[Premium-Rx] Insertion Loss and Return Loss of Passive Filters

[George Georgevits]

Dear Group,

I have a technical question which has perplexed me for some time now. I
think it is slightly off topic, but it is nevertheless very interesting. It
concerns passive filters, and in particular, what criteria should be used to
evaluate their performance.

As I understand it, passive filters have to be matched to their designed
source and load impedances in order to function properly. Somewhere in my
great store of junk, I have a 600 ohm 8th order (I think) passive low pass
filter, and I remember when I was given it, I fooled around with it and
could only make it work properly when it was terminated with and fed from
600 ohms.

Now consider the situation where we have, as an example, a passive band
reject filter which was designed to operate in a 50 ohms environment and
intended to cover the range 1 - 30MHz with a minimum loss of 30dB across the
entire range. If I measure the insertion loss in a 50 ohm environment, I see
that the filter does indeed cover the full intended range with better than
30dB loss. However, when I measure the return loss (from either end), I see
that it is a very poor match to 50 ohms over most of the intended operating
range.

So my question is this - has this filter been poorly designed, despite the
fact that the insertion loss criteria are completely satisfied? Or another
way to put this question is - should the filter also have a good return loss
performance (ie. be a good match to 50 ohms) over the entire band of
operation, whilst maintaining its band rejection performance requirements,
in order to be considered as correctly designed?

I think this question is also applicable to band pass applications in
communications receivers post the first mixer.

I would be most grateful for a few words of wisdom on this one.

de
George Georgevits
VK2KGG