On Sat, 23 Dec 2006 11:32:04 -0500, ken scharf
wrote:

Yet another idea is a single conversion with a 9mhz IF. Since I have a
bunch of surplus 9mhz filters (they are 8 pole units with 3.2 khz
bandwidth) I was also thinking of a rig with these. True the filters
are a bit wider than common today, but if I put THREE of them in cascade
(between IF stages) they should do a good job.

A single 9 MHz IF filter unit with I/Q detection (to handle the
opposite sideband) could be an other alternative and do the rest of
the filtering in audio stages.

I would use a DDS VFO,
but would also used a tuned (not broadband) front end. I have enough
toroids and multi section variable caps in the junk box for that.

The _unloaded_ Q values shown by toroid manufacturers are not very
spectacular (in the 200-300 range at most). If you aim for a filter
loaded Q of 100, there are going to be a considerable loss (several
dB), so placing the filter before the first RF amplifier stage will
deteriorate the noise figure quite badly, which can be a bad thing on
upper HF bands.

With a preselector loaded Q in the 50-100 range would still cover an
entire WARC band without tuning and with wider bands and tunable front
end filters, a 100-500 kHz segment would still be present at the mixer
input at full amplitude. Thus, the mixer would still need to be strong
to handle all those signals in that range.

A preselector filter will most definitively help in keeping out strong
broadcast band signals (e.g. the strong 49 m BC band in Europe) from
the mixer, but it does not help much against strong amateur signals in
the same amateur band.

Paul OH3LWR