[Elecraft] KPA100 - Low Voltage output from High Voltage Bias Supply Circuit

[Michael Haygood]

Hi,

I've been using my K2/100 for many months with great results.  After a 
recent storm containing lightning the next time I fired up the K2/100 I 
received the "HI CURR" message and the "HI REFL" messages upon initial 
transmit.  This was unusual, so the next thing I did was consult the 
troubleshooting guide in the KPA100.

Using a wattmeter and connected to a dummy load, I measured the output power 
of the K2/100 with the power outupt contol set to less than 11 watts and 
noticed that for a setting of 5 watts the output of the K2/100 was about 10 
watts.  For a power output setting of 15 watts, the output was very high - 
about 100 watts.

The troubleshooting guide indicates that for a HI CURR message the steps are 
to set CAL CUR to at least 3.50 Amps and verify the KPA100 T-R component, 
R21 - R22, T1 and nearby components.  I verified that all of these 
components were good; the CAL CURR was already set to 3.50.

Additionally, I re-calibrated the power out using R26 and R27.  When 
performing this procedure, I recalibrated at 5 watts output on the 
wattmeter.  I had to change the original setting of R26 alot to 
re-calibrate.  After completion of this procedure, I verified the output at 
various power levels.  The output was too low on all bands when the power 
output setting was set to 100 watts - so I re-calibrated using an output 
power of 50 watts.  The results revealed that the power output was still off 
by alot when setting output power to 5 watts and 100 watts.

The next thing I did was verify that the bias current was properly set.  I 
followed the Bias Current Adjustment procedure step-by-step from the KPA 
manual.  The bias current was set properly, but upon performing the last 
step in this procedure I noticed that the voltage on the left side of R12 
was only 49.7 voltage instead of 100 to 110 volts when the power knob was 
set above 11 watts!  This gave me a clue that perhaps something was wrong in 
the high voltage bias circuit.

I then consulted the circuit description for the high voltage bias circuit 
and learned that the 18.432 MHZ crystal oscillator, Q8 and T3 are used to 
derive the T-R switch bias voltage (90 - 150V).  D1 - D8 are used for 
rectification and voltage doubling.

I connected an oscilloscope to the 18.432 MHZ crystal and verified that it 
was "oscillating" on the scope - although it did have a weak signal using a 
1X probe on the scope.  I then consulted the DC voltage chart for Q8 and D1 
through D8.  Additionally, I did a diode check using my DMM and D1 through 
D8 all passed the diode check - no open or shorted diodes.

Knowing that the voltage form the high-voltage bias supply circuit was about 
one half the minimum designed voltage of 90 volts, I suspected that perhaps 
one of the diodes in the high voltage supply circuit may not be working 
properly since the diodes are used for rectification and voltage doubling.  
I measured the anode and cathode voltages with respect to ground for D1 
through D8 and here are my measurements:

D1:  Anode  15.7 V  Cathode  49.7 V
D2:  Anode  34.9 V  Cathode  15.7 V
D3:  Anode  14.5 V  Cathode  34.9 V
D4:  Anode  19.7 V  Cathode  14.5 V
D5:  Anode  14.9 V  Cathode  19.7 V
D6:  Anode  13.5 V  Cathode  14.9 V
D7:  Anode  -4.5 V  Cathode  -1.2 V
D8:  Anode  -1.2 V  Cathode - 0.00 V

The D4 cathode and the D3 anode connect to the output of T3 - the voltage 
measured at this juntion was 14.5 volts; the DC voltage chart reveals that I 
should have measured about 21 volts.  So I verified that T3 had no open or 
damaged windings and verified the 18.432 MHZ crystal waveform across the 
"input and output" windings of T3.  Is 14.5 volts too low at the D4/D3 
junction?  This seems to be one of the "input" voltages to the diode 
rectification and doubling circuitry.

I next checked the voltage on the Q8 EBC.  Voltage readings for Q8 were E - 
zero volts; B was -200 millivolts; C was 11.6 volts.  These readings seem 
reasonable with the exception of the Q8 base which should have read 
approximately -2.5 volts.  Perhaps the output of the crystal is too low - R4 
has a value of 39K (connected between the base and collector of Q8).

After all of this work "troubleshooting" and measuring voltages in the high 
voltabe bias supply circuit I have the following questions:

1.  Could the 18.432 MHZ crystal output be too low, thus not generating 
enough "derived" voltage for the high voltage supply circuit?

2.  Could Q8 - which seems to "amplify" the crystal waveform be bad or not 
biased properly to amplify the crystal signal?

3.  Could one of the diodes in the rectification and doubling circuity be 
bad?  However, all the diodes did pass the diode check using the DMM.

I'm stumped, and feel as if I'm close to solving this problem.  Perhaps 
there is someone on this board who could provide some feedback about this 
problem.

Thanks in advance.

73,

Michael Haygood
KI5E

_________________________________________________________________
http://im.live.com/messenger/im/home/?source=hmtextlinkjuly07