Thomas Vox Repair Info

Vox and Hiwatt Discussion

Moderators: pompeiisneaks, Colossal

Post Reply
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Thomas Vox Repair Info

Post by R.G. »

I just got some more of the Thomas Vox repair books on line. They're at TheBookPatch.com, and/or thomas-vox-repair.com.

The list now includes:
- Vox Owner's Safety Net
- Vox Footswitches
- V1141 (Super Beatle) Repair Supplement
- V1181 (Westminister 60W) Repair Supplement
- V117 (Sovereign) Repair Supplement

I put them up as I get time to finish off the similar/different from amp to amp in print format. The Vox Owner's Safety Net is a compendium of how the sub-circuits work across the entire line. The others are expanded details of the individual models. Depending on your skill level, you may not need the overall how-it-works, or you may only need the insight, but not the details.
User avatar
rooster
Posts: 1616
Joined: Wed Apr 02, 2008 8:50 am
Location: Pacific NW

Re: Thomas Vox Repair Info

Post by rooster »

Wow, R.G., you are on the hunt. Sure, I'd like to see these. Q: Are you finding that all the various transistors are still out there? Or do you think there are better replacements so it's not a problem? The reason I ask is that I just revamped/rebuilt a Neve 1073 circuit and I tracked down all the NOS transistors and find that the preamp is quite remarkable, better by far than it was with the newer parts in it. It's really made me think about the quality and the mojo of things old. And contrary to my behavior/results, there's a very competent outboard gear guy who's a wizz with rebuilding things that used newer/better/faster/quieter transistors who said that when he was done he just didn't understand why ppl like the 1073 circuit. :shock:

In fact, does the Thomas Vox amp use a Motorola 2N3055 output transistor? Probably so. I found those up in Canada and glad I did. Hope things are good, R.G.
Most people stall out when fixing a mistake that they've made. Why?
R.G.
Posts: 1223
Joined: Tue Dec 02, 2014 9:01 pm

Re: Thomas Vox Repair Info

Post by R.G. »

Sorry I missed your post for so long. I've been hit-and-miss on my network work lately.
rooster wrote: Wed Apr 05, 2017 2:08 am Q: Are you finding that all the various transistors are still out there? Or do you think there are better replacements so it's not a problem?
Maybe - with a couple of exceptions, and some philosophizing. The whole "original part numbers/makers/NOS" question is complicated enough to be more typing that I can do at one sitting.
1. The original transistors are not still out there; Thomas Organ did not make transistors, they bought them from the few companies that could make them, back in their day. But as was common at that time, they did not buy the manufacturer's ordinary "2Nxxxx' parts. They bought "house numbered" transistors, (presumably...) selected to their own special specifications. So no, barring finding an accidental find of a dusty shelf bin of 85-5054-2 transistors somewhere, the original parts are not still out there.
2. The original Thomas-blessed replacement industry numbers are still out there. Even Thomas didn't supply the original transistors in all cases. They did send some shops supply kits of replacement transistors, but they also sent repair shops their list of commercial replacements to buy. I have posted an enhanced version of this list d on geofex, and also list suggested replacement part numbers in each repair manual. In most cases, with some asterisks and some ACK! exclamations, yes, you can buy industry-suggested part numbers matching the Thomas-suggested commercial replacements. And there are a number of dusty-bin sources of transistors with the nominally-correct numbers stamped on the part. That may or may not help because...
3. A 2N2925 is NOT just a 2N2925. Transistors are better thought of as a datasheet, not physical devices. Dataheets define a range of a huge number of parameters - voltage, current, power dissipation, gain, frequency response, switching times, RF parameters, it goes on and on. Just the gain specs are usually at least a 3:1 range, and could well, in the 1960s, be a 10:1 range. To the transistor manufacturer, any piece of silicon he cooked up that met the min/max specs on the datasheet could be and was sold as that part number. In fact, it was common to cook a batch of silicon dies, test them, and dump them into bins for different part numbers on test. And to dump the same dies into different bins, or dip "2N2925's" out of different bins if they met the datasheet specs and a big order was in hand. The house-number movement was one reaction to that, to try to get less-wildly varying parts to manufacture with. But Thomas, like most electronics manufacturers, did not 100% test in coming parts to spec, relying instead on the transistor makers' guarantees.

And today, you CAN buy many of the transistor type numbers that Thomas specified as commercial replacements. But with 50 years of improvement in semiconductor ovens, I guarantee you that you cannot buy any new manufactured transistor as wildly variable as the ones from the 1960s. We (humans, that is) are simply dramatically better at cooking up consistent silicon. You're still stuck with anything out of the oven that meets the specs on the datasheet being labeled with the part number, but the parts you get tend to be either more tightly clumped or to dramatically beat many of the original specs. Usually especially frequency response and noise.

4. You simply can't get some of the original part numbers, even newly manufactured. This is especially true of the JFETs Thomas used in the V11*3 series. The whole semiconductor industry is moving away from discrete JFETs entirely. You can get replacements of a different type number that can be made to work, with some tweaking, and the repair manuals talk about that tweaking.
The reason I ask is that I just revamped/rebuilt a Neve 1073 circuit and I tracked down all the NOS transistors and find that the preamp is quite remarkable, better by far than it was with the newer parts in it. It's really made me think about the quality and the mojo of things old. And contrary to my behavior/results, there's a very competent outboard gear guy who's a wizz with rebuilding things that used newer/better/faster/quieter transistors who said that when he was done he just didn't understand why ppl like the 1073 circuit.
I am certain that both of these are 100% true. However, I'm also sure that neither of them indicate a fundamental truth about replacement semiconductors.

The reason I think that way is that as a design engineer responsible for designing circuits to be reproduced in the thousands, I had to do designs that would work with ANY "2Nxxxx" transistor the uninformed purchasing department bought, generally from the cheapest seller. I had to be a student of the ways that parts vary - how much each parameter varied, and how the performance of the whole circuit board varied if every stinking part on the board was either at minimum specification or maximum specification, or a mixture of mins and maxs. The number of possibilities of ways the parts' tolerances could interact was explosive. So those of us who worked in this snake pit had the rules of dealing with variation in parts beaten into us. Getting something to work was viewed as no big deal; you can get one of anything to work. But can you get ten to work with unselected parts from different manufacturers? Now do ten THOUSAND copies, and if the fallout is over 3%, you're fired.

We really quickly learned that one observation of a good result was interesting, but not totally reliable. It was a show-me-the-numbers game. I completely believe your Neve rebuild was great. If you did 200 such rebuilds, half with NOS parts and half with modern replacements, and the 100 with NOS parts showed better results on documented testing, I'd say you're really on to something. Note that this is not a comment about you or your work - it's a story about the nightmares *I* have over part variation. :shock:

As they say, education is what you have left when you've forgotten how you got that scar. :D
In fact, does the Thomas Vox amp use a Motorola 2N3055 output transistor? Probably so. I found those up in Canada and glad I did.
I don't know that Thomas did use Mororola for all their specially-selected 2N3055s. I do know they told their repair centers NOT to use certain brands of 2N3055 for replacements; I have a copy of the tech bulletin. I also know that modern 2N3055s are better and more consistent than old ones; and that they're often counterfeited; and that the NOS datasheet for the 2N3055 from Thomas' day showed that the 2N3055 was just barely capable of working in a +/-32V power supply amplifier and surviving. And that On Semi (who is the direct successor to Motorola's power transistor division) lists the 2N3055 and the MJ15015 on the same datasheet these days, with the MJ15015 having the spec for the much higher voltage and safe operating area to live in a power amplifier setup. That makes the answer clear to me - use the MJ15015.

I know my views are largely heresy in this "if it ain't old, it's cr@p" era, but I have a very large bias toward "if I can't rely on it to work on stage every single time, I don't want it" as well. And I realize that this is one of my personal failings. :lol:
User avatar
rooster
Posts: 1616
Joined: Wed Apr 02, 2008 8:50 am
Location: Pacific NW

Re: Thomas Vox Repair Info

Post by rooster »

Wow, I just read this last post RG. Thank you for the info!! The 2N3055 replacement is something I need to try.....
Most people stall out when fixing a mistake that they've made. Why?
Post Reply