Tweet
After re-populating the failed amp circuit with new xstrs, resitors (where needed) and diodes, electrolytic caps that would have taken a hit from such widespread failure, I finally had the module ready to power up. I first used my HP 6227B 0-25VDC Dual Tracking supply to check it out, and, after finding I had installed an MPS A42 backwards, I corrected that, and had a working amp, using an external -15V supply to apply the Unmute voltage to turn the muting FET’s Q1 & Q2 off.
When I moved it back into the chassis, I found I was unable to dial in the offset voltage to zero the output buss voltage to 0V. I was seeing -0.25V at the bases Q3/Q4 of the input stage. The other module (Left Ch) had -0.1V at the bases, and I was able to dial it to 0V on the output. After making measurements on the dual diff amp front end, I found this:
Q3 base -0.25V
Q3 collector 77.4V
Q3 emitter -0.86V
Q5 base -0.25V
Q5 collector 79.4V
Q5 emitter -0.86V
Q4 base -0.25V
Q4 collector -77.4V
Q4 emitter 0.35V
Q6 base -0.25V
Q6 collector -77.4V
Q6 emitter 0.35V
If I shorted out R3 bias resistor for Q3/Q4, voltage at the bases of Q3/Q4 dropped to -50mV. I was wondering if I’m dealing with leakage from the JFET’s. I swapped the pair of them out, and the base voltage increased to -0.3V just to spite me. I pulled the 1M offset adj pot VR1, finding it ok. Its’ series resistors R14/R15 measured fine, as did the four 100k emitter resistors for the front end. I moved the module back out to power it externally, and played with the unmute voltage, finding the threshold for unmuting the JFET’s was about -7VDC. -15V was being applied to J1 Pin 2, then applied to D2/R7/R6, so I tried putting a 10V zener across C4. That seemed to change the base voltage, so, I moved the module back to the chassis, and tried that at nominal supply voltages. Made no difference. I had a 6.2V zener, and tried that, no difference, nor did an 8.2V zener.
So, I pulled the module back out, and removed the two JFET’s, just to see if I was dealing with some leakage from them. No difference, still had -0.3V at the bases of Q3/Q4.
Then, placing a 10k across R3 (39k), that dropped the base voltage to -80mV. I removed R3 and installed a 10k there, put the JFET’s back into place and put the module back into the chassis. The base voltage of Q3/Q4 settled to -100mV, the same as on the other module. That at least allowed me to adjust the DC offset for 0V on the output.
I just don’t see WHAT is causing the neg voltage at the bases. Is it from NOT having matched NPN’s & matched PNP’s for the front end? Even better, having both matched NPN’s/PNP’s and matching each other on a common substrate?
Earlier, after having found I had Q5 installed backwards, and correcting it, I ended up replacing both Q3/Q5, just in case that mistake caused issues, and also Q4/Q6, while looking for the cause of the negative base voltage.
I looked all over the PCB for any leftover issues from the total failure of all the semi's, less 3 signal diodes, but didn't find any additional surprises. It does at least adjust now for 0V on the output, bias works, and both modules behave the same. I didn't find any other issues in the chassis from the module failure.
The amp schematic is the PM500, last schematic in the file attached.
WT800C schematic.pdf
When I moved it back into the chassis, I found I was unable to dial in the offset voltage to zero the output buss voltage to 0V. I was seeing -0.25V at the bases Q3/Q4 of the input stage. The other module (Left Ch) had -0.1V at the bases, and I was able to dial it to 0V on the output. After making measurements on the dual diff amp front end, I found this:
Q3 base -0.25V
Q3 collector 77.4V
Q3 emitter -0.86V
Q5 base -0.25V
Q5 collector 79.4V
Q5 emitter -0.86V
Q4 base -0.25V
Q4 collector -77.4V
Q4 emitter 0.35V
Q6 base -0.25V
Q6 collector -77.4V
Q6 emitter 0.35V
If I shorted out R3 bias resistor for Q3/Q4, voltage at the bases of Q3/Q4 dropped to -50mV. I was wondering if I’m dealing with leakage from the JFET’s. I swapped the pair of them out, and the base voltage increased to -0.3V just to spite me. I pulled the 1M offset adj pot VR1, finding it ok. Its’ series resistors R14/R15 measured fine, as did the four 100k emitter resistors for the front end. I moved the module back out to power it externally, and played with the unmute voltage, finding the threshold for unmuting the JFET’s was about -7VDC. -15V was being applied to J1 Pin 2, then applied to D2/R7/R6, so I tried putting a 10V zener across C4. That seemed to change the base voltage, so, I moved the module back to the chassis, and tried that at nominal supply voltages. Made no difference. I had a 6.2V zener, and tried that, no difference, nor did an 8.2V zener.
So, I pulled the module back out, and removed the two JFET’s, just to see if I was dealing with some leakage from them. No difference, still had -0.3V at the bases of Q3/Q4.
Then, placing a 10k across R3 (39k), that dropped the base voltage to -80mV. I removed R3 and installed a 10k there, put the JFET’s back into place and put the module back into the chassis. The base voltage of Q3/Q4 settled to -100mV, the same as on the other module. That at least allowed me to adjust the DC offset for 0V on the output.
I just don’t see WHAT is causing the neg voltage at the bases. Is it from NOT having matched NPN’s & matched PNP’s for the front end? Even better, having both matched NPN’s/PNP’s and matching each other on a common substrate?
Earlier, after having found I had Q5 installed backwards, and correcting it, I ended up replacing both Q3/Q5, just in case that mistake caused issues, and also Q4/Q6, while looking for the cause of the negative base voltage.
I looked all over the PCB for any leftover issues from the total failure of all the semi's, less 3 signal diodes, but didn't find any additional surprises. It does at least adjust now for 0V on the output, bias works, and both modules behave the same. I didn't find any other issues in the chassis from the module failure.
The amp schematic is the PM500, last schematic in the file attached.
WT800C schematic.pdf