LM3886-based Guitar Amp Problems

I originally posted this in a couple of Reddit sub-forums, and it was suggested to try here:

I'm working on a Behringer GX212 guitar amp. While the schematic is for a similar model, the power amp section is identical to what I have. When I received the amp, it had at least one bad output, one was shorted and loudly humming, the other dead quiet. After checking for possible shorts or overvoltage issues, I replaced the electrolytic caps on the power amp board for the 40 and 15 volt supplies, as well as on the I/O lines. I replaced the shorted chip, and now I have two dead quiet outputs. I checked the mute circuit, which I suspected might be an issue. I used a 68K resistor from pin 8 to pin 4 to create a current flow of greater than .05ma, which should disable the mute circuit. It did not change, and I have 40v between pins 8 and 4 without the resistor. The resistors, diode, and transistors in the mute circuit appear to be good, so I'm at a bit of a loss. I'm not sure I should be seeing the 40v at those pins. Link to schematic below. Anyone with more experience able to lend a hand?

Thanks!

 
I'm working on a Behringer GX212 guitar amp. ..... I replaced the shorted chip, and now I have two dead quiet outputs. I checked the mute circuit, which I suspected might be an issue. I used a 68K resistor from pin 8 to pin 4 to create a current flow of greater than .05ma, which should disable the mute circuit. It did not change, and I have 40v between pins 8 and 4 without the resistor. The resistors, diode, and transistors in the mute circuit appear to be good, so I'm at a bit of a loss. I'm not sure I should be seeing the 40v at those pins. Link to schematic below. Anyone with more experience able to lend a hand?
Hi, a few problems with your post.

* unmute current is not >.05mA but >.5mA
Since that would be the minimum unmute current, I would go for at least 1mA EACH

* since both chipamp mute pins are in parallel, you need double current, so 2mA .. at least.

68k can provide meager 0.58 mA, absolutely insufficient to unmute, try 22k or 18k or 15k there.
For reference, measure voltage from pin8 to ground, I guess you should see 1 or 2 diode drops there(-0.7 to -1.4 V).

* you can not see "40V at those pins" , if anything negative voltages.
One is the negative rail, the other I suspect is some unmute transistor emitter.

When you mention negative voltages, please also write the "-" sign when proper.

Try it and post results.

Plan B: the trivial solution is that those chips are Fakes 😱 ; where did you buy them?
 
Hi, a few problems with your post.

* unmute current is not >.05mA but >.5mA
Since that would be the minimum unmute current, I would go for at least 1mA EACH

* since both chipamp mute pins are in parallel, you need double current, so 2mA .. at least.

68k can provide meager 0.58 mA, absolutely insufficient to unmute, try 22k or 18k or 15k there.
For reference, measure voltage from pin8 to ground, I guess you should see 1 or 2 diode drops there(-0.7 to -1.4 V).

* you can not see "40V at those pins" , if anything negative voltages.
One is the negative rail, the other I suspect is some unmute transistor emitter.

When you mention negative voltages, please also write the "-" sign when proper.

Try it and post results.

Plan B: the trivial solution is that those chips are Fakes 😱 ; where did you buy them?
First, thank you for replying with some great information.

I did mistype the unmute current value. I used the formula (Vcc-2.6)/.0005 to arrive at the starting value needed to draw the 0.5mA, then dropped a bit to increase the draw. I did not realize it needed to/could be as large of a draw as you've indicated, so I was running conservatively. I did end up using a 15k, between pins 4-8, resulting in a current draw of 2.5mA. This did cause the chips to unmute, and they are both functioning.

I rechecked my measurements. These ae the results:
  • Pins 1-7: 39(+)
  • Pins 4-7: 39(-)
  • Pins 8-7: 0V
  • Pins 4-8: 39V(+)
  • Pins 10-9: 1v, 80Hz Sine Wave
I did not mention any negative voltages in my original post, so there was no need to add the (-) to them. Thank you for the helpful suggestion, though.

As for Plan B: The chip I replaced was a NTE Replacement chip, P/N NTE7197. Source is unknown. The other chip is original to the amplifier, and is a JRC chip.

As both chips function with the resistor, it seems I'm on the right track with the mute circuit in the amp. Do you have any thoughts on why it may not be unmuting? Both transistors passed a simple diode test, and the resistors test at their rated values. Going to dig a little more into it tonight, as I have it on the bench, anyway.

Again, thank you for the information. I now know the chips are good, and they work, which was a bit of a worry. I look forward to your thoughts on the mute issue, and if I resolve it before, I'll post what I find.
 
  • Like
Reactions: JMFahey