you somebody theory ?
Thank
To be honest nothing is frequency independent. Your waveguide has a bandpass characteristic between cutoff and the propagation of secondary modes. Therefore I assume you would want to absorb all this band?
It is for waveguide R120, 9.84-15 GHz (only first mode), and I need S21=-30 dB in all band and good S11 (-15 - -20 dB).
I know, that it is possible, but i dont know, where is the "trick" - thickness film ? Sq. Resistance ? .....
Thank
Interesting question. One idea might be to load the waveguide completely with absorber. This causes little reflection when done right and absorbs all inflecting power.
I have no idea about the practicality of this design.
Load the waveguide completely with absorber - i think, that is not good idea, because the absorption depends on how many wavelength pass through abs. material - so higher freguency = higher absorption.
But commercial attenuator have glass plate with resistance thin film of material - it is without freq. dependen... and is it possible tune it (shift in waveguide)
I would be tempted to transition into something like finline, where the fields are contained more to one location. Then I would design some sort of structure that had less resistance in the card at higher frequencies. Speaking in lumped element talk, if you had a shunt loss, and the distributed resitance was in series with a small inductance, then as you go higher in frequency (more wavelengths in the resistive card), the resistance would look higher due to the inductive part. You would, of course, have to use an EM simulator to get it right (or a lot of cut and try in the lab).
The "inductor" would be just a necking down of the resitive material before it touched the waveguide top/bottom edge. The necking down would look inductive.