How to dampe standing wave in reflective filters

[An_RF_Newbie]

Hello,
In a application I use two back-to-back reflective filters. As you know, these filters shapes standing wave between them by unwanted signals in their stop band. My main input signal is about 1dbm and unwanted signal before these filters is less than -20dbm. I need to know how to bypass the standing wave between these filters. (Frequency less than 6Ghz)

 

[dick_freebird]

Yeah, problem is that a reflective filter ruins the
50-ohm match (at frequency band of interest, or
at all others, depending on whether band reject
or band pass).

Perhaps there is a clever matching network which
acceptably degrades the forward power while
letting the filters still work. What about a simple
series 50 ohms between the two filter stages?
Not that I have much of a RF clue, just some
radio stains on an old lab coat.

 

[biff44]

there are a couple ways to do it.

1) put a 3 dB attenuator in between the two filters. that way bounce back signals are 6 dB less (3 db one way, and 3 dB coming back) and you might be able to live with the loss
2) build a diplexer filter so that out of band signals to a different port (which you would terminate in 50 ohm load). the theory behind it is to make a bandpass and bandstop filter, connected together at a juncion, but instead of a "chebychev" or a "max flat" design, you would use a "singly terminated" design. In narrow band applications, you can just get away with the two filters and some sort of reactance canceling element near the junction
3) use a ferrite isolator in between the two filters (we used to use them all the time, back in the day, but seldom do now due to the added expense)

 

[An_RF_Newbie]

Thank you very much biff44, could you let me know how you define 3dB value attenuator?

 

[ge]

John.Jin search on keyword “pi pad attenuator” or order them with your favorite distributor.

 

[An_RF_Newbie]

Thank you dear ge,
I need to know the optimum attenuation value for suppress the undesired standing wave.

 

[ge]

We have no method to determine the “optimum“ value. Likely the value will be between 3 and 6dB. Build one and have an attenuator kit on hand.

 

[biff44]

Thank you very much biff44, could you let me know how you define 3dB value attenuator?

is that a trick question?
you order one specifying "3 dB". (or some other value)

Like from Digikey:

If you measure a generator output with a power meter, and set the power to read 0 dBm (1 milliwatt), then install the 3 dB attenuator, after the attenuator the power meter will now read -3 dBm (0.5 milliwatts)

 

[dick_freebird]

Here's another idea - what about using a tuned stub
to take the time-of-flight reflection and cancel it at
the filter input port? That would be a narrowband
solution, but this sounds like an inherently
narrowband problem what with a filter cascade?