Fixing Thomas Vox Amps

[R.G.]

Any Vox amp fanciers out there? 

OK, this is getting to be a reality. I got back enough reports from beta testers on the rebuild boards to have some assurance that someone can build one of these and retrofit it into a busted amplifier and have it work.
See:http://geofex.com/Article_Folders/Repair%20Boards%20for%20US%20Vox%20Amps.htm
Here's what happens in a Beatle/Guardsman/Buckingham/Viscount:

instead of this:

Another board covers the small combos and Berkeley II.
I feel like a proud pappa.   

I just ordered a 17"x7"x3" Bud chassis base from Mouser to put one of the boards in. I think this will just plug into the effects loop of an amp and give a Vox front end.

On other fronts, a former owner of a Vox UL730 is testing out the preamp section from a UK Vox done on the same kind of PCB. When I get word back from him, it will be possible to do a UL700 preamp front end for other amps.

[mth5044]

That's insane. Very nice. What a headache it must have been to fix the old boards. Are the boards clones of the original circuits, or have you improved the guts as well? I don't see the IC's on the original boards, what's up with them?

How long until we can buy one of your PCB's and make our own? 

[R.G.]

It was so bad to work on the old boards that pretty much no amp tech will work on them. I only know of four techs *in the USA* that will. The general idea was to make "fixing" one of these amps be to clip off all the wires at the controls and jacks, then wire in the new board. No tracing wires through the bundle and breaking other wires by flexing the bundles.

The signal circuits are verbatim the same as the old ones, just new resistors, caps, and 2N5088 transistors instead of the old special-part-numbered 2N2925 types. I did fix the power circuits, subbing in linear regulators (those to-220 packages) for the resistor/capacitor voltage supplies in the originals. The ICs are CD4052 CMOS switches that replace the old hard-wired 4P3T switching so that the signal doesn't have to go off the board on six wires to get to the switch, which was its own nest of wires.

But except for that, it's the same circuit schematics.

I was holding off on letting anyone use them until someone other than me had taken a board, built it, and successfully put it into an amp. But we're there now. The boards work. I have about half a dozen from the first proto batch.

[tubegeek]

Thank God you use your super-powers for good, 'cause surely you could have gone the other way, Tex.

[mth5044]

Hey RG, I was checking out your TV Amp Repair Doc (GEOFEX) and am wondering why you chose three 7818's for regulation, reserving one for the normal and brilliant preamps, one for the bass and reverb recovery (which goes on to feed the switching chips), and one all for the mixer/limiter. It looks like you are using the TO-220 package, so you'd have 1A on tap, which seems like it would be enough for a handful of transistors, especially with the reverb driver running on the 24V.

Was this specifically because you didn't want to have to use heatsinks? Or possibly because it was easier to lay out, the parts are half a dollar each, and you had PCB room to spare?

Thanks for a peak into that brain.

[R.G.]

It was a compendium of those, plus some star grounding. 

I did start with TO-92s and was reminded when I started looking up parts that the TO-220 and the TO-92 voltage regulators are very nearly the same price. Ordinary three terminal regulators don't need a minimum load, so why not use the higher-power parts and let them loaf.

The question behind that is why one? I originally split the power sections for what may turn out to be silly reasons. The original Thomas Vox amps had individual power and ground domains, each star-wired back to decoupling caps. I did on-PCB star grounding for the Visual Sound Workhorse amps, and it worked really well. They were as quiet and hum free as any tube amp I've ever heard, so I knew it worked. I wanted to make sure that the works-like-a-Vox boards didn't have grounding or interference problems, so I checked to see if the extra board area for separate domains for each circuit section would fit; it would, so I did.

In retrospect, yeah, it's overkill. It is likely that one 7818 would have done it. But three lets you de-power a section during debug to see what's wrong. It was handy and more reliable, and only $1.00 more in parts for a big board, so I went with three.

Eliminating heatsinks was a welcome side effect. I actually laid the board out for heat sinks for all the TO-220 parts, just so there would be space. Then I tested the temperature rise on them without heat sinks on the first populated board. Only the 7824 really needed it, and that was marginal. The reverb driver(s) were even cooler, but still warm, so I suggest that any builders use sinks there. The recommended sink is $0.23 in ones at Mouser.

As a bit of review, there are now about 8-10 (I'd have to go count) of the Vox preamp works-alike boards out in the world now. They all work properly, sound like the original, but I get a comment from each builder - there's none of the hiss I remember.     I love it! 

[R.G.]

I forgot to say - checkout period is over. I'm pretty sure they work as designed. I got one trace in the wrong place, and reversed a TO-220    but they do work. The UL700 preamp boards  work too.

I got sidelined on documentation. Originally I did the docos for a board as simply as possible, just parts list, schematic, and some helpful advice. As the first adopters fed back info, that grew to about 85+ pages  per board. Then I started getting commentary that this or that didn't work after the board was in the amp, and it slowly dawned on me that the questions on "well, how do I get this... to work if it's not the board?" meant that I needed to write up "how to fix the rest of the amp, too". Grew to about 180 pages.

At this point, I thought, why not tell every Vox owner how to fix the rest of the amp, as not all the problems were related to the repro PCBs, so I word-processed out the "how to fix the rest of the amp" into what has become "The Vox Owner's Safety Net", being a compendium of what I know about Thomas Vox amps. That ran to about 120 pages last count. And I separated out the info specific to an amp into a number of "technical supplements" more specific to one amp per supplement than the general stuff in the Safety Net. So someone who wants to get/keep their Vox amp running can get just the general stuff in the Safety Net and the supplement specific to their amp without wading through the data on the entire Vox line.

So right now, there is the two variations of the replace-the-preamp PCBs, the V11n1 series and the V11n3 series and about 60 pages per on how to get the PCB populated and running. There is the Safety Net, and (quickly counting on fingers and toes) twenty-something Technical Supplements, one per model. While I was at it, I added in data and supplements for the "Suitcase Amps", the Cambridge Reverb, Berkeley II, Pacemaker and Pathfinder, and the Berkeley III, as well as some of the bass amps.

I got blisters on my fingers from typing.  Well, not really, but I did get a little stare-y eyed. Some very brave people did the proofreading of the drafts, and I'm near being able to make the whole mess public. I just want to get there before any more of them go into dumpsters!   

[mth5044]

That's great news! Congrats on the successful rework PCB's. I'm sure the last few decades have seen a lot of these amps bite the dust, hopefully you'll be able to save the remaining few. I look forward to reading the documents you've blistered up for, if you happen to publicize them. This is clearly a labor of love for you (since I've been reading posts you made about them from pre-new millennium), glad to see it coming to fruition.

While I could ask a dozen more questions, I'm sure they will be answered by the documents (if and when they come out - I'm in no hurry), but I do have one question for you regarding a post you made on Vintage Amp forum that is likely not in the documents.

I spent part of the afternoon working on a Cambridge/Berkeley equivalent. The reverb drive and the power amp drive were the only real hitches, as both of these have driver transformers in the original device. I put in two LM1875s, one for reverb drive and one for the power amp. Some techno-tweaking on the 1875 gives a power amp that's softer clipping than a typical solid state amp and also higher impedance to loosen up the grip on the speaker a bit and let some of the speaker voice show through, somewhat like a tube amp with its dismal damping factor does.

I'll probably have something usable for prototyping in a couple of days.

Had you ever revealed what you did to the 1875 for the poweramp, your 'techno-tweaking'? I see you've already went with a transistor based reverb driver in the preliminary documents, but I can't seem to find any more posts by you about the LM1875. This isn't directly related to the TV amps, but you had mentioned it worked well for them.

Danke

[R.G.]

It's a mixture of the voltage feedback that's in most power amps and current feedback to lower the output impedance. The technique predates Rod Elliot's write up on the technique, but his is one of the more lucid write ups. See Elliot Sound Products web site for the article on raising the output impedance of a power amp.

What you do is use a resistor in the return side of the speaker and feed the voltage from that back into the inputs of the amp. It changes the rock-solid voltage source nature of a modern feedback power amp into a part-voltage source, part-current source. Speaker eccentricities shine through better. The trick is used in many current solid state guitar amps, including ones from Fender and Marshall.

I also put a soft clipping network in front of it to keep the power amp itself from hard clipping. The Thomas Vox amps did this in the late 60s with their "watchdog power limiter", probably to save their power amps from dying, but it happens to sound good as well as preventing the archaic power output stage from going into meltdown.

The LM1875 for the reverb stage didn't need these. It was a one-package way to implement the solid state reverb tank driver that was used in the later Thomas Vox V11n3 series, notably the V1143 Beatle.

[R.G.]

Oh and it's OK to go ahead and ask. On something this big, it's always possible one of your questions is something I should have put in the docos and didn't think to do. 

[PRR]

> The technique predates ....

I believe it is in Black's original write-up of corrective feedback, though not specific to loudspeakers.

It is in early 1950s Fisher hi-fis, and as a user control.

I just noticed it is in the Shure VocalMaster, however using a transformer(!) to sense current.

http://shurevocalmaster.com/documents/Shure_Vocal_Master_pro_va300%20schematic.pdf

I guess they thought the speaker jack needed to be *grounded*. (Also Shure had in-house transformer facility.)