EEVblog Electronics Community Forum

Electronics => RF, Microwave, Ham Radio => Topic started by: CopperCone on September 05, 2017, 09:44:34 pm

Title: what decides of a microwave band pass filter modes at passband multiples?
Post by: CopperCone on September 05, 2017, 09:44:34 pm
I noticed that a band pass filter I have will let through the harmonic of the pass band through a 'resonance' that is way more then what the slope would indicate. The area at twice the pass band is very spotty, and looks like this /\/\/\/\/\

I watched a video online that explained this, but it was relevant to microstrip filters.

I looked at some datasheets for bandpass filters, for instance the ones offered by pasternack, and they show a bode plot to about 2.5x their pass band, and this effect is not present.

I thought this effect was not present in waveguide filters, it according to my measurement it is.

What decides this? The waveguide filter (which I took apart), looked like it had little disks suspended by rods inside of it adjusted by tuning screws.

Are there any clues as to what a filter might do? (for 2nd hand purchasing).
Title: Re: what decides of a microwave band pass filter modes at passband multiples?
Post by: T3sl4co1l on September 05, 2017, 09:53:39 pm
I suppose a responsible filter would have extra shapes inside, or tuned dampening, to reflect or absorb higher modes.  But that's costly.

Low frequency filters have the same problem, where component parasitics reduce the stopband attenuation.  Solution?  Add more filtering, with smaller values, to preserve the stopband.

Same idea for distributed filters, but the higher modes are closer to the passband, so the design must be done more carefully.

Tim
Title: Re: what decides of a microwave band pass filter modes at passband multiples?
Post by: CopperCone on September 05, 2017, 10:07:53 pm
ok so its a crap shoot  :(
Title: Re: what decides of a microwave band pass filter modes at passband multiples?
Post by: CopperCone on September 05, 2017, 10:11:47 pm
Is the shape of the filter in any way indicative of this? I.e. modified waveguide vs a 'aluminum box' (which indicates milling) with rf connectors on opposite corners?

Like manufacturers tenancies?

I'm inclined to think that adding some doodads to a piece of waveguide is the 'cost sensitive' manufacturing technique... is it standard to not have harmonic peaking on 'machined' products?
Title: Re: what decides of a microwave band pass filter modes at passband multiples?
Post by: T3sl4co1l on September 05, 2017, 10:24:32 pm
I doubt it. And all those methods will have higher modes, just in different ways (waveguide vs. cylindrical/coaxial resonator modes, plus oddball modes dependent on the exact positioning of the screws).  You have to make smaller features to reflect or divert the upper modes.

Tim