Home Stereo Amplifier Overload Indicator

[bobloblaws]

Hey guys, I'm not sure if this kind of gear gets discussed here, but I'm stuck. I have a Parasound HCA1000A solid state stereo amplifier here (schematic attached) and the right channel overload LED is permanently illuminated (amp powered on of course), even with no signal present. The left channel does not have this problem. I did some troubleshooting and found that both Q003 (right channel) and Q004(left channel) both have roughly 15 volts on the base and emitter, while Q0004 has 0V on the collector but Q003 has 5 volts on the collector which is dropped to 1.8V by R008 and is enough to turn on the LED. The surrounding components seem to be OK as far as I can tell (comparing left channel to right) so I figured Q003 had to be the culprit. However, after replacing it with another general purpose PNP transistor there is no change in the scenario and the original tests fine with the diode check on my DMM. I thought it would be an easy fix having the opposite channel to compare it with but like I said, I'm stuck. Anybody have an idea what might be going on?

[The Dude]

You say, "Q003 (right channel) and Q004(left channel) both have roughly 15 volts on the base and emitter". The problem you describe would indicate that Q003 is being turned on. You'll need more accurate measurements than "roughly 15V". A half a volt will make a big difference here. Please measure the voltage between base and emitter on both Q003 and Q004 for comparison and to verify Q003 is being turned on.

[bobloblaws]

You say, "Q003 (right channel) and Q004(left channel) both have roughly 15 volts on the base and emitter". The problem you describe would indicate that Q003 is being turned on. You'll need more accurate measurements than "roughly 15V". A half a volt will make a big difference here. Please measure the voltage between base and emitter on both Q003 and Q004 for comparison and to verify Q003 is being turned on.

Sure, measured to ground both Q003 and Q004 have 15.7V on the emitter. On the bases, Q003 has 15.1V and Q004 has 15.5V.

[The Dude]

So Q3 is turned on. My guess is that the right channel has enough DC on its output to trigger the LED. How much DC do you have on the output of the right channel compared to the left?

[bobloblaws]

So Q3 is turned on. My guess is that the right channel has enough DC on its output to trigger the LED. How much DC do you have on the output of the right channel compared to the left?

No DC on either output.

Why is Q3 turned on and not Q4? Are you looking at the differential between base and emitter? i.e. 0.6V for Q3 and 0.2V for Q4?

[The Dude]

Yes. At .6V, the transistor is biased on. Does that channel work normally otherwise? It sounds the same as the left?

[bobloblaws]

Yes. At .6V, the transistor is biased on. Does that channel work normally otherwise? It sounds the same as the left?

Yes, other than this issue it's working fine.

[g1]

Have you checked out D007, and the resistors connected to base of Q003 ?

[bobloblaws]

Have you checked out D007, and the resistors connected to base of Q003 ?

Yes, as far as I can tell they are OK. In circuit at least, I get the same measurements as on the left channel.

Now that I know that less than half of a volt can be the difference between the transistor turning on or not I think I need to look further upstream. For example on the left side there is 15.3V at D104/R006 while there is 14.3V at the right side equivalent.

[The Dude]

You might also unsolder one leg of D104 for the right channel to see if the voltage is coming from the amp or the LED circuit. The right channel isn't drawn on the schematic, so I'm not sure what its designation might be.

[bobloblaws]

You might also unsolder one leg of D104 for the right channel to see if the voltage is coming from the amp or the LED circuit. The right channel isn't drawn on the schematic, so I'm not sure what its designation might be.

It is D204. I snipped it and the voltage is on the LED side. That narrows it down. I had already replaced R005 as I had snipped it earlier before I disassembled everything to replace Q003. That would basically leave R004 or D007 (to G1's point) unless the LED itself was the problem.

[The Dude]

When you replaced Q003 with a substitute, were you sure to get the basing diagram correct?

[bobloblaws]

When you replaced Q003 with a substitute, were you sure to get the basing diagram correct?

Yes, I had to bend one leg over another one. I used a 2N3906.

[Jon Snell]

Q001 & Q002 detect DC on the outputs and will stop any more damage by blocking the drive the the speaker relay. What is the DC voltage on the junction of R129 & R132, It MUST be zero.
If it is, then start looking for a fault in the protection circuit but that will be rarer that rocking horse droppings!

[bobloblaws]

This is one of the weird ones I run into from time to time. (Jon, that rail is 0V and it definitely is in the protection circuit.)

I decided to experiment and see if I could get the base/emitter voltage on Q003 to be in line with that of Q004. Eventually I tried replacing R005 with a 68K. At that point the LED stayed off and base/emitter voltage was 0.46V. The voltage at the collector, which had been 5V when I first tackled this issue and 1.9V after I swapped out Q003 for a 2N3906, was around 0.9V but steadily dropping. At that point I changed R005 back to 100K. The base/emitter differential was still 0.46V and the collector was still dropping and eventually settled on a negligible reading of 0.008V. So then I put the original bipolar in Q003 and it still seems stable. But the datasheet for the 2SA970 shows Vebo as 5V. Based on what I learned earlier I'm guessing that means that if the emitter/base voltage hits 5V the transistor will turn on again and I'd be right back to where I started. And right now it is very close to 5V.