[Elecraft] KPA100 PA Bias Adjustment Step

[J. Edward (Ed) Muns]

In the recent ARRL Sweepstakes, I used my K2/100 for the first time on SSB
since I built the KPA100 back in May 2002.  The only thing that gets me on
SSB is the Northern California Contest Club's plea to members for maximizing
the club score for certain contests.  Much to my chagrin, I got immediate
reports that I had "RF feedback" in my audio, or "distortion", or "narrow
audio".  I quickly went off-the-air and ran Spectrogram to move my SSB BFO
settings to a better, I thought, position.  I went into the Sweepstakes and
never heard another comment until right at the end when W6TER worked me and
remarked that "your audio is terrible".  Not what I wanted to hear.

The end result of my troubleshooting is that my KPA100 PA bias was set at 70
mA instead of the specified 400 mA.  However, this mistake was very
deliberate on my part.  I'm guilty of mis-interpreting the 6th step on page
47 of the KPA100 manual (Appendix G, Rev. A).  Maybe the following
explanation will help others avoid the same issue.

When I built my KPA100 field test unit back in May 2002, I had a question
about this step and mistakenly made a very literal (but, wrong)
interpretation of the wording.  To give the builder background, the manual
explains that the base K2 draws 200-400 mA in receive mode.  The addition of
the KPA100 adds another approximate 200 mA.  The PA transistor bias is to be
set at 400 mA, so this instruction step is aimed at having the builder set
R6, PA BIAS, on the KPA100 board to the sum of all these currents, or around
900 mA.  (Each builder is to measure his power supply currents, do the
addition, and come up with his own unique current setting.)  Here is a
comparison of the correct, and my original incorrect, calculation:

Power Supply current components:
                   CORRECT	INCORRECT
K2 RX 		200-400	disregarded!
KPA100 fan       	     50	     50
KPA100 T-R switch	    150	    150
KPA100 PA bias	    400	    400
                  -------	-------
                 800-1000	    600

My mistake was interpreting this instruction step as follows:

"Set the POWER control for 11 watts.  The current should increase
by about 200 mA (150 mA for the KPA100 T-R switch, and 50 mA
for the fan, running at low speed).  Record this current: _____mA.
Add 400 mA (200 mA per power transistor) to this to determine the
target reading for the bias-setting step below: _____mA."

I thought "... this current ..." in the third sentence was "The current ..."
in the preceding sentence.  WRONG!  "... this current ..." is the TOTAL
current reading of the base K2's receive current and this 200 mA increase
described in the preceding sentence.

Needless to say, biasing the PA transistors at 35 mA. each vs. the specified
200 mA. meant that my SSB signal was probably operating the PA transistors
in the non-linear portion of their curve.  On CW, no problem was noticed,
and as Eric pointed out, on CW there is enough filtering of any harmonics
generated that the output is OK.  On SSB, though, it was another story!

I think I've terrorized Wayne enough over this that he'll probably add to
the KPA100 errata sheet to help any future builders who possess my
particular logic pattern irregularities!

73,
Ed - W0YK