Ampeg V4 distortion model service/repair

A friend of mine bought an Ampeg V4 declared defective a while back for next to nothing and after gathering some more dust and dirt it has finally ended up on my workbench.

It is a distortion model V4 and I have never worked on one before. As there isn't a lot of information about this model on the web, and it doesn't seem to get the love of the earlier models (it's mostly the distortion people seem to dislike), I thought I'd do a write-up of the work I did and list some of the problems I encountered.

As you can see from this second picture, the serial number tag/sticker is missing, and unfortunately the one inside the cabinet has faded to the point where it is unreadable - there is no telling exactly when this was made.
The schematic on top of the chassis is revision D

Ok, so some words of caution before we look at the guts of this amp.
I play guitar, I sing, I tinker with tube amps and other musical electronic gadgets in my spare time. I'm also an electronics engineer. While I'm old enough to be opinionated and not care what anybody else thinks about me, I came to be an electronics engineer rather late in life and vacuum tube technology was not on the curriculum. I know only the very basics of vacuum tube tech and do not pretend to understand everything going on in this circuit. I don't do mojo though. If this were my my amp, I'd replace every capacitor and carbon comp resistor with polypropylene film capacitors and metal film resistors just for reliability sake. Would that alter the tone? Sure, but so would replacing them with whatever is the current hype (paper in oil anyone?).
What we perceive with our ears is highly subjective and more often than not very dependent on how much we paid for some part/piece of equipment.
I am however not immune to provable facts and science - when I'm wrong, I'm wrong. It happens (often enough) and I welcome factual discussion and a chance to learn something new.

So let's go.

Some shots with the covers removed.

Right away we know someone's been in here before. A couple of parts on the wrong side of the pcbs. That large red capacitor, C21, the replacement was too large to fit on the other side. The other 2 capacitors, two 220nF in parallel to replace C9 (470nF) - really? The 2 resistors R8 and R38? No idea why they're on this side of the pcb. Also, what's up with the tinned traces? Yes, that fuse also belongs on the other side of the pcb.

The other thing that immediately catches the eye is the major amount of grime inside the chassis. You'll notice it is more pronounced on the left side of the chassis (there's a spot on the power transformer where a wire has rubbed it off). You'll also notice that C301 (the left-most of the power supply filter capacitors in the image above) has been replaced. I think it's safe to assume that the grime is the residue that precipitated when it went up in smoke.
Regarding C301, the negative terminal on that capacitor has half of B+ on it. The way the replacement is mounted is unsafe. The outside of the can (negative terminal) has been taped up with gaffer tape, except for the top...ouch if you accidently touch it.

At this point I should explain that the goal for now is just to get the amp running so my friend can decide whether he wants to keep it and play it regularly. So for now, I'll leave everything not broken alone, including noise and hum issues.
He also is a bass player, so I'll probably be looking to change some coupling caps to larger values later on.

There are also a few resistors and capacitors that look cracked but I leave everything as it is for now

The tubes. Oh my. 2 power tubes were missing. The 2 left in there were russian. One was a 6Π3C-E (6P3S-E), the other was unreadable but looked identical except for a slightly taller glass bulb. The printing on the preamp tubes has become unreadable except for a 12AX7 for V1 (wrong) and a 6CG7 for V3. V201, which should be a 6AN8-A (triode and pentode) was an unidentified dual triode. The others V2, V4 and V5, unknown dual triodes.

Except for the 6AN8-A I had working tubes to test the amp with in my own supply so at least I would be able to verify the preamp before having to order new tubes.

First test: After checking the fuses, power it up without tubes and measure all the supply voltages. All was well. I did resistance measurements on the output transformer, all's well here too. I give my friend a call tell him the iron's fine and that alone is worth more than what he paid for it.
I did replace the impedance selector switch as contact was flaky. Cleaning could probably have solved the issue too and I might still do that and put it back in.

Next: Clean the tube sockets, plug in the proper preamp tubes and fire it up again. Measure plate voltages, cathode voltages. Some are ok, some are not. This is were the fun starts.

So I start with channel 2 (notice how I'm already biased against the distortion channel), V2's voltages are off. Cleaning the PCB, reflowing the solder, measuring part values. I find nothing wrong. The tube socket looks suspicious though, re-tension the tube socket, we're in business - or so I thought.
I'll jump ahead at this point and say that I only found 2 non-functional parts in the amp, a cap and the 100mH inductor of the mid tone control.

Back to the tube socket. Re-tensioning only solved the problem temporarely. After 2 or 3 tube insert-remove-cycles the problem was back again. Taking a closer look at the socket the contact material seems very soft and brittle and on top of that, my re-tensioning efforts have left them scarred.

Time to replace the socket.
This is not for beginners (soldering-wise) as 1970s pcb technology was not what it is today. Add to that the 40+ years of age and you'll be lifting pads and traces in no time. The important thing is not to keep the heat on too long.
If you don't mind destroying the old socket go ahead and dremel off each pin right underneath the socket. That way you can then desolder and remove each pin separately. You'll need to remove a couple of the surrounding parts if you're going to go that way.
I wanted to keep the socket so I removed most of the solder with a solder pump and the rest with a desoldering braid. Be especially careful with the desoldering braid or you'll pull the pads right off the board. After that I was able to remove the socket by carefully wiggling it out.

Above: work in progress before the socket change. PCBs partly cleaned, no reflowing done yet.

With a new socket soldered in the voltages were now good and stable for channel 2's first stage. I continue in this way, cleaning, reflowing until I'm satisfied that all the preamp tubes' operating points are stable and where they are supposed to be. V4 socket also had to be replaced.

Intermission:
Voltages on V3 were way off too. Specifically for the second triode. There's no plate voltage indicated on the schematic, but our good friend Georg Simon O. tells us that 0.6V across 330Ω (R22) equals a current of 1.82mA (which in hindsight should have seemed low right away). For simplicity's sake we multiply that by the value of the plate resistor (R21) and get a voltage drop of ~18V across it. Starting with a B+ of 352V we get a plate voltage of about 330V - hmmm? So I double checked the resistors and capacitors and for good measure, I replaced the tube socket - I'm getting really good at it by this point.
All of this didn't change anything and neither did trying 3 different 6CG7 tubes.
I brooded over this for a bit and while comparing schematics of other V4 models, I noticed that while the circuitry for the reverb driver is almost identical, the cathode voltage is not. With a B+ of 400V and exactly the same plate and cathode resistors, an early model V4 has 5.6V on the cathode (and 270V on the plate) of the second triode of the 6CG7. This was much more in line with what I measured. Obviously one of these schematics had a wrong voltage indicated and I believe it's the one for the distortion model.
Oh, and another thing, not a mistake as such, more of an inconsistency. C305 is marked on the schematic as 20µF, whereas in this particular amp, it had a dual section 20µF/20µF axial electrolytic capacitor with both sections paralleled - ergo 40µF. And this was the original cap for sure.
Intermission end, on with the main item.

Time to fire up the signal generator and oscilloscope. Ampeg has graciously provided some AC voltage readings on the schematic. I put a 10mV RMS sine on the input and start tracing the signal with the scope. I see low signals pretty early, right after the first stages. Channel 1 has no signal at all after the volume control. Time to do some more cleaning. This time the potentiometers and switches on the control pcb.
The solder joints on R104 (channel 1 volume pot) are cracked, so no connection there. I clean the pots and switches with contact cleaner and reflow every solder joint on the control pcb.
Much improved levels after more signal tracing.
Something is still wrong with the mid tone control. As mentioned earlier, I find the 100mH inductor is open so I replace it. I also find C8 to be open. It looked bad, discoloured from the get go and the signal tracing confirmed it.
While I'm doing this, I also replace (for space reasons where necessary, C9, C21) and move the parts from the solder side back to their original positions.

At this point the preamp looks like it's working fine. I hook up the preamp out to an amplifier and play a bit to have a listen. Everything works, no scratchy pots or switches. I'm sure that the amp will run fine with a full set of new tubes - it's time to order.

Let me just add here that the distortion indeed sounds "strange" and I don't like it at all. It sounds like a clean tone with some fizz added on top. Of course, I don't even know if this is what it's supposed to sound like. I don't think all the hate is justified though since you can just turn it off if you don't like it.

While I'm waiting for the tubes, I continue the work and clean and reflow the second pcb with the driver and phase splitter (V201) on it, and reflow all remaining off-pcb solder joints. I also clean the powertubes' sockets.

Above: My work is done, for now. I did not clean the chassis as the only sensible way to do that would involve ripping everything out.

Some thoughts while we're waiting for the tubes to arrive. I'd recommend replacing all electrolytics with modern equivalents. Some improvisational skills will be needed to install them as they will be much smaller (but you need more of them) than the original parts, but they will be more reliable. I'd also recommend replacing the coupling caps with polypropylene film caps. They are small and can stand up to the heat inside the amp very well. The ones that are in the amp are all kinds of different sizes and types and a lot of them do look bad, discoloured and even cracked.
While I found only 1 dead capacitor in this amp and the rest were not leaky, I did not measure capacitances and they are bound to have drifted at least somewhat (in addition to their most likely higher tolerances than todays equivalents) - and the same probably goes for the resistors.
One more thing: when I say reflow, I mean I remove most of the solder with a solder pump and apply fresh solder.

Tired of reading? The tubes have arrived so here are some more pictures:

Above: Channel 1 Output measured across an 8Ω dummy load with a 10mV input signal, volume at 12 o'clock, distortion all the way down.

Above: Channel 1 Output measured across an 8Ω dummy load with a 10mV input signal, volume at 12 o'clock, distortion at 12 o'clock.

Above: Channel 1 Output measured across an 8Ω dummy load with a 10mV input signal, volume at 12 o'clock, distortion dimed.

Above: Channel 1 Output measured across an 8Ω dummy load with a 10mV input signal, volume at 4 o'clock, distortion at 2 o'clock - output on the verge of clipping ~94W.

Above: Channel 2 Output measured across an 8Ω dummy load with a 10mV input signal and the volume at 12 o'clock.

Above: Channel 2 Output measured across an 8Ω dummy load with a 10mV input signal and the volume at 3 o'clock.

Above: Channel 2 Output measured across an 8Ω dummy load with a 10mV input signal and the volume dimed, ~57W RMS.

Above: R306, R307, R308, measurement on R306.

Above: Solder side around V3, that is why the pcb is discoloured there.

Above: Component side of the preamp pcb - V3 working real hard.

Above: Component side of the poweramp pcb - hottest components are R204, R206, R209, R217, R219 and R220.

Troubleshooting help: chassis layout and part designators for the poweramp pcb solder side, your mileage may vary (also, see note below).

Note: There's an inconsistency between schematic and layout, from the schematic grid resistors R213 and R221 are fed from R211 and R214 and R222 are fed from R212. On the pcb R211 feeds R213 and R222 and R212 feeds R214 and R221. Consequently there is no "correct" way to label the plate resistors.

With all the tubes in, another problem pops up: microphonics - located on the poweramp pcb. My first guess was the 6AN8A of course but after a lot of tapping I'm not so sure. I can tap on the pcb anywhere (and the chassis around it for that matter) and hear it clearly through the speakers and, at least subjectively, tapping the tube does not produce the loudest noises (hard to say though, as depending on where I tap it sounds different). It might be (a) bad resistor(s), capacitors or even the pcb itself. Was that why some of the traces on the pcb were tinned? In any case, there will be some more troubleshooting to do.

Microphonics update: We received a replacement tube from the supplier under warranty (which is nice) and this one was specified as low noise and low microphonics. The issue has definitely improved (so I'm almost sure that tube is the culprit) with a different sound/noise and much lower volume when tapping the tube or chassis. How this will hold up under high(er) volume conditions I don't know yet.
I fear that in order to get this amp gig-worthy it might be necessary to spend a lot of money to find a 6AN8 that has very low microphonics.
Btw, the first tube we got was a JAN Sylvania 6AN8A and the replacement was an RCA 6AN8.
Tube manufacturers take note: We'd like a current production 6AN8(A)!

Final thoughts: There was nothing seriously wrong with the amp other than wrong tubes, lots of grime, cold solder joints, bad sockets and 2 dead components. To service it properly I would exchange all the power supply filter caps and most likely the coupling caps as well. And I would also replace the electrolytics in the bias supply circuit. The amp is quite noisy, lots of hiss, especially the reverb, so there's bound to be some more troubleshooting to be done should my buddy decide to use the amp. There's also a lot of visible corrosion on the input jacks so I guess I would replace those (and most likely every other jack as well). And lastly - purists avert your eyes - I'd install a master volume.
The supply voltages measured were all a bit low (at 229VAC line voltage) with respect to the indicated values on the schematic: A: 524VDC, B:320VDC, C: 313VDC, D: 287VDC, E: 338VDC and F: -31.4VDC. Since the line voltage range for the export model is specified as 200-260VAC, I guess those values are ok. It's possible that with new filter caps those voltages will rise a bit.
I'll say Ampeg made some improvements over the earlier model V4. I like the separate power transformer windings for the screen grid and bias voltages. They also perforated the chassis top cover for improved airflow and they got rid of the 6K11 triple triode for the mid tone control. Also of note is that contrary to the earlier V4 this one has no negative feedback from the output transformer secondary. They did plan for it though, R210 and C206 are on the layout (not the schematic) and connect to the junction of R202 and R203 which are the cathode resistors for the driver tube. Whether any of these changes translate into a better sounding amp is up to your ears.
Another thing: the way Ampeg built the plate supply with the 2 electrolytics in series and the balancing resistors across them, well those will also bleed off the charge in the filter caps once the amp is turned off. Because this model has a separate screen supply winding, the charge from that capacitor (C302B) won't bleed off and I've measured voltages close to 300VDC on that cap after several hours of not being powered on. Make sure to drain that capacitor before working on the amp.
As for the sound, I haven't had a chance to really play it at a decent volume level (plus there's still work to be done) so come back for updates and maybe even sound samples.

That's it. For questions, remarks, corrections or discussion: rook-e@gmx.net. Hope you enjoyed this, stay safe and don't electrocute yourself.

Tools used: Wavetek Model 184 Signal/Sweep Generator, Tektronix 2430A Oscilloscope, Fluke 87 Multimeter, FLIR TG167 Thermal Camera.

Last updated 2020-03-18.