[Elecraft] KPA100 Problem solved

[Don Brown]

Hi Don, Leo and all

Maybe a little circuit description of the Temperature measuring circuit will
help locate the problem. This circuit is similar to a temperature probe I
worked on at Tektronix. If a constant current is forced through a transistor
the base-emitter junction voltage will track fairly close to 10 mV change
per 1 degree C. A large value (15K) resistor causes a constant current to
flow in Q3 and U7B, Q4 buffers the base voltage of Q3 and forms a
temperature regulated drive for the PA transistors. The base voltage at Q3
is also sent to an analog input on the KPA100 MPU. The firmware uses this
voltage, that varies with temperature, to calculate the heat sink
temperature. This works because Q3 is screwed to the heat sink so its
base-emitter junction will be at the heat sink temperature.

If Q3 was not screwed to the heat sink it would measure the ambient
temperature in the K2. It should not read a very high temperature. I think
the tightening of the screw may have flexed the board or the transistor,
causing a bad connection somewhere to reconnect. The problem may come back.
There is not enough current flowing in Q3 for it to self heat. What is worse
if there is a problem in the temp measuring circuit it could over drive the
bias on the PA transistors and cause the transistors to overheat. I had a
KPA100 today I was working on with the bias misadjusted. If you keyed the
mic the KPA100 would draw 10 amps without speaking into the mic. The normal
draw should be a little less than 1 amp. After adjusting the bias everything
was fine. I mention this because this is the symptom you would have if Q3
was not working or a bad connection in the temperature measuring circuit.

If a split lock washer is used on Q3 and Q4 without a flat washer or fiber
washer the split lock will sometimes crack the case of the transistor if
tightened to much. I use a fiber washer like the ones used to mount the
speaker on Q3 and Q4's mounting screws to cushion the transistors.

Don Brown
KD5NDB


----- Original Message -----
From: <W5EI@cs.com>
To: <elecraft@mailman.qth.net>
Sent: Monday, January 20, 2003 3:29 PM
Subject: [Elecraft] KPA100 Problem solved


> Leo,
>
> I don't believe anything was actually getting hot, because the PA hot
message
>  appeared upon intial power-up when the heat sink was quite cold to the
> touch.  The high temp reading in the CAL tpa menu setting would also show
an
> extremely high temp reading even when the ambient temp. in my shack was
only
> 5 deg. C early in the morning.  It's almost as if the erratic readings
might
> have been caused by a crack in the case of Q3, a possibility suggested by
> Don, KD5NDB.  I could not find such a problem when I looked at Q3; it was
> quite loose however.
>
> 73,
>
> Don
>