[Boatanchors] Re: IF Filter testing...

[J. Forster]

Barry L. Ornitz wrote:

> Hi John,
>
> This is a good idea and I have used a noise bridge in this manner with
> a receiver to simulate this technique.  You still have the problem of
> identifying which peak is the desired one and which are spurs.

Not really. Because the filter is essentially operating steady state, the true
curve is shown.  Sweep time is not limited by filter Q.

> To see a good filter response curve this way, you still have to do
> lots of averaging with the spectrum analyzer output.  Filters "ring",
> especially if driven to hard.

It's much harder to overdrive a filter with white noise than with a pure tone.
The thing is running pretty much in steady state. In practice you don't see the
dynamics using this method.If you bring a tone into the BP of a high Q filter
and then out, it will ring for a while.  If you use an ANA or tracking
generator, and look at a 'rabbit ears' (i.e. flat phase filter), and sweep too
fast, this effect will distort the shape. If the peak shape is not distorted,
it's much easier to sort out what the thing really is. I've used this with SAW
filters, which are a dogs breakfast using other techniques. Also, a decent
spectrum analyzer will cover 1 to 1800 MHz in one sweep, which helps to identify
spurious responses.

You could also use a frequency selective voltmeter  or good receiver as a
detector, BTW
-.John

>         73,  Barry     WA4VZQ     ornitz@tricon.net
>