BA115 DC on speaker output

[chris1954]

Hi - I've found multiple issues with this amp....replaced the zeners (D8 and D9) - one was shorted out - now the LED comes on versus no LED prior. I've also replaced Q2 and Q5. But I'm seeing -34v at the speaker and when I run a signal into the amp, I see it die around Q6 area.

Any ideas would be helpful!

Thanks!

[Enzo]

Work with no load until the amp is stable and not making DC.

Don't even think about signal while there is 34vDC on the output. D8,9 shouldn't affect the power light, but do you have good power rails? No system works if the power supply isn't right. You have +40 and -40, more or less? I don't care if they are 42v or 37v, as long as both are about the same. And do you have both +16 and -16? Easiest way is to find the two large resistors, R66,68, 270 ohm 5 watt. on each, 40v at one end and 16v on the other.

If a 16v is missing, that will shove U3 in the power amp over to one side. So what is on pins 1 and 7 of that IC? Whatever is on pin 7 is likely to be what is on the speaker too.

You had to replace the outputs? OK< any time you do that in any amp, the drivers are instantly suspect, as they likely were stressed by the bad power transistors. Also always check the associated resistors, in this case: 47 ohm and 220 ohm. I usually replace the drivers even if they SEEM to be OK.

Once the amp settles down to zero volts DC on the output, THEN worry about signal.

[chris1954]

Thanks! The only reason I was worrying about signal was that I thought the amp was fixed. I had no LED until I replaced the Zeners. So I hooked up a crappy speaker and wham - it blew. Checked for DC and there is was. I'll look into your suggestions....super thanks for your reply!! I'll report back...

[chris1954]

Hi Enzo - Here is what I found....

D8,9 shouldn't affect the power light, but do you have good power rails? No system works if the power supply isn't right. You have +40 and -40, more or less? I don't care if they are 42v or 37v, as long as both are about the same.

>>>>>>>>>Chris: Yes. + and - 35v or so.



And do you have both +16 and -16? Easiest way is to find the two large resistors, R66,68, 270 ohm 5 watt. on each, 40v at one end and 16v on the other.

>>>>>>>>>Chris: No - Here is where things are wacky. R68 is 10.7v / 35v (and running hot), R66 is -16v / -35v (warm too). The 10.7 is suspect...but I'll listen to your thoughts!


If a 16v is missing, that will shove U3 in the power amp over to one side. So what is on pins 1 and 7 of that IC? Whatever is on pin 7 is likely to be what is on the speaker too.

>>>>>>>>>Chris: U3 pin 1 is 11v, pin 7 is 4v, pin 8 is 11v, and pin 2 is 11v. Speaker is 35v


You had to replace the outputs? OK< any time you do that in any amp, the drivers are instantly suspect, as they likely were stressed by the bad power transistors. Also always check the associated resistors, in this case: 47 ohm and 220 ohm. I usually replace the drivers even if they SEEM to be OK.


>>>>>>>>>Chris: I measured R34, 41, 35, 31 - all are 220 Ohms. R33 and 49 measure 47 ohms.

[Enzo]

The two resistors DO run hot normally, but if the one with 10v on it is extra hot, it is because more current is drawn through it pulling down the 16v. They dissipate more than 2 watts, which is plenty to get hot.

My first suspect on the 10v instead of 16 is the zener is dying. otherwise something in the circuit could be loading it down. Such a loading would most likely be a bad op amp, any of them getting hot?

[chris1954]

Are you referring to D6, D7, D8 and D9 on the zener's?

I didn't feel any of the op amps getting hot...but I'll check again. How can I test if an op amp is bad? I can't remember how to do this....from my school days....a long long time ago!

And yes - R68 with the 10V is for sure running hotter.


Again - I really really appreciate your help. Big time.

[Enzo]

No, the zeners in the power supplies that set the voltage, in this case 16v zeners. In this case D11, D12. If D11 collapses, then instead of -16v, you get -10 or -8 or -5 or whatever. If it outright shorts, you get zero volts there and the resistor gets REAL hot.

I test op amps by function, they work or not, and they don't have voltage on their output that doesn't belong there. I can unsolder it and pull it out and see if the voltage in the circuit rebounds to 16v. I still suspect the zener first.

[chris1954]

Checked D11 and D12 - both are fine. Even replaced them and same issue. No op amps are warm to the touch.


There is a consistent 10v on each op amp....this makes sense....and everywhere where there should be 16v - there is 10v. Makes sense too...


It would seem to me that where the 16v is created is where the issue might be....but again, I look to you, the expert.



Thinking out loud, +40v is dropped to +16v. I'm now wondering if R68 is the issue. It measures ok, but it seems like after R68, we only have the 10v. But again, I'm not an expert. What creates the voltage drop from +40 to +16v?

Also could the driver transistors be an issue? Q4 and Q1? hmmm....just thought of that.

[Enzo]

The zener and the op amps are all connected together. The zener decides the voltage. You can test the zener as a diode to see if it is shorted, but the only way to see if it functions as a zener is to apply working voltage to it, something your meter does not do.

If the diode is bad and only makes 10v instead of 16v, then everywhere 16v would be there would be 10v instead. Of course something on the 16v line could be dragging it down, like a stuck parking brake.

What creates the voltage drop? On a plain resistor, the current through it, that is Ohm's Law. But here, the end voltage is not set by current, but by the zener diode voltage. If the resistor measures 270 ohms or so, then it is fine.

You had 16v 5 watt zeners in stock??

If you replaced the zener and still get 10v, and the resistor is 270 ohms or even remotely close, then we look at the load. I see five op amp ICs. I see a couple connections to the power amp, but they are through 1k resistors, which tells me a dead short to ground through 1k would only draw 16ma, not enough to load it down. Couple diodes at the input jack, I doubt they are shorted, but easily checked. C39 could be bad.

If after sitting there a few minutes, my op amps were not getting warm, then I would probably start lifting them from the circuit to see if the voltage recovers.

[chris1954]

Progress.....I pulled out the drivers. After I pulled out Q1, +16v appeared. Gonna order new Q1 and Q4 transistors!

Also when I put Q4 back in and left out Q1 - no DC at the speaker. I did ohm both Q1 and Q4, and they both look ok...but I'm just gonna replace both of them.

Turns out that lightning took out this amp....



Does this all make sense?

[Enzo]

Not quite, by "ohming" I assume you used resistance measures? To see if the part was shorted? What about open? You should use the diode test function of your meter to test the base to collector and base to emitter junctions Then also emitter to collector both directions to make sure no short or leakage exists.

Q1 was killing your 16v, so obviously it has a fault that your meter is not detecting. I wouldn't have thought Q1 could kill the 16v, but I guess I was wrong. At least I was right to suspect the drivers.

[g1]

I wouldn't have thought Q1 could kill the 16v, but I guess I was wrong.

I didn't think so either. An extra 16mA? It must be running very close to capacity under normal conditions. Or perhaps bad solder at the R68 connections?

[chris1954]

Sorry - I did both...looked for open and the diode function. Agree on the fault the meter is not detecting....unless there is something else....


Someone also asked about solder at R68 - I resoldered both R68 and 66. Also checked C39 and everything along the +16v line.....


Hoping new drivers fix the problem....

[chris1954]

Update - Not good. Replaced the drivers - 340 and 350. Now I'm blowing fuses....argh!

[chris1954]

Hi Enzo and all - I gave up for a few months...but back at it. Looks like one of the mosfets IRFP-140R (Q5) is shorted. Must have been bad - I got it from eBay...after reading, seems like that is NOT the place to buy parts...so I have 2 new MOSFETS on order - Q2 and Q5.