SVT III Pro bass amp distorting, low output

[PremiumPlus]

Hi all,
I don't post much but I have a problem that I hope someone with Ampeg SVT III Pro experience can help me with. The amp sounds really bad. First came new tubes, no help. I put it on the bench and scoped the preamp; it's clean. Injecting a sine wave into the output gives a distorted top half of waveform. I adjusted the bias and set it for minimal crossover notch, which helped but there's still a problem. The Mosfets appear to be good; they all check okay as far as I can check them in-circuit. No shorted or open semi-conductors...which makes me think it's probably a dried up electrolytic.
I pulled the power amp board out (what fun) and the solder connections look good.

Is there an achille's heel on this particular amp that you guys know of? Anything specific to check? The thing is a real hassle to get the board in and out of for solder side access, as you know.
I'm thinking that tomorrow morning I'll start by changing a half dozen caps, then put it back in and start tracing again.

EDIT: the output sine wave now looks nice on the bottom and has a hump at the top of the wave.

By the way, I couldn't find a schematic when I was working on it this afternoon, so I was flying blind. I'm an experienced tech, and retired EE, so now that I have schematics I feel a lot better. But I'd really appreciate any input that you all could offer, if you've experienced a similar failure mode.

Thanks!

[Enzo]

The preamp uses all 12AX7s, but the power amp uses a 12AX7 and a 12AU7. Do you have the correct tubes in it? The 12AU7 is in my view the highest failure rate part in the amp.

You say the top side of the output looks distorted. How does it look driving NO load?

When you get the schematic, look at the two small supplies above and below the output main rails. Q3, Q4.

[PremiumPlus]

I've got the power amp board out of the amp right now, after I check the components around those two supplies above the main rails and check a couple of big diodes over on the left side of the board next to some 10 watt resistors that look hot I'll reinstall and check the output under no load.
Now that I've got schematics and waveforms to guide me it's going to be a WHOLE lot nicer to work on. I'll report back later today. Jerrold and Enzo, you have emailed me, thanks guys!

[PremiumPlus]

I went through the power amp board and couldn't find any bad components, I reset the bias using the recommended procedure, and it biased up just fine. The power supply voltages are all good, maybe a bit high but I can't find out what's going on with this thing. The top side of the sine wave is still distorted. It sounds ugly, and again, if I scope the preamp output it's fine. It's in the outputs.
Do mosfet outputs ever fail by degrading in performance? I've never seen it but something is definitely funky. It's kicking my butt and I don't like it when I lose!
I'm done with it for today but tomorrow I'm going to start over, and just follow the signal through the power amp stage by stage and hopefully that will let the problem surface.
Any other suggestions, gentlemen?

[Enzo]

How did you test for MOSFETs? They can short, but they can also open. And when they fail they can take out the 5w 0.47 ohm ballast resistors. Are any of those open? Measure across each one to see if any of the MOSFETs are not conducting current. Any of the 47 ohm gate resistors open? There are four 13v zeners, are any open or shorted?

[PremiumPlus]

The mosfets I checked for by checking forward voltage drop, they all gave a reasonable reading and none were shorted or open. Same with the driver transistors, and I also measured the voltage drop across the 0.47 ohm resistors. I set the bias so the the highest one reads around 36mV and the lowest is about 18mV. No open resistors. The 13 volt Zeners, I'll have to check and make sure, but I did take my Fluke 87 and used the diode test function to look for forward voltage drop. I resoldered all the mosfets, the big caps and power resistors, the molex pins, and any other joint that might have had physical stress.
The high voltage is there, if anything it's higher than spec., like 330 volts where 300 was the spec. I figure that's just line voltage, mine runs 125-127 volts sometimes.
I dunno fellas, I think I'll just try to follow the signal through the amp with my scope tomorrow and see where it leads me.
I do have a little electrical fan motor noise coming through the speaker now that was not there before I rebiased with a target of 25mV across the 0.47 ohm mosfet resistors. Unplug the fan and the noise is gone.

[Enzo]

Does it make a difference between output loaded and not loaded?

[nevetslab]

Does the distortion change with adjusting the Tube Gain control on the front panel? that control varies the plate supply voltage on the 12AU7's. You can also look at the drive voltage from the 12AU7's to where it feeds the MosFET Power Amp output stage. Does that waveform show similar waveform distortion as you're seeing on the output from the amp? The upper and lower transistors do play a role in the drive voltage to the upper and lower MosFET's.

I suppose it's possible the upper rail xstr 2N5415 could be replaced. Does the waveform distortion only happen at high output levels, or at any level? Just wondering about the back-to-back 13V zener diodes from the output buss to the gate drive rails.

I looked thru my service notes from 2014 when I had last serviced the three of these amps in our inventory. Two of them I had pulled the main power amp assembly from the chassis, found and repaired a lot of solder joint fractures...I had done that work first prior to powering up the amps, as I was in a Preventative Maintenance mode going thru our rental inventory. I didn't find any notes shedding light on what you're having.

[PremiumPlus]

Enzo: It's the same unloaded or loaded.
nevetslab: I think the tube gain doesn't matter, I'll check again. I'm going to go downstairs in a bit and get to work, report to follow...thanks fellas.

EDIT: Tube gain control doesn't improve or make the distortion worse.

[PremiumPlus]

Guys, could the electrical noise I'm getting from the fan be related to the distortion issue? The noise varies in volume and goes away completely when the bias pot is moved. But at proper bias, you can hear the motor noise through the speaker, and it varies in frequency with the fan motor speed. If I brake the motor with my finger the noise frequency drops accordingly.
I would think that these are two different circuits, and would not affect each other, but could this be affecting one half of the waveform (the top, where the distortion is)? I'm going to try unplugging the fan and see if that makes a difference in the top waveform "hump".
At this point I'm looking for anything. This one's a dog. Really frustrated, I've been working on amps for decades and every once in a while you get one of these...grrrr.

[Enzo]

Did we ever try a different 12AU7? It has two halves that have to yank the output stage up and down.

Is the signal distorted at TP4?

[PremiumPlus]

I've got a good stock of tubes, and I put ALL new tubes in it. And I tried a couple of different 12AU7s and 12AX7s in the power amp with no effect. Right now the amp's on the back burner, I had to do something for tomorrow's church service, so since the preamp is working fine I took it back and ran the XLR line out to the FOH board for the mains, and took an old Crown DC-300A amp, plugged the 4x10 into the left channel and ran the "Preamp Out" line off the Ampeg through the Crown for onstage monitoring. It sounds surprisingly very good. That amp's a real workhorse, I recapped it 15 years ago and use it every few years, never fails me.

I'll make a note to check TP4 when I bring it back to the bench but that's going to be after Easter. So thanks for the input so far, guys...this one is barking like a dog. I was amazed at how sensitive the bias pot adjustment was. Why did Ampeg use such a large trim pot there? Seems like it should have been a lot smaller so there would be more physical motion per amount of adjustment.
Oh, and that noise I mentioned is now more of a hum. In a way it reminds me of what a tube amp sounds like when the output tubes are going south. But it's isolated to the power amp.

[MarkusBass]

Just to confirm that you have set the bias correctly, can you measure the voltage on the ballast resistors (0R33 or 0R47)? What about a photo of the distorted signal on the oscilloscope?

Mark

[nevetslab]

I'll pull one of the SVT3 Pro amps from our Rental Inventory to have a look at the Fan circuit and see it I see/hear interference noise from the fan's current running thru the heaters of the two tubes V1 & V2. Am I correct in assuming the voltage established across V1's heater in Q17's emitter circuit establishes a constant current source for the for the rest of the circuit? 150mA heat er current, then the fan (24VDC 3.3W Rotron fan would 138mA @ full speed, and it would be in series with the two tube's heater, with it's voltage controlled by Q16/IC1B and the Temp sensor LM35. What would cause the interference noise?.

[tomtech]

this is an interesting thread tome because today I started working on a friends amp (SVT3 PRO) that has similar symptoms. Only on my amp, the bottom half of the sine wave, as viewed from the output, is distorted and low magnitude compared to the positive half. In the schematic I have of the power amp section, the sine wave signal looks healthy and symmetrical at TP2 (R4), but starts to look bad at TP3 (R5). Looks like R4, R5, R6, and R7 are OK. +-100 v supplies good. +-65V supplies good. 310v present for the V11 v12 B+. I dont see that the above thread ever resolved the issue so that I might apply the knowledge to this amp. Anyone have suggestions for this one?