Greetings and Salutations,
I got myself a PA 7E. Unfortunately, the a** who sold it to me thought it was a good idea to damage it in such a way that the new owner would just throw it away.
Long story short, I repaired the right side, but the left side is still tripping the protection circuit. It turns out that the output has an offset which triggers the protection circuit, and the pots don't change anything when turned either way.
Now, this is what makes it really strange (at least for me): Connecting the left side to my lab power supply (+-68V), that damaged amp works flawlessly. Unfortunately, that is the max I can get out of my supply. I know the datasheet says it should be +-60V, but the supply of my and other people's PA 7E seems to be running at +-80V. Anyways, I have also connected the (bad) amp to the other power supply side where the other amp works, and it does the same over there: the output has an offset, and the pots don’t change anything.
What could be causing this kind of behavior, and how could you repair something like this? Could a 12V difference trigger bad diodes and transistors and revert their behavior when under-volting? I mean, these are old semiconductors—is that something rather uncommon? I have come across half-bad semiconductors, but never ones that would behave differently with a 10% voltage increase/decrease.
Thanks in advance and Best Regards from Germany
I got myself a PA 7E. Unfortunately, the a** who sold it to me thought it was a good idea to damage it in such a way that the new owner would just throw it away.
Long story short, I repaired the right side, but the left side is still tripping the protection circuit. It turns out that the output has an offset which triggers the protection circuit, and the pots don't change anything when turned either way.
Now, this is what makes it really strange (at least for me): Connecting the left side to my lab power supply (+-68V), that damaged amp works flawlessly. Unfortunately, that is the max I can get out of my supply. I know the datasheet says it should be +-60V, but the supply of my and other people's PA 7E seems to be running at +-80V. Anyways, I have also connected the (bad) amp to the other power supply side where the other amp works, and it does the same over there: the output has an offset, and the pots don’t change anything.
What could be causing this kind of behavior, and how could you repair something like this? Could a 12V difference trigger bad diodes and transistors and revert their behavior when under-volting? I mean, these are old semiconductors—is that something rather uncommon? I have come across half-bad semiconductors, but never ones that would behave differently with a 10% voltage increase/decrease.
Thanks in advance and Best Regards from Germany
What happens when you power the amp to the same rail voltages via a variac (using its native power supply)?
If it doesn't react when changing bias and DC adjustment potentiometers I would suspect that the problem are there.
What are DC voltage at the three arrows?
Is signal passed to the output before the output relay (red arrow)?
What are DC voltage at the three arrows?
Is signal passed to the output before the output relay (red arrow)?
Unfortunately I don't have a variac on hand right now. But the output of the Nakamichi transformer after the diode-bridge is straight DC, that was checked with a scope. One of the old bridges was bad so I changed both for new ones.
With +-68V the amp is performing normal. - just a small offset which can be compensated with the pots.
That was also my suspicion, and this is indeed whats triggering the protection circuit. With the original transformer at +-80V there is a voltage offset of around 8v before the relay which keeps it from getting activated. At the green arrows I also do get a voltage difference of around 0.5V. One side at 0.55v and the other at 0.5V, if I remember correctly. I will double check and report back when I am home. But that sounds just about right, having a 10% (0.05V) difference at the drivers and and a 10% difference at the output (8V).
I totally forgot to mention that the drivers were both changed as they were bad. the were substituted for new original MJL3281A and MJL1302A. They measured very close in HFE at 1mA, maybe a 1% difference at most.
Between the green arrows you should have around 1,6 Volt difference. 0,8 on high side and -0,8 on lower side.
If you do not have something in this range please check voltage drop over R114 and R116.
I assume you have checked voltage drop between TP101 and TP102 which should be 40mV when idling current is set correctly.
If bias circuitry are damaged you probably have no voltage drop between TP101 and TP102.
If you do not have something in this range please check voltage drop over R114 and R116.
I assume you have checked voltage drop between TP101 and TP102 which should be 40mV when idling current is set correctly.
If bias circuitry are damaged you probably have no voltage drop between TP101 and TP102.
I finally got home and measured 0.64V between BE of the BJT 1302, and 0.5V between BE of the BJT 3281. I get around 80V at the output stage. The voltage drop across R114 is around 1.1V, and across R116 it is around 1.26V. I really do not understand why this is only happening with a ±12V increase.
Voltage drop over feedback resistor and what DC voltage on the bias spreader side of that resistor?
Have you set the VR102 to have 20mV over the test points TP101 and TP102 ?
That's a bit low. Should be 40mV. You got the instructions in the document you provided.
0,5 V over a BE junction raises questions if the transistor is turned on at all.
Have you set the VR102 to have 20mV over the test points TP101 and TP102 ?
That's a bit low. Should be 40mV. You got the instructions in the document you provided.
0,5 V over a BE junction raises questions if the transistor is turned on at all.
Last edited:
Be very careful adjusting VR102. If you set resistance too low, bias current will become dangerously large.
Yes.
The good side is working flawlessly at 20mV while burning much less energy. Output signal looks good as well on a scope.
The voltage drop across R109 is not measurable; it fluctuates from around -0.3V to 0.1V. After a few seconds, the output also starts to fluctuate.
Very odd!; I've put a 500W bulb in series with the 230V AC input And the device works as expected. The caps are charged to +-86V and the output relay turns on. Output signal looks good as well.
I've set the voltage between TP101 and TP102 lower than the service manual specifies, thereby increasing the resistance of VR102.
Great...
Working amp AND new lightening.
Win-win...
Must say that I'm out of clue what it can be.
Need to have it in front of me, and that is not gonna happend, so I have to pass further ideas.
Good luck...
BC of Q102 measures 0V against the output of the output transistors, looks good to me.
I am curious why the supply rails measure so much higher than the nominal 60V. Is this a different variation that’s not shown in the schematics/service manuals? What’s your mains voltage? Any tap options on your power transformer?
You mentioned a previous owner had worked on the amp. Are you certain he didn’t replace transistors with counterfeit parts?
Unless I missed it, you haven’t explicitly mentioned how large the output offset voltage is. From a comment in post 5, I infer it might be about 8V. Would you elaborate?
Note that this design does not include the output stages within the global feedback loop. The output stages should have a gain of 1 and have only a few tens of mV offset, measured from output (at relay input) with respect to wiper of VR101, i.e. low enough offset to be cancelled by adjusting VR101. If you see large offset, the problem probably is somewhere in the output stages. This conjecture underlies my earlier questions.
I can suggest a test to isolate output stages from the front-end, but will defer that until further discussion.
Good luck!
You mentioned a previous owner had worked on the amp. Are you certain he didn’t replace transistors with counterfeit parts?
Unless I missed it, you haven’t explicitly mentioned how large the output offset voltage is. From a comment in post 5, I infer it might be about 8V. Would you elaborate?
Note that this design does not include the output stages within the global feedback loop. The output stages should have a gain of 1 and have only a few tens of mV offset, measured from output (at relay input) with respect to wiper of VR101, i.e. low enough offset to be cancelled by adjusting VR101. If you see large offset, the problem probably is somewhere in the output stages. This conjecture underlies my earlier questions.
I can suggest a test to isolate output stages from the front-end, but will defer that until further discussion.
Good luck!
The mains Voltage in my region is 230V +-10%.
There are no taps available to change the output voltage.
The inrush current PCB has 220V-240V etched into it.
Either there are different versions out there or there is an error in the schematics. // yes that is possible, there is even an error in the TEK2465B service manual...
Unfortunately yes. but the transistors looked legit to me.
That was a mistake on my part. I had way higher offset.
Output stage is fine. no issues there.
@Nico Ras had the right answer. There were indeed leaky transistors, which showed their bad characteristics only under given circumstances.
Some 2sc3333 and the 2sa1321 where leaking at higher voltages.
They will be substituted with fairchild transistors.
Thank you all for the help!
Some 2sc3333 and the 2sa1321 where leaking at higher voltages.
They will be substituted with fairchild transistors.
Thank you all for the help!