[Elecraft] QSK CLICK

[Henry Gardiner]

Fred and others,
   I've lost track of who originated what mod to eliminate that QSK click 
that started for me when I installed the DSP option into the K2.  What 
provided a good enough solution for me was the 100K resistor in series with 
pin 17 /MUTE of the MCU U6 on the control board.  Why this worked I can't 
tell you for sure.
   The down side is that I think I had to remove a soldered IC so I could 
cut a trace and install the resistor.  Getting the IC out was no problem 
for me since I have the appropriate vacuum desoldering equipment and 
experience to do it without hazard.  Then too maybe I'm confusing this mod 
with another mod.
   Under some conditions - maybe 1-3% of the operating time - a light 
thumping remains in my full range headphones for short periods from unknown 
causes.  It's so transitory that I never did anything further about it.
   Looking further down the road: One thing I found important is that the 
BFO has to be stable if the DSP is to track the crystal 
filter.  Spectrogram in waterfall mode works quite well using noise to see 
how well the two filter systems track as the K2 warms up.  If the K2 is 
connected to a stable signal generator putting out a signal on the K2's 
receive frequency, one can also check to see how the PLL reference 
oscillator is doing as things warm up.  The line for the received carrier 
and the width of the two passbands should ideally maintain a constant and 
rational relationship.
   I ended up gluing an LM34DZ temperature-reading IC to the ground plane 
between the PLL reference crystal and its associated varicaps.  This feeds 
an opamp which in turn feeds a correction voltage via a high value resistor 
to the varicaps.  Now the drift is somewhere between 0 and -3 Hz from a 
room temperature cold start to the maximum QRP sustained key-down 5W output 
temperature with the lid on.  However, the crystal quality control has 
quite likely been improved since my #4017 and maybe most people get good 
results with the stock solution.
   I added a similar pair of ICs in the BFO area to correct the BFO.  The 
results were +0/-3 Hz or better over the same temperature regime.
   I run the filter gain in the DSP at about -6db to avoid flat-topping of 
the signal in the DSP just outside the corners of the passband.  This 
avoids the harsh sounds.  The best setting could be expected to change from 
rig to rig.
   My narrowest DSP filter width is the minimum that can be set - 40Hz I 
think.  Using the 'soft' filter response for this setting avoids most of 
the ringing.

73s, Henry AC5LA