Marshall Trem Reverb 20

A Marshall amp I've never worked on, or even seen!

The tremolo on Marshalls is always awful though, so I can't see why this one should be any different. It really puzzles me - given that their entire company was built on copying a Fender Bassman, why didn't they just copy a Fender tremolo circuit too?

That's exactly what I'd do - change it to the same values as a BF Fender Princeton (also bias-modulated, and the layout is identical) circuit and it should work. The caps are all the same, it's just the resistors that are different. Your only problem will be getting a 3M reverse-audio pot for the speed control, but 2M Lin might give a usable range. There's an additional .022 cap to ground in the Fender circuit but that should be easy to add.

 

Ah - the 2046 Specialist. I thought for a minute there was some super rare 20Wer (2xEL84).

I've also heard the same stories about overheating PCB due to the upside down valves.

 

Ah... wait a minute. I have seen one of those, there's one sitting in my garage! I'm not quite losing my marbles, the difference is that this one is a Park. I haven't opened it up yet, and it hasn't been turned on in maybe twenty years... perhaps a good thing.

I thought it was some obscure Park model, but now you come to mention it I have seen pics of the Marshall version too, and I knew about them cooking the circuit boards. The funny thing is that it looks quite different (and much cooler) in the Park color scheme with a blue grille cloth.

 

Well, I finally got around to working on this thing today, and I'm very surprised.

It's actually quite cool.

I wouldn't say it sounds great - rather dark and muffled-sounding, and gets a bit strident and midrangy (not bright or sparkly) if you turn the tone up - but it does work well enough, and doesn't really sound like any other amp. The reverb and trem do both work well, with a good useful speed and depth range on the trem - and it is dead stock too (apart from replaced rectifier diodes), so AmpNerd's one must be faulty, rather than just bad by design.

What's really odd is that the reverb and tone controls are slightly interactive - the lower the tone is set, the more reverb you get! I can see why from the schematic - the tone control is almost acting as a pan between the dry feed from the volume control and the reverb return.

The only actual repair it needed was to resolder a broken filter cap solder joint - I haven't changed the caps since they're an utter pain to get at and with a very slow Variac-up they appear to be fine... and I somehow doubt this is going to be anyone's working amp . The board is badly heat-damaged around the EL34s and ECL86 though, and flexes alarmingly when removing or inserting the tubes. It really is very poorly designed from a structural point of view.

What's even more utterly stupid is that there is no ventilation at all for the chassis - the whole of the top is closed in by the upper back panel, which extends well below the tubes so trapping all the hot air up there. Surely a simple vent at the top would have been a good idea? I've just drilled two more holes in the side rails and mounted the back panel a couple of inches lower down...

It's basically a giant Marshall Mercury with reverb, really! . Interesting as a curiosity but I won't be too sorry if it's as long until I have to work on another one.

Anyway, here it is...

Much cooler than the Marshall...


Back panel mounted lower (the upper screws are in the original holes for the lower ones)...


Chassis outside - note that the reverb transformer HT connections are on the top of it and completely uninsulated...


Do I smell burning... ?

 

Is the reverb tank fixed to the back of the chassis using the four small holes you can see in the back view?

I couldn't work out what these were for - but there is evidence around the holes that nuts were fitted at some time, some rotary scratching and red ink:


And also the tremolo footswitch wire is led sideways about four inches before exiting the back of the amp, which would only be necessary if something was in the way... I wondered if it had been a heat sink, but I thought that would have been remarkably advanced for the time it was built (and I couldn't see why anyone would remove it) - but a small reverb tank would be about that size too.

It would be a bit of a puzzle if so, since the wires that go to the tank in the bottom of the cabinet do appear to be stock as well, and their solder connections at the amp end are also inked.

Here's the transformer:


Connections left to right are: HT, ECL86 plate (shielded), ground (+ both shields), tank send (shielded).

I did re-rout the send cable into the chassis through the same grommet as the others, then back out through the grommet for the footswitch cable - which was missing - to tidy it up and prevent any tug on the (very long and completely loose) wire pulling it off the transformer. The odd thing is that the wire length was exactly right for that to have been intended in the first place... and yet it wasn't done like that.

It's almost as if this one was modded at the factory - it's not impossible that someone else could have re-inked the connections, but I think it's unlikely on an amp like this where there's no great value in originality.

Strange, since it looks absolutely dead stock otherwise, with the exception of the rectifier diodes and the power tubes.

Curiouser and curiouser...

 

HT = "high tension" = B+ (in this case).

Maybe just an archaic English bit of jargon that never made it over the pond?

 

Unfortunately it's back together again and I'm a little unwilling to take it apart... You'll know exactly why. I'll take the chassis out if you really need me to, but I'd rather not lift the board. I've got the original full-res versions of those pics though, so I'll do the best I can with those for now...

HT = High Tension / B+.

That's the red wire going over the back of the board:


The shielded cable is the reverb return. The cable to the ECL86 plate goes to the underside of the board (and is very short, making lifting it a pain), so you can't see it in the pic.

For some reason the grounds (black wire also disappearing over the back, and the shield) are just soldered to the flat traces, not to the turret provided for the purpose...

There were originally two separate footswitches, according to someone who has seen one of the Marshall versions, and they do have separate cables. I have just the one for the trem (marked 'T' on it, I assume the reverb one had an 'R') and the reverb one had been cut just outside the amp, so I removed the cable completely... no-one really turns the reverb off anyway.

If the trem is too shallow and too fast - and stops working at all when the speed is low - have you tried replacing the cathode bypass cap? It sounds like the circuit has too little gain and rolls off at too high a frequency.

 

No - the first trace is one of the B+ connections to the dual cap can, which is held to the PCB by the cord that passes over the three traces further down.

It may have been labeled 'red' in reference to the cap color codes. I can't remember if that is the red terminal or not (the other one is the one nearest the black ground wire).

That's the wire that goes to the transformer - it's the plate connection for the pentode half of the ECL86.

It sounds like you have the reverb wired up right, but the tank on yours may just be crap. Mine has clearly been converted, now we've put the puzzle together! There was obviously once a small tank on the back of the chassis, but now there's a full-size one in the bottom of the cabinet. The long wires that reach it are 'factory' too though from the ink on their connections and the type of cable (identical to the footswitch wiring), so I'm guessing that this one was factory modified - maybe after being returned under warranty? Or a left-over Marshall chassis that was rebuilt as a Park in a factory clear-out? Who knows...

Who would have thought this thing would turn out to be so 'interesting'?


Also interestingly, this one has improved tonally since I've fixed it. I ran it for six hours straight yesterday, and it's significantly more chimey and even has a bit of sparkle when the tone control is up all the way now. Quite nice. I'd guess it hasn't been turned on in nearly 20 years - the power tubes that were fitted were mid-80s Chinese types, and I'd guess the reason it was put away was that broken filter cap connection, which caused violent popping noises and loud hum as it made intermittent contact... and might even have been caused by the stress of changing the tubes.

I'm not sure I'd go so far as to say I like it, but it's not bad. If only they'd built it better.

I even cleaned it too .

 

Ah... I can see them now.

Even more charred than the Park! It's interesting that it's not just the EL34s that are the problem, that big power resistor next to them is if anything even worse...

The reverb tank position - and the reason the back panel goes right up to the top of the amp (which I think is a significant contributor to the overheating) - makes sense now.

I still think it looks cooler as a Park though .