Thomas Vox V1032 troubleshooting help?

[R.G.]

Drewl is right. No.

However, let me run these down for you. I mentioned a bit in the Vox Owner's Safety Net that Thomas service schematics contain some mistakes. You're in the middle of some of those.

The component changes are no doubt real, but I suspect they're tuning the values based on "Hey, Fred. We're getting complaints about [insert whatever here] from the dealers. Can you fix [that] with some part value changes?"

My first thoughts on the changes"

C2: 10pf (18pf)

No substantial change. Probably a minor tweak for better RF stability.

R15: 15K (27K)
R17: 100K (330K)
R19: 52R (82R)

These changes are playing with the DC current through the primary of the reverb drive transformer, and probably the available overhead of the driver circuit. This circuit has appeared in a lot of the Thomas Vox amps in one form or another, and the voltages (and hence the implied currents) are different, even though all the Thomas Vox amps that used a reverb transformer used the *same* part number transformer. In any case, neither of the schematic voltages are real.

Georg Ohm shows us some things. If there is 9V across R19, then it has a current of 110ma if R19 is 82 ohms and 173ma if R19 is 52 ohms. Q4 would dissipate between 1.8 and 3W and would rapidly start to glow in the dark. If the 1.3V number is right, we get a much more reasonable 18 to 25 ma, and much lower dissipation.

However, neither the 9V number nor the 1.3V number works with the 11.5 or 15V base number. Bipolar transistors cannot have their base separated from their emitter by more than about 0.7V and remain linear amplifiers. None of the four sets of voltages subtract to about 0.7V, so there is no combination of them that can actually be what is happening. I have the actual numbers somewhere in my simulator, and I can re-run it for the new resistor values. But you can't get that set of numbers from those transistors and resistors.

R15 being lowered provides more available current to drive the base of Q4. Q4 current rises until the voltage on its emitter gets big enough to turn on Q3's base, something above the 0.7V we need for all NPNs. 1.3V might be right, as that number would let Q3 turn on a bit and start stealing some current from the base of Q4 to keep its current from continuously increasing. So the two transistors' DC conditions balance where Q3 monitors the voltage across R19 and cuts back on Q4's base current to keep it relatively fixed. With more current available through R15, R17 was lowered so Q3 could get more base current for eating as much of the increased Q4 base current as needed to keep things stable. R19's value winds up setting the DC current in T1's primary.

But at least three of the four voltages shown for Q3 and Q4 on the two schematics are incorrect.

C14: 470pf (180pf)
Also, there is a 680 cap to ground at the junction of C12/R18/Q4C

Both of these have the effect of wiping some treble off the reverb signal. C14A cuts treble on the primary/drive side of the reverb, C14 cuts it on the recovery side. They thought the reverb was too bright.

So no, these are not your problems. They're minor tweaks to the stage.

[MWaldorf]

Ah, good to know. At least I don't feel like too much of a shmuck for not noticing these earlier.

FWIW, I play surf guitar so I like my reverb bright. So +1 for not having the extra capacitor cutting treble!

So I guess we're back to the question of how having the path to ground on the Q2 emitter going through the e-tuner is affecting things? Even though the values on the RC circuit are the same in both the V1031 and V1032 schematics, is it possible that the V1032 should be different?

I looked back at photos I took before I pulled the original board and C4 was 0.47uF. However, it was a replacement, so I don't know if that's what it had been. R8 is 100r and looks factory. Interestingly, C4 and R8 are in a different location than shown on the layout, as shown in this image. Makes sense since they're shown connected to ground on the layout, and according to the schematic, they shouldn't be.

V1082_board.pdf

Of course, with so many errors in the schematic, I guess it's a toss up which is right?

Don't you just love when someone who doesn't understand what they're talking about speculates away?

[R.G.]

Thomas Organ was faced with a different problem than we are today as far as PCBs. They could barely get PCBs at all, and they sure didn't want to go get new ones made. So they used the same board layout and tinkered with it as needed to make things work. That was possible because they used rows of humans placing parts in the board, not pick-and-place computers. Today, we'd go correct the PCB layout. Then, they figured out how to place the parts a little differently to make it work.

There is speculation that the unused pad area on the "Suitcase Amps" PCB was for an eventual distortion feature that never got added. But it made sense to use it to relocate C4 and R8 so that the E-tuner 2.7R resistor could supply the ground for Q2's emitter. I think that accounts for the location differences.

There is nothing at all wrong with speculating, so long as you know you're speculating. That's one way to learn - do the reasoning and see if it pans out.

[R.G.]

Did some digestion. Nothing that's really definitive, but some suspicious.

W4:0 Good
W5:0 Good
W6:0.4 ACK!! How the devil is a pad with a solid copper trace back to the star ground area showing 0.4V when the other two with
similar traces show 0 ohms? Could be a meter/measurement issue, but is this real? Could be the whole problem if real.
W13:0 Odd, but not a disaster
W14:0 Odd, but not a disaster
W15:0 OK. Should be 0 with a jumper for R55.
W19:0 OK, should be 0
W27:0 OK, should be 0
W28:0 OK probably. Will be 0 if the speed pot is turned to minimum.
W29:0 OK, should be 0
W31:0 OK probably. Will be 0 if the depth pot is turned to minimum.
W34:0 OK, should be 0
W41:0.1r ?? why 0.1, not 0? Meter funniness?
W43:0.1r ?? why 0.1, not 0? Meter funniness?
W45:10r ?? should be about 100R

W6 is downright suspicious if true.

[MWaldorf]

R.G.,

I double checked on my DMM, and it looks like the meter isn't accurate at the level of tenths of ohms. I clipped the leads to each other and got a reading of 0.2 ohms. My DMM is a Uni-T UT58D, which shows accuracy of +-0.8%. So W6 along with W41 and W43 should probably be considered 0.

W45 is definitely reading 10 ohms. The red lead to the pilot lamp is connected to the center tap of the power transformer and the black lead to W45.

Also, I made a couple of videos to show some of the odd behaviors of the amp.

First one shows troubleshooting with 440Hz/40mV signal, with the probe on the input, and various legs of the volume, treble and bass pots and at TP1 and TP2. Even though there's signal being amplified with the volume at zero, the treble and bass controls don't affect the tone of the signal until the volume of the amp is turned up.
https://youtu.be/5HdQoIV3YgA

Next shows the tremolo behavior with the 440/40 signal. With the volume on zero, when the tremolo depth is turned up, there's a new pitch output that doesn't increase when the volume is turned up.
https://youtu.be/CXJqAPyKoek

Finally, here I'm playing around with the E-tuner, tone and tremolo behavior
https://youtu.be/1crWNMTO5cE

[R.G.]

Um, Mel, did you recently offend some minor deity? Or perhaps play a guitar on and ancient indian burial ground?

[MWaldorf]

I'm not sure if I've offended deities (if so, oh, deity in question, I apologize!), but there's an indian burial ground less than a mile away, so I certainly have played guitar in its vicinity!!!

[MWaldorf]

In hopes that I might make penance to the deities of solid state circuit design, I've been trying to learn more about bipolar transistor amplifiers in hopes I might better understand how to troubleshoot my amp, and so far I gather that there are two kinds:

common emitter, where the signal passes from the base to collector, amplifying voltage

emitter follower, where the signal passes from the base to the emitter, amplifying current

The two stage examples I've seen in basic lessons show either two common emitters, with the collector of the first connected to the base of the second, or a common emittor feeding an emitter follower.

However, on the Cambridge Reverb, it looks like Q1 is an emitter follower and Q2 is a common emitter. Is this uncommon? Do I have the whole thing wrong?

[R.G.]

You have it mostly correct. However, Thomas was solving a different problem.

The other characteristics of common emitter versus emitter follower (which is the same as "common collector") are their input impedances. The emitter follower has a much bigger input impedance at the base, roughly the emitter resistance times the current gain, in parallel with the biasing resistors. Thomas used the emitter follower first to get a high input impedance and not load down a guitar signal, leading to treble loss.

The CE stage after the emitter follower runs at a gain of about 6-8, which is low for a CE stage, but suited the tremo effect's needs and the common design of other amplifiers.

I'm still beating my head on what's going on. All the easy explanations for volume being noticeable at "0", decreasing, then increasing again have been tried.

[R.G.]

Mel, I've been thinking some more. Always dangerous, I know.

Do you have a 470uF cap to stick into the C35 position? I've speculated myself into thinking that maybe C35 is needed after all. Shouldn't be, but - hey, we've tried out all the other theories I've had, and it's certain that sometimes I can be wrong about theories.

If this doesn't work, I may go back to the "Indian burial ground" theory.

[MWaldorf]

Bingo! C35 makes the weird volume behavior go away! There's still distortion in the signal, but now the volume control is acting as a volume control should!!!

So, as for the distortion, I noticed some interesting behaviors.

First, here's the signal (440Hz, 50mV) at input 1:


Here's how it looks at the CW leg of the volume pot, with the treble and bass both at 10. The distortion is the same as before.


Then I started playing around with the tone controls. Here's the signal with the treble on 10 and bass on 0. Roughly the same amount of distortion, though the amplitude of the signal seems to have increased from 10mV to 15mV.


Here's the signal with treble on 0 and bass on 10. Much less distortion and much higher output - nearly 35mV


Here's the signal with both the treble and bass on 0. Signal is back down to 10mV, without much distortion.


Here's the signal with the treble and bass on 10 and the MRB engaged. Maybe a little less distortion, but a big jump in output - 30mV


So, there seems to be something amiss on the path through the treble control. This got me thinking - is there anything out of sorts? Then I remembered, when I ordered parts for the build, I mistakenly ordered non-polarized caps for C5 and C11. With everything else weird on the amp, I totally forgot about this until now. Could this be the source of the distortion?

[loudthud]

Do you have a 470uF cap to stick into the C35 position? I've speculated myself into thinking that maybe C35 is needed after all. Shouldn't be, but - hey, we've tried out all the other theories I've had, and it's certain that sometimes I can be wrong about theories.

I'm a little lost here. On the Vox schematic C35 bypasses the +17V Source. There was a mention that the replacement board had some three terminal regulators. (Is the schematic for the replacement board posted anywhere?) Is one of those regulators used to provide the +17V source? The +17V goes to the preamp, the input to the reverb driver after the Volume control and to the power amp driver stage. So signal on the +17 can bypass the Volume control and I guess get to the power amp when the Volume control is all the way down.

[R.G.]

Bingo! C35 makes the weird volume behavior go away! There's still distortion in the signal, but now the volume control is acting as a volume control should!!!

OK, maybe that wasn't such a wild guess. I need to do a bit more thinking before I am sure of why this happens, but hey, one step forward.

So, as for the distortion, I noticed some interesting behaviors.

First, here's the signal (440Hz, 50mV) at input 1:

And that waveform shows visible crossover-notch distortion. That cannot be coming from the preamp stage in this amp, as it is purely Class A single ended, and can't cross over. I think that's in the signal generator at the input.

I think.

So, there seems to be something amiss on the path through the treble control. This got me thinking - is there anything out of sorts? Then I remembered, when I ordered parts for the build, I mistakenly ordered non-polarized caps for C5 and C11. With everything else weird on the amp, I totally forgot about this until now. Could this be the source of the distortion?

Generally, you can put in an NP cap anywhere a polarized one was specified. There are some funny things about different ESRs and the capacitance value as the signal swings from + to - and vice versa, but no, that's not likely to be the issue.

Tone and volume stacks have odd behavior as viewed in a scope trace. There still could be something odd, but if the funny volume effect is gone, try it with guitar if you can. Is the odd distortion still there?

[MWaldorf]

Here's the link to the repair board documentation
Repair Board

R.G. make the board to have the option to build a power supply like the original Thomas Vox design, or to have an updated one that he made with a full wave bridge instead of the diodes, fuses, and a regulated power supply to the pre-amp. I built mine with the updated power supply, in which the RC circuit is replaced by a regulator, diode and different value resistor. It seems that, at least in my case, having the original C35 in place in addition to the regulated supply actually makes a difference.

[R.G.]

I'm a little lost here. On the Vox schematic C35 bypasses the +17V Source. There was a mention that the replacement board had some three terminal regulators. (Is the schematic for the replacement board posted anywhere?) Is one of those regulators used to provide the +17V source?

No, it's not on line. But in general, it's the same as the stock schematic with the exception of those voltage regs. The power supply work on the preamp boards for the entire line - Beatle, Guardsman, Buckingham, Viscount, Berkeley2/Cambridge, Pacemaker and Pathfinder - was given the same treatment. That being, where they used 17-19v, use a 7818, and multiple of those in the "Big Head" amps with more sections. Where +24 was derived from a higher voltage for a signal stage, I put in a 7824.

The "+17" is actually supplied from a 7818 regulator fed from the main DC supply. It comes out as +18, but then equivalent preamps in the series run from +17V nominal to +19V nominal, depending on the amp, and I reasoned that a clean 18V volts in the middle would be OK. It has been so far in other boards.

The +17V goes to the preamp, the input to the reverb driver after the Volume control and to the power amp driver stage. So signal on the +17 can bypass the Volume control and I guess get to the power amp when the Volume control is all the way down.

Yeah. Actually, I think what's going on is that I got fancy in terms of feeding the stages exactly what I thought they ought to get.

I think the real problem is R103a, 68 ohms. It's the equivalent resistance to the voltage divider resistors that fed the +17V supply in the original. But I got it set up so there's no filtering after it with R35 not there, and I think that point was mixing +17V currents and injecting them through the top side of the reverb recovery transistor.

There's a long string of me thinking about why to put that R103a in there, including not wanting to produce a too hifi sound by over-regulating the preamp supply, but that reasoning now seems suspect. I think I might have gotten the same bingo! if I'd said "short R103a". I probably worried too much about the esoterica of getting the power supply to be equivalent instead of just making it cleaner.

And I just got word from an early adopter of the Big Head preamp board that it seems to have much less of what the real vintage amps had - some tiny amount of sound with the volume control all the way down. I think this issue may be endemic to the Thomas Vox Design. We'll see. I'm going to cross-feed him with some of this new-found info.

I figure I can get the final design wrinkles ironed out of Thomas Vox amps about the same time that no one else has one of them to fix, or cares.