Posted by
Geoffrey Mackenzie-Kennedy-2 on
Aug 12, 2006; 5:22pm
URL: http://elecraft.85.s1.nabble.com/Up-conversion-tp392867p392908.html
Jack Smith wrote:
> Hence, the resonator Q requirements for a 5 KHz filter at 40 MHz are not
> grossly different than for a 500 Hz filter at 4 MHz, which is quite
> achievable. But for a 400 Hz filter at 45 MHz, the required resonator Q's
> get into the million range.
-------------------------------------------------------------------
Quite right! The first roofing filter that I built for trial used 12
inexpensive 36MHz crystals, not far from the IF to be used, with holder
capacitance compensation and strung out in a ladder. While it's 6db
bandwidth was a nominal 3 kHz and its shape factor was very good with
virtually a symmetrical response, its ability to handle large signals was
not good and a similar filter at the actual IF would have compromised the
overall receiver performance. For a reason that I do not yet fully
understand the problem appeared to centre around crystal #5, counting from
the input.
The 3kHz and 1.5kHz roofing filters in use each consists of three four pole
"crystal filter modules" each separated by very high IIP3 low noise
push-pull amplifiers running at very low gain, but power hungry. To reduce
the 6db bandwidth in the1.5kHz filter modules I have resorted to a "brute
force" method by adding shunt crystals (Dishal) in each module. With the
amplifiers the overall loss can be controlled, keeping in mind the overall
receiver performance requirements.
Much too expensive for a commercial rig, there must be better ways to beat
Q.
73,
Geoff
GM4ESD
_______________________________________________
Elecraft mailing list
Post to:
[hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
http://mailman.qth.net/mailman/listinfo/elecraft
Help:
http://mailman.qth.net/subscribers.htmElecraft web page:
http://www.elecraft.com