[R-1051] mechanical filters

[John Kolb]

On Sun, 10 Aug 2003, Steve Hobensack wrote:

> The 3.8 khz mechanical filters in the r-1051b are just
> too wide for crowded ham band conditions. Has anyone
> succeeded in narrowing up the passband? What would
> happen if two 3.8 khz filters were cascaded ? Any
> thoughts on this will be appreciated.
> ....Steve....KJ8L

The specs for the LSB and USB filters in the R-1051 are:
3.2 kHz min @ 3 db, 6.1 kHz mzx @ 60db, 16 +/- 2 db
insersion loss, 1000 ohms input Z, 330 ohms output Z,
resonated with 130-150 pF. Filters are color coded with
dots near the terminal with required C, orange=130,
yellow=140, green=150 pF.

I'll try to find time to plot one and put the response
curve at <http://members.cts.com/king/j/jlkolb>

If putting two in series, it won't narrow the nose of the
filter much. The attenuations of the two filters will add.
Assuming the filters are identical, and assuming no leakage
around the filter, the freq on the response curve that was
3 db down will now be 6, 30 db down becomes 60, etc. Thus
shape factor improves greatly. Passband ripple will also
be doubled.
<http://members.cts.com/king/j/jlkolb/site/curves/LF-D6S.PDF>
is a spreadsheet calculation I did of two ceramic filters in
series with those assumptions.

Since the filter has 14 db insersion loss, if putting two in
series, an additional stage of gain will be required, preferably
between the two filters to match between the 330 ohms output
and 1000 ohm input Z.

I've got some 500 kHz 2.7 kHz wide filters on my For Sale page,
but these would require a bit of design change to match
the impedances properly.

John