Ampeg BA-210 SP

[lowell]

This amp has an issue in the preamp supply. The +16v supply is only showing +4v. I'm repairing that as we speak, however while troubleshooting THAT issue somehow R28 in the power amp feedback loop toasted. I cannot for the life of me figure out why. DC resistance of C17 checks out ok. I belive R28 and C17 are resonance components. If R28 fried and C17 is ok then some AC surge must've killed R28... right?

[Enzo]

That resistor and cap are the amp's stability network, also called the zobel network. Without it, the amp will oscillate. And if the amp goes into oscillation anyway, then that resistor can burn up. The cap is probably fine.

That 16v rail is pretty important, after all there is an op amp in the power supply, and the bias string runs off +/-16v too.

[lowell]

Ok the preamp rails are good now. However still no worky. I found some bad voltages on pins 1-3 (15v) on U5:A, replaced it. There is still about 1.3v on all 3 pins of U5:A. D10/13 seem good w/ the meter but maybe D10 is injured? I believe there should be close to 0v on pins 1-3 right? I also found -15v on pin1 of U1:A so I replaced that too, but the -15v is still there on pin1. I'm not getting any oddball readings on the other opamps. The power amp bias is a bit low at around 10mv across the source resistors but I can't see that being a real issue.

At this point the amp makes a loud hum and some static unless I turn the LEVEL knob UP a bit, then the hum pretty much goes away. There is no signal coming out of the speaker out OR the headphone out. I disconnected C11 which couples the preamp signal to the power amp and the noise is all gone. Any ideas?

[lowell]

Well it's working now. Replaced U2. Had to signal trace to figure out U2 was bad. It wasn't giving me any full on bad readings w/ the volt meter though. Still not sure why U1 has that -15v on pin 1, but the headphone out is working so seems all is well and good. In the future I'll remember that a bad preamp voltage rail can likely lead to killing opamps.

[lowell]

wait a minute... naturally after reassembling this "piece" it starts acting up. There is a real loud static. Open'd her back up to find U2:A is bad again. There is apprx -12v on all pins of U2:A. Pin 3 is 10k to ground w/ J5 disconnected. Pins 1/2 are indeed connected and are 5meg to ground. I believe pins 1/2 should be OL to ground, not 5meg. Could a leaky C52 kill this opamp? Fyi there is still plenty of signal coming out of the amp when played through. U2:B voltages check good and there IS signal out of the amp so my guess is that U2:B is good and U2:A keeps frying. If the DSP circuit is causing the issue I don't know what to do here.

[lowell]

Rg speaks of back to back diodes across opamp inputs, and series input resistor. Maybe I need to add these to u1:A? I don't think nodding this w protection circuitry should be my go to fix. But maybe.

[gtrdon]

This may sound stupid but if the owner of this amp doesn't care about line out or headphone use just remove U1.I have done this to a few of these amps with Loud Tech's blessings under warranty.
HTH

[lowell]

Whoops. I meant U2:A. Enzo, jm, conner, rg? Could use an expert opinion on this one

[gtrdon]

I have had a hard time with these amps.I have found that the 5532 op amp in U1 will fail and cause the voltages to be very low for the rest of the preamp try removing U1 and power up the amp.

[R.G.]

The real problem I can see is that it's a marginal design.

They used an output stage where most of the amplification is coming from an opamp, then used an output stage with gain, and in addition used the opamp in inverted mode because they have the output stage being signal-inverting. This hoses the expected compensation on the opamp, and leaves it prone to oscillating unless everything is ...just...right..., and it's impossible to keep things just right over the long haul. The problem is compounded by using MOSFET outputs which have high gain right out into the hundreds of megahertz. They added C13 around U3b to try to add more compensation; sometimes this helps, sometimes it makes the problem worse.

Burning out R28 is what happens when it oscillates; the cap looks like a short circuit and lets the full output appear across the resistor. This is a symptom, not a cause, just as Enzo said.

A power amp built like this will fry itself on the slightest of provocations. Sorry - it's just that way. Worse yet, it's happy to fry itself if there is any oscillation coming in from the preamp, too.

The problem is compounded by the power supply for the low voltage sections. They used zeners instead of voltage regulators, just like the Fender "de ville" amps.

I don't know what's on the "DSP" board; if that thing feeds ultrasonics back in through U2A, that could set off a suicidal power amp.

If I had to get this thing to work, I would
- replace C52 with a 10uF NP cap, not polarized. They're using a polarized cap at zero DC bias, and that is always bad. I don't know that a bad C52 would necessarily hose it, but no sense in taking chances. I have not looked through the whole schematic, but someone who makes this mistake once probably made it other places. Polarized caps need polarizing DC.
[AGH! Yep, C11 is the same thing. ]
- between R3 (10K) and C53, I would insert a 470R resistor. This will let C53 roll off any ultrasonic junk coming out of the DSP and will not materially change the audio band. Having done this, I'd put a pair of back-to-back diodes across the + and - inputs to U2a. It can't hurt. These are never turned on unless something really ugly comes through the input wires, and the 470R you put in will protect the diodes if something is odd out at the DSP.
- I'd check out the grounding connectors to and from boards. A loose ground to a DSP board can inject all the ground junk on the signal line. Shoot, check out all the connectors. They're probably the silly little insulation displacement types. Good for computers, bad for audio power equipment.

Having done all that, I'd repair the power amp and run it through its paces with a signal generator into C11.

They're playing protection games with Q9 and Q6. Q6 shuts down the signal if the output gets too close to the power supplies, and Q9 shuts down the signal if (something) happens wherever (HPGS) goes. But run the power amp to power into a dummy load, diddle the frequency and see if you get fuzzies on the traces or odd peaks in response that betray a tendency to self destruct.

Only then connect up the reconstructed preamp and see if it will be stable.

All of this is to say - I don't really know what's wrong, based on what you've said and the schemos. But there are several places where it could be going wrong. I've added this to my list of amps never to buy.

[lowell]

Do the back to back diodes have their anodes or cathodes together? Also, what wattage/peak inverse ratings should these have?

[Enzo]

Back to back means the two of them in parallel one pointing each way. So in essence anode A and cathode B wired together at one end and then cathode A and anode B wired together at the other. The idea is that whatever direction the current might flow, it sees a diode.

Plain old 1N4148 diodes will work fine.

[lowell]

Ok and the idea here is that the input voltages will never be more than .7v apart right? This leads me to think that op-amps in a basic audio amp arrangement always have input voltages that never exceed .7v. Is this correct?

I'm led to thinking this because, with back to back diodes, if there was 1v going to the non-inverting input there would be .3v reaching the inverting input right? And, wouldn't this lead to some signal cancellation on the output?

[MarkusBass]

Ok the preamp rails are good now. However still no worky. I found some bad voltages on pins 1-3 (15v) on U5:A, replaced it. There is still about 1.3v on all 3 pins of U5:A. ....
Any ideas?

Lowell,

It looks that you are trying to get rid of symptoms instead of curing the reasons of the problems :-(. You said that +V rail was only +4V and now it's OK. Can you tell us what was the reason? Maybe this would explain all other issues.
Also if there is a DC on output of a IC, I suggest to check what is the voltage on the input. In this case probably the IC was bad but it could be also bad voltage on the input (so possibly previous IC in the signal chain was bad and not the one you replaced).
I suggest that you fix the amp with the DSP board disconnected (simply by unplugging J5 and turning the Effects Level pot counterclock-wise).
In every case I suggest to understand what is the cause of a problem (bad voltage, etc) instead of replacing IC one by one. I replace IC only if I'm sure that it is bad (or it uses a socket so replacing it is very easy).
The problem that I see with the design of this amp is the headphone amplifier. It uses NE5532 without any protection on the output. So if you plug in a low resistance headphones, the outputs are shorted to the ground. NE5532 may not work correctly in such a case or it can even fail. There should be at least some resistors on each output of U1. You can put small value resistors (47 - 100 Ohms) instead of JW51 and JW52 jumpers.
Please note also that the CD input is DC coupled with U2:B - there should be at least some capacitors. Otherwise, if the CD output is bad, you may cause the amp failure.
First I would check that the power amp is OK (with an generator, an oscilloscope and a dummy load), then I would fix the preamp, and finally I would check whether with the DSP module everything is still OK.

EDIT: are both rails 16V, or one is 16V and the other 14.5V?

Mark

[lowell]

Hi mark thanks for chiming in. I fixed the preamp voltages by replacing D11 and C39. I then replaced U2 and the amp seemed great, until 5 minutes after reassembling. When i checked, U2A had DC on on all pins. I guess I should check to see if the dsp is putting out dc. DC on the input of an opamp isn't destructive is it?