nice to hear from you pete !

this page explains how the graph was produced using a HP 3590A Wave
Analyzer and several other pieces of equipment:

http://members.cts.com/king/j/jlkolb/site/MFtest.htm


"Pete KE9OA" wrote in message ...
One other thing I forgot to mention............you will see the skirt
selectivity of a filter change in the cheaper filters, depending on the
drive level from the RF source. This would be caused from feedaround effects
inside of the filter. Probably, the best way to measure a filter in
question would be to use an HP 8753 Network Analyzer; this is what the
filter manufacturers would probably be using. When I was working for
Motorola, I did get some filter samples from Toko, and what they used was
and HP 3577 Network Analyzer. Alternatively, you could use a fairly good
spectrum analyzer, such as an HP 8590, along with a tracking generator, if
you are not interested in such things as input/output return loss, group
delay, etc.
The whole point of this is that unless you can duplicate the test
measurement setup of the manufacturer, you MIGHT be comparing apples and
oranges. Just a few thoughts.

Pete

Mike Maghakian wrote in message
om...
also the fact that you use excellent drake receivers means you know
that these receivers I mention use cheap filters and that shape
factors are important. If drake used filters like the receivers I
mentioned they would not be the classics they are today.

N8KDV wrote in message
...
Mike Maghakian wrote:

most receivers don't come with adequate selectivity. now for a small
amount and with a bit of skill you can change that. check out the
LF-D6 ceramic filter on my web page.

http://maghakian.home.att.net/

Oh yes. what receiver(s) are you proposing that these filters be put in?