[Elecraft] KPA100 problem fixed

[David A. Belsley]

Well, gang, the problem with my KPA100 that I reported a couple of weeks 
ago is fixed, thanks in good part to Gary.  I discovered early on what the 
problem was, but finding the cause was a bit more problematic.  I was 
getting HI RFL and HI CUR displays, at first rather intermittently, then 
for longer stretches.  Some voltage measurements quickly determined that 
pin 18 of U2, the IC that controls the relays in the various paths of the 
LPF, was much too low, about 4 volts instead of 13vdc. This is the pin that 
controls K3 and K4, the relays in the 160 meter path of the LPF, and the 
low voltage was allowing these relays to switch on when operating on other 
bands.  Clearly this would rather upset the KPA100 on every band but top 
band.  A quick check also showed that it was working just fine on top band. 
The resistance on pin 18 was also quite low, initially about 2K ohms, where 
Gary indicated it should be in the meg ohms.  I cut pin 18 to see if U2 was 
bad in that path, but the resistance stayed the same.  Also the resistance 
was lower than it should be on the other U2 output pins and 12CTRL.

Some preliminary measurements didn't show anything in that branch to be 
obviously bad, but, of course, you can't see under the relays.  An e-mail 
to Gary brought an interesting notion that it might be a trace with 
incomplete etching.  Apparently this has happened.  There was nothing 
obviously wrong in inspecting the traces or associated components.  Gary 
suggested hitting the pad of pin 18 with a current-controlled 12 volt 
source to burn away any impurities or hidden bridges.  That indeed helped. 
The resistance went up to 16K and voltage went up enough to prevent K3 and 
K4 from closing, but neither was as high as it should be.  So, while the 
KPA100 worked, I wasn't happy.  I finally pulled the board and shipped it 
to Gary.  He zapped the trace again with 12vdc and also with the 150 volt 
t/r source.  That apparently did the trick (taking the I/O chip with it). 
It turned out that C12 was also open, but that clearly was not involved in 
the problem.

In any event, once everything was put back together, the KPA100 was working 
beautifully and had the proper several meg ohms of resistance on the 12CTRL 
line.

Again, you simply can't beat the Elecraft service.  Gary was super support, 
giving great advice, experience, and skilled service.  The turn-around was 
within a week.  As one of the chaps I was qso-ing with today remarked, 
"hard to find things like that now-a-days."  That's for sure.  Thanks Gary, 
thanks Elecraft.  And also many thanks to John, K1JD, and Ken, K3IU, who 
both offered at an early stage to come visit (no short trip), bringing 
another KPA100 to try, just to be sure the problem was with the amp and not 
the K2.  Before they had to make that trip, however, I was able to 
determine where the problem was in the amp.  But I was sure grateful to 
them for such a wonderful offer.

best wishes,

dave belsley, w1euy

----------------------------------
David A. Belsley
Professor of Economics