Vox Buckingham Amp - Reverb Transformer Issue

[Frank_NH]

Hey all,

Since I'm working from home , I've had an opportunity to finish fixing up my Thomas Vox Buckingham project amp.  I've replaced all the electrolytic caps and it sounds pretty good (relatively noise free compared to its original state).  However, the reverb is not working correctly - I noticed that as I turn up the blend knob, the output gets noticeably lower.  I've checked all the reverb circuit board wiring and voltages and they seem to be OK. 

However, the reverb circuit uses a transformer, and based on my research, I think it's wired wrong (see image below).  I think the ground and reverb out wires should be on the left and the 25V supply connections on the right.  I didn't do anything to it and it's possible someone messed this up at the factory.  Also, I can't seem to find any information on this transformer - is there a replacement/equivalent if I need one?

Thanks in advance for any help!

[duck_arse]

erm - the image you've linked to is residing in an outlook account somewhere - we can't see it unless we log into your account. so, what's the password?

[Mark Hammer]

Calling RG!  Calling RG!  That's the guy you want to speak to regarding anything pertaining to solid-state Vox amps.

[R.G.]

mmm.... huh? uh.. I'm awake! Brazil!
No, wait...
OK, better now. 

Frank: good on you for fixing the old Thomas Vox stuff! One of my several sub-missions in life is to help save these things.

You did the right thing replacing all the caps first. It's a huge first step, although a PITA one, as you know. 
For some detailed answers:

However, the reverb is not working correctly - I noticed that as I turn up the blend knob, the output gets noticeably lower.  I've checked all the reverb circuit board wiring and voltages and they seem to be OK. 

Given that, the problem may be in the reverb recovery circuit. The reverb blend knob is a simple pot-in-the-middle fader between the buffered dry signal from the driver circuit to the output of the recovery circuit. If you're getting dry signal that gets quieter as you turn up the blend knob, the recovery circuit is not providing signal to the hot/reverb side of the blend knob.

This can be either passive (broken wire to the blend knob hot lug, busted blend pot) or active (the recovery circuit is actively providing zero signal). If it's passive, I'd expect there to be dry signal at all positions of the blend knob, just quieter at max reverb positions. If it's active, I'd expect the dry part of the signal to go nearly to zero at max reverb position. These are guesses based on looking at the circuit.

An active-type fault could be anything that fails to provide recovered reverb signal. This includes the recovery circuit itself before the blend pot, the reverb tank, the cables to/from the tank, the transformer, the transformer driver, etc.

However, the reverb circuit uses a transformer, and based on my research, I think it's wired wrong (see image below).  I think the ground and reverb out wires should be on the left and the 25V supply connections on the right.  I didn't do anything to it and it's possible someone messed this up at the factory.  Also, I can't seem to find any information on this transformer - is there a replacement/equivalent if I need one?

Let's start with hoping it's not the reverb transformer, because there are no replacements commercially available. The only direct replacement would be to cannibalize another amp (ugh!! sacrilege!) or winding/re-winding one. Rewinding is not impossible, just a worse and more painful PITA than cap replacement. But on the bright side, it would be incredibly rare for the reverb transformer to be dead. These are just not stressed in this circuit like power or output driver transformers are.

It is possible it was mis-wired at the factory. My amp tech friend once repaired a Fender that was having BAD intermittent pops, and it turned out to have the wires to the output tube socket lugs wound into the lugs, but not soldered. Worked that way for thirty years.

The reverb transformer has a high resistance winding and a low resistance winding. I measured those one time but I haven't turned up my notes yet. If I remember correctly, the primary was 125 ohms, the secondary much less. It's worth checking the resistances on the windings. It's also worth noting that as long as you're unsoldering wires, you can just try it both ways. This will not kill the transformer if you hook it up wrong for a while. They're tough.

I mention doing this by measuring the resistances on windings because Thomas transformers don't necessarily conform to a specific color scheme. I've seen a few counter examples on power amp driver transformers. Not many, but enough to make me trust, but verify.

Do you have access to an oscilloscope? The simple thing would be to look for signal on both sides of the blend pot and then track it back through the circuit.

[Frank_NH]

erm - the image you've linked to is residing in an outlook account somewhere - we can't see it unless we log into your account. so, what's the password?

Hmmm...posting these images are always problematic for me.  Here's a link - let me know if you can access it now.

https://1drv.ms/u/s!ApPayh2_367xiCFC8AU-FWHTRg5-

[Frank_NH]

R.G.  Thanks much!  I'll check the blend pot for problems.  I'm also planning to remove the transformer and reversing the wires.  If the transformer is bad, well I'll have to find an old junker that has an intact transformer.  Or maybe learn to rewind my own  .

For reference, the transformer is shown in this link on your geofex site:

http://geofex.com/Article_Folders/Repair%20Boards%20for%20US%20Vox%20Amps.htm

You can see that your ground/reverb tank inputs are on the left (with the 1 nF ceramic cap to ground) and the power connections on the right.  Mine are reversed.  The cap on mine however is in the correct spot on the left!  So obviously that was wrong and made me suspect it was a factory problem.

What may have happened is that this amp was upgraded at some point, as one thing I noticed is that the input jack is isolated from the chassis by nylon washers which probably helps with noise.  I think this was recommended by Thomas Vox after the release of the V1121 version of the amp along with other changes. When I first got the amp, it was quite noisy but changing out all the electros helped that a lot! 

Another interesting anomaly - the peak limiter resistors connected to the two transistors did not entirely match the schematic.  On the first transistor, Q210, they were 10K collector (correct) and 2.7K emitter (should be 4.7K).  For the second transistor, they were 680 ohms collector (should be 2.2K) and 2.2K emitter (correct).  Maybe they wanted to change that second transistor resistors to 680/560, and I tried that but the emitter voltage seemed too low (0.2V).  So I'm going back to the original values (2.2K/2.2K) and hope that is correct for the transistors I have.  Adjusting the peak limiter pot didn't seem to do a whole lot to the sound - maybe this is the problem?

At any rate, I know just about every wire and circuit component by heart now in that amp.  It's a pain to work on for obvious reasons but rewarding nonetheless.  And for those who are wondering about the tone (I have the 2x12 Vox cab) -  by themselves, the normal and brilliant channels are mostly clean with a little hair, so don't expect tube amp tone.  But running an overdrive pedal up front sounds fantastic!  And the tremolo sounds great as well.  I would comment about the fuzz but that needs to be fixed (the relay which switches the fuzz on wasn't working for some reason).

[R.G.]

You sound like you're already past needing it, but there's a Buckingham repair manual in the works for thomas-vox-repair.com. The Buck differs only in the power amp from the Guardsman and  Beatle of the same series. The preamp board at thomas-vox-repair works in the Buckingham just like the Guardsman and Beatle as well.

If you have isolating washers on the input jacks, it's definitely been repaired at. The grounding system in all of these amps make for a little hum that's impossible to remove without some special tricks. One of them is isolating the input jack bushings and grounding those only to the preamp board. I use nylon-nose jacks when I do this, but iso bushings work too. The other tricks are to break the preamp-output jack ground loose from chassis and ground it to the PCB and to the chassis through a 10 ohm resistor, and adding a return wire for the on and standby lights.

The preamp chassis can then be tied to the power supply ground with one wire, and perform it's main duty as a shield. The return wire for the panel lights runs AC, and if the preamp signal ground is attached to the chassis, the light's 120Hz current rides on the same power return as the chassis and inserts the wire resistance times the current as a ground offset of the signal out to the power amp. This took me a long time to figure out. 

When you get the reverb working, here are some sprinkles to put on the icing. You can replace the input transistors with 2SC1815 or 2N5088 (different pinouts! watch that!) and get lower input noise. You can keep that lower input noise by soldering a silicon diode reverse biased across the base emitter. This prevents input transients from reverse breaking the input transistor base-emitter, and making it noisier. Other hiss points are the reverb input, and the three transistors in the mixer limiter.

[duck_arse]

image is visible here now - which usually means everyone can see it.

[Frank_NH]

Well, I thought it would be good to update my post.  I've been working on the reverb circuit and still can't get it to work correctly.  I get a very weak signal at full blend.  I've checked all the voltages and they seem OK, though the collector of the driver transistor connected to the transformer has a voltage a bit lower than the schematic (18V versus 23V).  I'm going to to check all voltages again just to be sure, but it's possible I may need a new transformer.

BTW I did measure the resistances on the primary and secondary sides, and they were about 120 ohms and 50 ohms respectively.  Should the secondary be a lot lower than that?

For reference, here is a picture of how it is hooked up.  I believe this is correct, but I haven't found any pictures on the internet of this transformer as originally wired in the Buckingham (it is hidden after all in the bowels of the preamp chassis!).  As viewed, the wire on the upper left goes to the collector of the driver transistor, the lower left to a +25V supply, the upper right goes to the reverb tank (shield is grounded), and the lower right to ground.  A 1 nf cap is installed across the right side in accordance with the schematic.



I haven't checked the blend pot but I'll do that tonight.  It gives me full signal and a weak signal at max.  Maybe something is going on there.  I may also get out my audio probe to see if I can spot where the signal goes bad.

On the positive side, I got the fuzz working and it sounds really good!  Oh, and I fixed the resistors on peak limiter circuit so that now I'm getting full volume from all channels.  In fact, it's so loud, I can't really go past 7 on the Normal channel.  Not bad for a 30W solid state amp!  So... if I could only get this reverb working...

[Gus]

Have you checked the 1nF cap?
Old disk ceramics can fail.

[Frank_NH]

Have you checked the 1nF cap?
Old disk ceramics can fail.

Yes, I did, and in fact all the ceramic caps I've thought were bad and removed have been fine.  No harm as they're easily replaced. 

Actually, I worked on it more last night, and I think I know the problem now.  The high current driver transistor with the funny heat sink shown in the photo looks to be bad.  This make some sense as the weak signal I'm getting was probably because the amp stage driving the reverb tank was bad.   The collector measured at about 18V while the base and emitter were nearly equal voltages at about 8.5V.  It's hard to test the transistor in circuit so I'll go ahead and remove it to check it off the PCB.  I don't know what a suitable replacement is (R.G. may know), but I'm sure a modern substitute exists.  Otherwise, I'll have to check out the vintage electronics parts market on eBay! 

[anotherjim]

It's a Darlington?

R.G.?

[Frank_NH]

Take the transistor out and diode test the junctions. It should test like 2 diodes with a common connection at the base. Sometimes you can do that with it in circuit but there's a chance the rest of the circuit will mess the readings if you do.
If it's NPN, a 2N3053 might be a candidate (40v 5W).

Thanks Jim, I'm going to do just that.  I've found a replacement part on eBay (2N2219, TO-39) and will install it if the existing transistor is fried.  It doesn't surprise me that these transistors would go after a while as they get hot and need a heat sink to dissipate the energy.   There's second driver transistor like this one in the power amp stage, and I may replace it as well given its age.

[anotherjim]

In case my quote puzzles anyone, I went back and edited my post while Frank was answering. After a hard squint at the schematic I found I twigged that with the base & emitter voltages and the symbol, the reverb driver has to be a Darlington.
Like the OT in a tube amp, any trouble on the secondary side of the transformer can cause excessive & damaging flyback voltages on the primary side. Semiconductors will be even more prone to damage than tubes. An intermittent contact in the cable to the spring could be responsible. If it were mine, I think I'd add a pair of protection diodes, like the 1N4001's on the output of this chip amp...

[Frank_NH]

In case my quote puzzles anyone, I went back and edited my post while Frank was answering. After a hard squint at the schematic I found I twigged that with the base & emitter voltages and the symbol, the reverb driver has to be a Darlington.
Like the OT in a tube amp, any trouble on the secondary side of the transformer can cause excessive & damaging flyback voltages on the primary side. Semiconductors will be even more prone to damage than tubes. An intermittent contact in the cable to the spring could be responsible. If it were mine, I think I'd add a pair of protection diodes, like the 1N4001's on the output of this chip amp...

Thanks anotherjim.  Interesting suggestion.  So, you would place the protection diodes at the output (collector) of Q202 (the driver transistor)?  That could be done with some careful soldering I suppose.  Here's the reverb circuit schematic for reference.

By the way, my reverb circuit was modified sometime in the past to incorporate a 220 pf capacitor between the base and collector of Q202 (it was soldered underneath the PCB!).  I think this mod among others made to later production units.  I noticed that later Beatle amps (version 1143) used a different driver circuit and dispensed with the transformer entirely!

[anotherjim]

I would also check the resistors going to ground from the emitter of the driver. If either of those failed open circuit, you would get higher voltage readings and probably the same on the bass and emitter when there should be 2 diode drops difference (1.4v to 1.6v difference).

Yes, the protection diodes join between the load (primary) and the collector. What they do is catch voltage surges - preventing them from going more positive or more negative than either the supply or ground. They have zero impact on normal operation.

An old organ I had used 38V and a single TO39 transistor drive - but they probably had a spring tank with a higher input impedance. A transformer can be needed to use a low impedance tank.

[R.G.]

Just catching back up.

I've replaced these things with TO220 "driver" transistors. The cheapest is the 2SC/KSC2073. These work. So do some others that are bigger power/current/voltage ratings. Latest I checked - which was last year! - they cost about US$0.50.

If you still need one, try one of these. The pinout, is of course not anything like the TO39. The TO220 >might< not need a heat sink. A TO220 can get rid of 2W with no sink. With a tab of aluminum bolted to it, it can probably to to maybe 4-5W.

These things have good gain for most of their current range. A darlington isn't needed. I have a PCB that replaces the entire works of the V1121/1151/1131/1141 preamp chassis, and the 2073 has worked well in these.

[Frank_NH]

Just catching back up.

I've replaced these things with TO220 "driver" transistors. The cheapest is the 2SC/KSC2073. These work. So do some others that are bigger power/current/voltage ratings. Latest I checked - which was last year! - they cost about US$0.50.

If you still need one, try one of these. The pinout, is of course not anything like the TO39. The TO220 >might< not need a heat sink. A TO220 can get rid of 2W with no sink. With a tab of aluminum bolted to it, it can probably to to maybe 4-5W.

These things have good gain for most of their current range. A darlington isn't needed. I have a PCB that replaces the entire works of the V1121/1151/1131/1141 preamp chassis, and the 2073 has worked well in these.

Thank you R.G.  I have a 2N2219 TO-39 transistor on the way and will see if it works.  I did test the old one after removing it and sure enough the base-emitter junction is bad, so away it goes.  Hoping a new transistor does the trick.  I'll report back...

Everything else on the amp works, and the only mod left will be to replace the 2 prong cord with a grounded one and remove the "death cap".

Thinking about the protection diodes, those could be installed right on the transformer, where you have connections for the source voltage, Q202 collector, and ground.

[PRR]

> 2N2219 ...on the way

As I read the plan (and R.G. will lasso me if I am wrong), that needs to be a 2N2219A.

You have 24V supply into a transformer coupled stage. Counting on one thumb, you "can" have 48V spikes. (And even more if not loaded, and a tank is poor load.)

The no-A part is 30V Vceo. The 2N2219A part is 40V Vceo.
https://www.mouser.com/datasheet/2/68/2n2219-19a-369102.pdf
Strictly, the least signal may zap a no-A part. The A part is almost good enough.

Both parts have higher Vcbo specs. And I kinda remember when Motorola was handing-out 10-packs of gold 2N22xx parts to anyone, even RCA engineers. As I recall at first they were lucky/happy to get 30V. Later they got better and they'd sort-out some 40V parts for a premium price. Perhaps at the end of the metal-can era they were all 60V parts, but "Shhhh!!" they had a different part# and price for the 60V rating.

As R.G. says, by the time of TO220 and TIP parts, general-purpose bipolars were all 40V, with 60 and 80V as nickel/dime upgrades. So even a no-A part has a good chance of living in this 24V circuit. Unless you have been digging in the early 1960s.

Another path is a Zener clipper. Strictly one Zener and one plain diode will do. For a one/few-off upgrade, two 47V Zeners will work and eliminates polarity confusion.

[Frank_NH]

> 2N2219 ...on the way

As I read the plan (and R.G. will lasso me if I am wrong), that needs to be a 2N2219A.

You have 24V supply into a transformer coupled stage. Counting on one thumb, you "can" have 48V spikes. (And even more if not loaded, and a tank is poor load.)

The no-A part is 30V Vceo. The 2N2219A part is 40V Vceo.
https://www.mouser.com/datasheet/2/68/2n2219-19a-369102.pdf
Strictly, the least signal may zap a no-A part. The A part is almost good enough.

Both parts have higher Vcbo specs. And I kinda remember when Motorola was handing-out 10-packs of gold 2N22xx parts to anyone, even RCA engineers. As I recall at first they were lucky/happy to get 30V. Later they got better and they'd sort-out some 40V parts for a premium price. Perhaps at the end of the metal-can era they were all 60V parts, but "Shhhh!!" they had a different part# and price for the 60V rating.

As R.G. says, by the time of TO220 and TIP parts, general-purpose bipolars were all 40V, with 60 and 80V as nickel/dime upgrades. So even a no-A part has a good chance of living in this 24V circuit. Unless you have been digging in the early 1960s.

Another path is a Zener clipper. Strictly one Zener and one plain diode will do. For a one/few-off upgrade, two 47V Zeners will work and eliminates polarity confusion.

Thanks Paul.  As it turns out, the transistors I'm getting from eBay are in fact 2N2219A.    I looked up the data sheets and it looked like the A version (as you correctly point out) was more hearty than the regular 2N2219.  They were inexpensive and came in a package of 14, so I'll have a few left over. Maybe I can make a 25V superfuzz! 

I like the idea of using Zener diodes for voltage protection (cheap insurance).  I could solder them directly to the transformer lugs along with the 1nF capacitor (C228), so an easy install.