Thomas Vox V1032 troubleshooting help?

[MWaldorf]

Well, after many months of being dormant, I've made some progress of sorts with the V1032. As luck would have it, I was able to pick up a generally unmolested V1031 in good working order on CL, which I have been using as a base point to continue to troubleshoot. I re-built the V1032 using the original board and the same spec parts as are in the V1031. For simplicity in wiring, I used ECB 2SC1815 and EBC 2N2219A transistors. Looking at the V1031, there are several components that are different than the values of the schematic. In particular, C14 in the reverb recovery section is 470pf as opposed to 180pf on the schematic, R39, part of an RC filter between Q6 collector and the speaker is 4.7kOhms vs 47kOhms in schematic, R48 on the Q11 emitter is 1kOhm vs 2.2kOhms in the schematic, and R50, in parallel with diode SD1 is 1.8kOhms vs 1kOhm.

With these changes in place I rebuilt the amp and fired it up. I get no sound from the guitar inputs, no reverb and oscillations when the reverb control is turned up. I do get a sound by putting a signal on the volume pot output, and the tremolo works. Hmmm, not exactly promising.

Next I measured the voltages on the transistors. Interestingly, the voltages are almost identical to that I got with R.G.'s replacement board, which suggests that whatever is wrong is likely not on the board. At this point, I took voltage measurements on the V1031 for comparison and found similar voltages on all transistors except Q6 and Q11 - same as before. For Q6, the V1031 reads C:16.1V;B:2V;E:1.45V and the V1032 reads C:13.5V;B:4.11V;E:3.54V. For Q11, the V1031 reads C:26.2V;B:11V;E:9.5V and the V1032 reads C:26.4V;B:10V;E:10.6V.

As the e-tuner had been a source of problems before with a capacitor leg shorting on the reverb transformer, I figured it might be a problem so I removed the e-tuner from the amp. Still the same voltage issues.

Since I apparently have some sort of grounding issue at the inputs, I figured I'd measure resistance to ground at all external connections to the board for both amps. I took the measurements with all controls on 10 and used the same reverb pan and speaker for both. What I found puzzles me - almost every value is different between the amps.

Comparing connection points, going counterclockwise around the board, with the value for the V1031 before the V1032 I measured:
Input from J1: 0 with input closed; >20M when open; same for V1032
Input from J2: 139k open or closed; 158k open or closed
Treble CCW: 6.4K; 11K
MRB Pedal: >20M; same for V1032
Volume wiper: 6.7K; 9.5K
Treble CW: 6.7K;22.3K
Reverb footswitch: 167K; 154K
Reverb CCW: 147R;151R
Reverb CW: 3K; 3.7K
Reverb wiper: 3K; 3.7K
Trem Depth CCW: 742K; 1.066M
Speaker: 3.8R; 3.6R
Q9E: 289R; 449R
T2Sec: 294R; 435R
Q8C: 291R; 447R
Q8E: 527R; 854R
T2Sec: 530R; 844R
+33.5V: 527R; 855R
Trem Footswitch: 19.3K; 22.4K
Trem Speed: 11R; 9.2R
T2Primary Blue: 620R; 975R
Reverb return: 182R; same in V1032
T2Primary Red: 657R; 934R
17V: 1.162K; 1.394K
T1 Send: 667R; 980R

So, there are some substantial differences. One note is that the V1031 appears to have the original germanium Q8 and Q9 transistors where the V1032 has silicon replacements, with R32 and R33 changed from 3.3 to 13 ohms to compensate. Perhaps this explains the differences on the values on the right side of the board.

Any thoughts on what to do next? I'm thinking I can swap the boards from one amp to the other and take measurements. This may be an issue with the R32/R34 values, but otherwise, it should be able to tell me if the problem is on the board or external to the board.

Finally, some pictures! First, here's the V1031 chassis

Next, here's the V1032 chassis

Also, since I couldn't find any photos of the input wiring, here's inputs 1 and 2

And here's input 3

The wire from Input 1 goes to the board; the wire from Input 2 hot goes to Input 3 middle and the wire from Input 2 middle goes to Input 3 hot.

[MWaldorf]

OK, some progress today. There were a couple of bad solder joints which explained the lack of volume at the input. So, now I've got sound, but it's somewhat distorted and not nearly as loud as the V1031. The reverb still motorboats, although there is lots of reverb happening behind the noise. I put a 0.5V 440Hz signal at input 1 and measured the signal at a few different places.

Here's the signal at the input side of R1 - nice and clean


On the other side of R1, lots of distortion. Oddly, it's all on the upper swing of the signal.


On the far side of R2 at Q2E, the distortion phase is reversed.


At Vol CW, it's all a mess


At Bass CCW, it seems half of every wave has cleaned up.


Any suggestions on next steps for troubleshooting?

Thanks, Mel

[R.G.]

Just found this, Mel. I'm in the middle of on eof those periodic life-frenzies, so my time to look at it is limited. I'll try to do a closer look a bit later.

But this kind of oscillation in general says that something is whining somewhere and the preamp is only picking it up on certain voltage conditions and amplifying it then. The change in character as it goes through the circuits may be from filtering in the various parts it hits, or may be just getting closer to the original source of the oscillation. It's the kind of thing that used to be common in pushpull solid state outputs, commonly referred to as "bottom side fuzzies".

[MWaldorf]

Thanks R.G.

FWIW, I decoupled the pre-amp from the rest of the amp by disconnecting one side of C11, and the distortion is still there. In fact it's worse.

Also, I just ordered a copy of your Vox Owner's Safety Net.

[MWaldorf]

I got my copy of the VOSN and read it through this weekend - lots to absorb!

I keep looking for differences between the V1031, which works, and the V1032, which doesn't. I've been getting essentially the same voltage readings with the rebuilt original board in the V1032 as I was getting with R.G.'s replacement board, so I suspect the issue is something off the board. The primary difference between the V1031 and the V1032 that I have is that the V1031 has the original 86-5043-2 output transistors and the V1032 has MJ15016G silicon output transistors. For the silicon output transistors, I'm using 13 ohm resistors for the Rbe bias resistors R32 and R34 (vs 3.3 ohm in the V1031 for the original germanium transistors).

I wonder, could these be the wrong value to use? In the In the V1032, Q6 has incorrect voltage readings - C/B/E of 13.5V/4.11V/3.54V - compared to the readings in the V1031 - C/B/E of 16.1V/2V/1.45V. The Q6 collector is connected to Q8C and Q9E through C22, R39/C23 and C19. Could the bias on Q8 and Q9 cause this voltage change? (shows how little I know about theory, right?)

[MWaldorf]

Holy Toledo - it's ALIVE!!!!!

In reading R.G.'s excellent Vox Owner's Safety Net, he talks about proper biasing when replacing the original germanium output transistors with silicon ones (p122-125). In the V1031 Repair Supplement, he had mentioned a biasing range of 12-18 ohms for silicon transistors (p.2) and in the Cambridge Reverb Repair Board prelim he mentioned using 13 ohm bias resistors (p.8). Not being very well versed in electronics theory, I went ahead and used 13 ohm resistors, and have been dealing with distortion troubles that I couldn't trace. In the VOSN, R.G. describes what happens if the bias isn't set high enough - "This leads to a bad crossover distortion notch, bad enough that every note has a readily discernable "fizz" of distortion, and soft notes may not be heard at all" - well, this seemed to be describing a lot of what I've been hearing. I put in some jumpers to the locations of R32 and R34 and using resistors in series and parallel tried out different bias resistances. It seemed the higher I went the clearer it became, finally clearing up at around 20 ohms. After testing I installed 22 ohm resistors and now the distortion is gone.

The moral here for those like me who aren't well steeped in theory - if you've got silicon output transistors, be prepared that the bias resistance needed will have to be tested, by trial and error if necessary.

I've still got a couple of smaller issues to deal with - the reverb starts oscillating when it's turned above halfway and there's an odd warbling in the tremolo that only occurs with the foot pedal connected and the effect switched off. I'm also trying to get a bigger bump from the MRB. But for right now, I'm just pretty darn excited that the amp is running!

[R.G.]

It seemed the higher I went the clearer it became, finally clearing up at around 20 ohms. After testing I installed 22 ohm resistors and now the distortion is gone.

The moral here for those like me who aren't well steeped in theory - if you've got silicon output transistors, be prepared that the bias resistance needed will have to be tested, by trial and error if necessary.

Nice work Mel. It surprises me that you had to go up to 22 ohms, but you did absolutely the right thing - raise the resistors little by little until the distortion just clears. What output transistors were you using? I may need to dig deeper into how to evaluate the biasing resistors. I was erring on the side of biasing too cold to avoid thermal runaway.

the reverb starts oscillating when it's turned above halfway

Never seen that one. Before the oscillation level on the knob, does the reverb sound right?

and there's an odd warbling in the tremolo that only occurs with the foot pedal connected and the effect switched off.

Try reducing the value of R40 (680). It could be that with a modern transistor in for Q10, it's not being quite shut off into non-oscillation.

I'm also trying to get a bigger bump from the MRB.

I had a guy with a Guardsman that wanted LESS bump from MRB, and he put a resistor in the line to the MRB footswitch. Following LC network theory, that reduced the Q of the L-C and produced a lower, broader hump from it. That same theory suggests that you would want to go to lower resistance to get a higher peak. Problem is, you may be stuck with the resistance of the inductor itself. There is one possibility that changing R10 would lower the loading, but that's kind of a dark horse, and might require tinkering all the values of the tone controls. I'll try to simulate this and see what can be done with the stock values.

It's critical to know - what's the DC resistance of your L1? That's as low as you can go for resistance in the L leg and it limits the inductor's peakiness. To get peaker than that, you'd have to change the inductor for a lower-resistance one. Can you measure that?

Again, good work!

[MWaldorf]

Hey, I think I've got this solved!!!

For the MRB, I played around with different values for C7. Stock is 0.068uF; I find 0.022uF seems to give a nice bump - if I understand it correctly, in an L-C network, the value of C7 affects the resonant peak, and lowering the value raises the resonant peak.

For the reverb, I tried lower values for C13 and C15 - the value of C13 doesn't seem to make a difference, and a lower value on C15 delayed the onset of the oscillation only slightly. Then I tried increasing the value of C14, and that did the trick. I started by added 560pF to the existing 470pF, which totally removed the oscillation, as well as much of the "drip" of the reverb tone. Then I backed down the value as much as possible without having the oscillation. 320pF did the trick.

The weird noises in the tremolo went away somewhere in the process; maybe I had a bad solder joint.

At this point, I'm not sure if I'll put the e-tuner back in the amp. I wonder - could I shoehorn a fuzz circuit in its place on the board?

[MWaldorf]

Hey, another MRB question, this time for the V1031 - on this CR, the MRB is barely doing anything. I couldn't get a reading on the inductor (at over the 20H max on my multimeter). I put in a 1.5H inductor (actually a 4KCT/600CT audio transformer) and now I get a barely audible change. I know the footswitch is working with the V1032 and I tested the footswitch jack continuity so I don't think the problem is there. Any clues where else to look? Maybe trying different values for R10?

[52 Bill]

I couldn't get a reading on the inductor (at over the 20H max on my multimeter). I put in a 1.5H inductor (actually a 4KCT/600CT audio transformer) and now I get a barely audible change.

Did you test the inductor with your ohm meter? I think that the original inductance value is 500mH.

The transformer idea will work, but the one you used may push the center frequency too low for the circuit and the resistance of the transformer coil may be too high to allow for a sharp center frequency boost.

Check the schematic for the correct value and if it is 500mH, use a Vox wah inductor.

[R.G.]

The inductor is 0.5H (500mH). The originals don't fail often, but when they do, they're usually open, high resistance.

If you're not getting any noticeable change, measre the inductor DC value. The transformer core will run the midrange resonance down by the square root of the difference in the inductances. So for a 1.5H instead of a 0.5H, the change will square root of three lower in frequency, or 1/1.732 lower.

If you're not getting a noticeable change, it's likely that the path through the inductor is high resistance - whether in the inductor or the rest of the path to ground.

[MWaldorf]

Interesting. I think the inductor in the big amps is 0.5H, but in the suitcase amps it's 1H. At least it in on the spec and in my V1032. I'll try a Vox wah inductor.

[R.G.]

If you have a measures-good 1H from the 1032, try it in the 1031. Gently.

[MWaldorf]

Good idea. I'll give that a try over the weekend.

[R.G.]

Did a check on the originals. But the schemo and parts list say 1H for the MRB inductor. My mind must be wandering.

Carry on...