Vox Buckingham Amp - Reverb Transformer Issue

[R.G.]

No lasso-ing needed.

I have to admit - I HATE to-39 "power" devices. The TO-5 package was "enhanced" to dissipate more power to make the TO-39, with mixed results. The TO-220 was designed from the start to be a power device. I personally would never choose the TO-39 for a several-watts position, but I am biased, and I realize it.

@OP: if you get the TO-39 device and you trust it not to fail, go for it. Me, I go for massively higher voltage-rated devices from new-manufacture and new semiconductor-process fabs. It's a personal quirk. I ... LIKE... newer, better semiconductors. 

[Rob Strand]

I have to admit - I HATE to-39 "power" devices. The TO-5 package was "enhanced" to dissipate more power to make the TO-39, with mixed results. The TO-220 was designed from the start to be a power device. I personally would never choose the TO-39 for a several-watts position, but I am biased, and I realize it.

I don't see it as a bias.  TO-5's are only good for 50 ohm output impedance function generators.   If you don't use enough heatsinking with TO-5's some output stage designs with go into thermal run-away.   TO-126 packages are a little better.  IIRC, some modern TO-220's are better than some of the old TO-3's.

[Frank_NH]

Thanks R.G. and Rob.  I should be getting my 2N2219A transistor today and will report back if it fixes my problem or not.

Meanwhile, I've been doing other refurbishment on the amp, including rebuilding from scratch a new back panel (photo) as the old one was completely trashed.  I used MDF and wood for the plate and covered it with vox style black tolex.  The odd shape made it a little tricky to cover but in the end it came out pretty well.  Now just need to install attachment screws, the three pin speaker connector, and the old nameplate. 

[R.G.]

Very nice. Consider putting a 1/4" phone jack in it. Just sayin'...

I did mechanical drawings of the "big head" woodwork cabinets for the Thomas Vox line somewhere. I always intended to build at least the head of a unit.

[Frank_NH]

Success!  Installed the new transistor tonight and I finally now have reverb in the Buckingham.    So in the end, it was the driver transistor and not the transformer (thankfully).  I couldn't use the old heat sink so I'm going to get another new one for TO-39 transistors.  The top of the transistor gets quite hot after some use, so I can see why it could fail after a while without adequate cooling.

I checked the TO-39 driver transistor in the power amp, and the voltages check out OK so I'm not going to touch that for now (even though I have a bunch of 2N2219A transistors from my eBay order...).

I'll also solder in the zener diodes before reinstalling the transformer in its hideaway home in the preamp chassis.

I'm still tinkering a little with the peak limiter and understand now how to set it.  One thing I'm going to add while I'm at it is a capacitor at the output of the peak limiter so as not to have DC voltage in the shielded cable between the preamp and the input of the power amp.  I noticed that the 1143 version of the Beatle amp had this mod, which makes a lot of sense to me.

[Rob Strand]

I couldn't use the old heat sink so I'm going to get another new one for TO-39 transistors.  The top of the transistor gets quite hot after some use, so I can see why it could fail after a while without adequate cooling.

Why not?   The old one is going to be better than those push on jobs.

I haven't checked the specs but the old one is probably rated at 35 deg C/W [## see below]whereas the push-on types are 40 deg C/W to 42 deg C/W.

The push-on types can fall off sometimes.

The larger screw-on types I've seen are about 30 degC/W.

The screw-on types seem very expensive these days.
-------------------------------
##
From the tabulated specs the Thermalloy 2211 is rated at 49 deg C/W.
Seems pretty feeble, don't know the conditions of the rating.

Lower deg C/W is better.

[R.G.]

Good work!! Congrats!

I'm still tinkering a little with the peak limiter and understand now how to set it.  One thing I'm going to add while I'm at it is a capacitor at the output of the peak limiter so as not to have DC voltage in the shielded cable between the preamp and the input of the power amp.  I noticed that the 1143 version of the Beatle amp had this mod, which makes a lot of sense to me.

I dug deeply enough into the Thomas Vox line to write up what amounts to repair manuals for them. The mixer-limiter is a complicated bit of circuitry. It bothered me as well that they put DC across the signal line out of the preamp and into the power amp. But then, the power amp input is capacitor coupled, and so the DC doesn't matter, and they saved a capacitor, which were expensive back when these were manufactured.

But it goes deeper than that. The voltage at the inputs of the mixer are on the order of 25-50mV comparable to the guitar level signal. Nearly all of the effective gain of the preamp happens right there. The gain of the input circuits is thrown away in the tone circuits, and the tremolo and reverb circuits only work well on about 25-50mV signals because of the modulation and mixing approaches. So the mixer gets that back up to ~1V for feeding to the power amp.

The actual limiter circuit is fascinating. It's a variable soft clipper. Under the very beginning of limiting, it's linear enough to be unnoticeable. In limiting, it acts like a diode clipper with some resistance in series, much as we use diode pairs to clip signals. If the limit level is set correctly, the power amp never clips, but is only fed a signal that maxes out below the level that would make it clip. The signal is (relatively) softly limited, so this solid state amplifier exhibits much softer overload sound than you would expect from any solid state amp of the era. That is - it does not have the sudden blat into hard-edged distortion on peaks that musicians associate with solid state from that era.
The limiter's DC bias is critical, and there are three variants of the limiter that I know of, with slightly different gains, between 1 and 2. The last version of the limiter used 680 and 560 ohm resistors on the PNP. I suspect that this was done to keep the power amp stable. The Beatle power amp is marginally stable, and can oscillate if the impedance at its input is too high. One version of the Beatle split the output resistance of the limiter stage so that two bigger sized resistors were used, one in the preamp and one in the power amp. So the power amp chassis actually had a resistor that was part of the DC operation of the limiter circuit. Obviously, this version needed the DC in the cable to not muck up the biasing of the limiter. Another variant used two caps in the power amp, with a low resistance to ground between them. This version didn't need the DC on the cable, but used two biggish caps.

It's a fascinating electro-archeological dig. I've spent pretty much my entire adult life fascinated by the Thomas Vox line, silly teenager that I was.