[Elecraft] SSB Tx audio on K2

[Ron D'Eau Claire]

Nice-looking sweeps, John. They are about what you should see in receive
with a noise signal and a spectrum display like "Spectrogram" gives you =
too.


"Ladder filters" such as the K2 uses are inherently asymmetrical. I'm =
not
versed in the mathematics behind the filter design, but I've built =
dozens of
these things using many crystals over the years based on the info W7ZOI
published in QST and in "Solid State Design for the Radio Amateur" long =
ago.
In my experience, the wider the bandwidth, the more asymmetrical the
response. All of my filters were designed for CW bandwidths - out to =
perhaps
1 or 1.5 kHz maximum, so it wasn't an issue for me. I was rather =
surprised
to see in the K2 what a real expert in filter design can accomplish with =
a
wider-bandwidth ladder filter for SSB reception.=20

Having messed with a lot of "empirical" designs, it's been really
interesting to follow the evolution of the modifications made to the K2
filters, starting with removing the severe "tilt" in the cleanup filter
about three years ago, and then the use of more tightly-controlled =
crystal
parameters along with the FB work John (KI6WX) has done with mods to
increase the SSB bandwidth.=20

Through it all, a basic difference is the USB and LSB remains quite =
audible
to me in TX and RX both. It's quite apparent in your plots. That slow
roll-off on the bandpass (left side of the green plot and right side of =
the
red plot) seems to me what we're hearing. It produces slightly greater =
bass
on one sideband and slightly greater treble on the other sideband. Of
course, the difference in your plots are entirely differences in the =
gain
settings you used. Since USB and LSB are selected by moving the BFO
frequency, not by changing anything in the filter, one plot is really an
exact "mirror image" of the other.=20

Ron AC7AC

-----Original Message-----
Season's Greetings, Everyone.

I ran an audio sweep of my KSB2 and posted the results at:

	http://www.qsl.net/kd2bd/ksb2.html

The results obtained are pretty interesting.


All the best..


73, de John, KD2BD