[Elecraft] K2 KSB2: Signal Loss through SSB filters

[Ron D'Eau Claire]

Neat solution, Dick!

A common approach used by O.T.s was a "signal injection" probe, either audio
or wide-band RF noise. It was used to insert a signal at either audio or RF
starting at the final audio output and working back toward the antenna
input, stage by stage, until the signal disappeared or dropped way down.
That would be the stage where the problem lay. 

Back in the days when most receivers were simple single-conversion superhets
using a handful of tubes, a number test equipment companies made such
devices just for that purpose. 

I guess that's not so well known any longer because its usefulness is
limited in today's much more complex radios. It often takes something better
tailored to the rig, situation and the gear available on the bench, as you
did. 

Congrats, and thanks for the report!

I, too, found a huge improvement in the wider band SSB filter. It's much
easier on my ears and gets excellent reports from others, where it seemed
that I was constantly fiddling with the BFO frequency at 2.1 kHz to split
the difference between audio that sounded muddy or too bright for my taste.


73, 

Ron AC7AC



-----Original Message-----
Problem solved.  I'd like to report the cause and suggest a signal  
tracing method that might be helpful to someone without much test  
equipment.

First, though, I'd like to thank WA6VNN and LA1PHA for their  
suggestions, and of course thank Don, W3FPR, for his customary fast  
response and detailed advice.

The problem was lack of continuity between a trace on the top of the  
circuit board and the pad on the bottom of the board.  I had damaged  
the board when removing the old parts and cleaning the hole, and it  
was not obvious to me from casual visual inspection.  A  short jumper  
on the bottom of the board to parallel the trace solved the problem.

I found the break with the aid of a separate general coverage  
receiver tuned to the K2 IF frequency.  My test probe was a short  
lead on the antenna jack of that receiver.  I fed a strong signal  
from an oscillator into the K2's antenna terminal and tuned the K2 to  
that signal.  Then I put my probe on the input to the KSB2's first  
crystal and adjusted the oscillator output so that I had a near full  
scale S-meter reading on the general coverage receiver.  Next, I  
worked my way with the probe through the filter network, touching  
each coupling capacitor connection between crystals.  I found the S  
meter on the general coverage receiver dropped to about mid scale  
between the input of X3 and the input of X4.  Sure enough, my  
Ohmmeter revealed a break on the trace at the output of X3.

This approach to signal tracing probably is not new to many old  
timers.  However, it is quick and easy, very sensitive, gives both an  
aural and visual indication that you're actually following the  
signal, and works for those lacking test equipment.  For more serious  
tracing with a general coverage receiver it probably would be good to  
make a real probe, use coax to the probe, and put a small (e.g., 5 to  
50 pf) capacitor in series with the line.

BTW, upgrading to the 2.6 KHz SSB bandwidth, from the original 2.1  
KHZ passband in my older K2, made a dramatic difference in audio  
quality.

Dick, K0KK