[Elecraft] SSB Filter Flatness anomaly

[Ron D'Eau Claire]

Hi, Fran...

There are two variations on the 2nd Xfil flattening mod. The first =
version
used a 120 pf cap. The second version developed by John Grebenkemper
improved the flatness further by replacing the 120 pf cap with a 100 pf =
cap
and adding a 5.6 k resistor between Pins 1 and 6 of U11 (this terminates =
the
output of the filter in the 5.6 k resistor). I found them a very =
worthwhile
addition to the original "2nd Xfil Flattening mod".=20

It sounds like you might have the parts for the second version and the
instructions for the original version.=20

Since the flatness varies according to the sideband, it looks to me like
there is still some residual tilt in your 2nd Xfil that is adding with =
the
residual tilt in the OPT1 SSB filter. That would account for the "sum" =
to be
tilted differently depending upon the sideband selected, and - true to
Murphy's law - the combination you happened to get provides the greatest
flatness on the least-used sideband.=20

I suggest replacing the 120 pf cap with 100 pf and adding the 5.6 k
resistor.=20

Also, while it is still recommended that L34 be tweaked for maximum =
signal
(or best s/n ratio if you can detect a difference), L34 will have some
effect on the final flatness. As you know, the setting of L34 is very
b-r-o-a-d, so you may find that you can use L34 to optimize the flatness
without compromising the performance of the receiver in any way.=20

73,=20

Ron, AC7AC
K2 # 1289

-----Original Message-----
I have been working on the filter flatness modification and have come =
across
something that does not seem to make sense to me.  The filter response =
only
looks like I would expect in the 'wrong' SSB mode.  That is things look
flatter in USB than LSB on 40 meters and LSB is flatter than USB on 15
meters.

For those who need a picture, go to
http://home.comcast.net/~f.belliveau/K2/SSB_Flat/FlatDesc.htm

Does anybody have a clue as to what might cause such an anomaly?

Fran