> >The only problem with filters is that they usually need something close
> to
> >their design termination impedance to have their frequency-responses
> stay
> >as-designed. A highly reactive load or one with an impedance far away
> >from the filter's expectations will upset it greatly.
>
> The mismatch has more effect on the passband than on the stopband -
> usually gives horrendous ripples in the passband with increased loss as
> the mismatch gets worse. However, really mismatching a crystal filter
> can give a Gaussian top for about the first 6dB, stopping ringing. But
> you then end up with about 20dB of insertion loss.
>
> 73
>
> Peter G3RZP
Yes, quite right. I didn't make my point well in that upsetting the
filter in its passband (i.e., where you are trying to make measurements)
will distort whatever it is that you're trying to measure. It may also
reduce the out-of-band attenuation to the point where you contaminate the
detectors with BC interference, as well.
73, Ward N0AX
-
---
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Problems: owner-topband@contesting.com
|