[Elecraft] RE: KPA100 "negative resistance" -- more

[Ward Willats]

Hello All:

I sent Tom e-mail last night when I just finished basic assembly of 
my KPA100 and was performing the resistance checks on page 41. I too 
discovered the dreaded "negative resistance" on the high current 12v 
input, J3, and part of last night's exchange found its way onto the 
list this morning. (My collector measurements on the PA finals were 
goofy too.)

Tom (and others) spoke of this in some July 2003 posts (mine is -2K 
ohms on a 20K ohm scale after a lot of settling down from about -4K 
ohms). Different scales give wildly different values.

More interesting was a minus 0.4 to 0.43 vdc  magic voltage on this 
pin (that no doubt is messing up the resistance checks.). Tom had 
0.35 vdc stray voltage if I recall from last night.

Now, understand, we are talking about a KPA100 just sitting in the 
Panavise with no power supply anywhere close, BUT I have an 
incandescent 75 watt swing arm lamp just 5 inches above the board. 
When I turn this off, the stray voltage immediately drops and tapers 
off slowly (as capacitors discharge?)

      Start     10s      30s    1m       2m     5m
Vdc   .33     .315     .279  .256     .235   .215

I wrote Tom last night I didn't think it was photosensitive, and was 
going into the "stray rectified AC currents" camp --  but after 
messing around with a maglite flashlight this morning I'm not so 
sure. In fact, if I shoot a tight beam onto the ceramic cap of one or 
the other power transistors, I can make the voltage immediately leap 
into the 0.46 to 0.48 range. Brighter == more. (I confirmed this with 
the AC incandescent lamp too.) It is pretty obvious the power 
transistors are the source.

Now, if I turn off all the work lights, and measure in just ambient 
room light, I get the correct (> 10K) value on + J3 and the 
collectors of the finals.

Tom just pushed on with alignment and installation and all turned out 
well, I am going to do the same. Both Tom and I were using bright 
incandescents -- and the light has to be shining on the ceramic for 
this to happen. Dunno what you get under a florescent.

So I think I'm going to be OK, but maybe all you EEs and physics 
mavens can shed some "light" on what is going on. It is kind of 
interesting and perhaps should be a note somewhere for future 
builders if indeed it is benign.

-- Ward / KG6HAF