Boogie Subway Rocket Reverb - help needed

uncledick

This has been a brilliant amp in the 18 years I've owned it but suddenly it's throwing a wobbly.  The audio out is quiet and distorted on rhythm and contour settings.  A bit louder on normal lead channel but not much.  I've swapped all the valves one at a time and no difference.  Just to eliminate the output stage, I plugged in some headphones.  Initially it was just the same as through the speaker but then decided to go into some kind of oscillation where it made a loud on/off hum/fart.  The on/off cycle was about 1Hz.  I then unplugged the headphones and it just goes into a mad oscillation which can only be stopped by turning it off.  Fwiw, I tried a jumper in the effects loop but it made no difference.  Reverb settings also make no difference.  Just before it started going really bad, if you hit the guitar fairly harshly, it was like there was a huge compressor in the loop i.e. the output would cut almost completely before slowly coming back a few seconds later.  Any pointers/tests before I lift the lid and get stuck in with a meter?

Thanks in advance
Dick

Maynehead

Does twiddling the fx loop level knob make any difference? With and without a jump lead?

uncledick

Maynehead said:

Does twiddling the fx loop level knob make any difference? With and without a jump lead?

No difference whatsoever.

ICBM

Wow. That's possibly the most bizarre fault I've ever heard of, which is saying something .

It sounds like the fault is in the preamp or the LT power supply if it was doing odd things through the headphones - as you know, the headphones are driven from a separate (solid-state) module and not the main power amp.

I would start by checking the voltages on every valve in in the preamp and see what you find.

uncledick

ICBM said:

Wow. That's possibly the most bizarre fault I've ever heard of, which is saying something .

It sounds like the fault is in the preamp or the LT power supply if it was doing odd things through the headphones - as you know, the headphones are driven from a separate (solid-state) module and not the main power amp.

I would start by checking the voltages on every valve in in the preamp and see what you find.

I'm on it!  Will report back.

uncledick

Ok, here goes:  As far as I can tell these are all in the ball-park although maybe slightly low (It is a 10 quid meter I keep for testing batteries & leads!)

All V1 - V4 in order and with respect to chassis ground - except where stated.  All controls set to zero.

Pin 1  140/140/139/230

Pin 2 all 0

Pin 3 1.2/1.2/1.2/2.5

Pin 4 & 5 V1/2 both 11.8Vac to pin 9.  V3/4 5.5Vac to pin 9.

Pin 6  142,141,94,247

Pin 7 all 0V

Pin 8 1.2,1.2,0.8, 2.55


Now it's out on the bench, the weird oscillation has stopped but it's still doing the whole quiet/farty thing.

ICBM

V3B looks odd - too low at both plate and cathode - but I haven’t got the schematic here, I’ll have a look tomorrow.

Heater voltages look wrong on V1 and V2.

uncledick

Slight update:  If I use my V Twin pedal as a preamp and go straight into the fx return, everything sounds fine.

ICBM

uncledick said:

Slight update:  If I use my V Twin pedal as a preamp and go straight into the fx return, everything sounds fine.

So it's not actually V3B that's the problem, since that's the return stage for the FX loop. Actually the voltage is about right, from the schematic - although they're all a bit too low. (V3 pin 6 should be 118V for example.)

It is definitely a preamp problem, or to do with the low-voltage supply which feeds the Contour circuit (immediately before the loop) and the switching, or the Contour circuit itself - all these are before the loop.

The odd AC voltages on V1&2 pins 4&5 are to do with it being a DC filament supply that's confusing your meter.

Further diagnosis without hands on is going to be quite difficult I'm afraid!

uncledick

Getting closer:  I don't seem to have a -30V line to the four transistors in the switching circuit.  I've found the 70Vac coming from the mains transformer.  I don't think either of the schematics I've found online are quite right so I might have to remove the board and trace it through.  The foot switch works which rules out that particular bit of the circuit so I'm probably down to about 8 components.  Help thus far is much appreciated.

ICBM

uncledick said:

Getting closer:  I don't seem to have a -30V line to the four transistors in the switching circuit.  I've found the 70Vac coming from the mains transformer.  I don't think either of the schematics I've found online are quite right so I might have to remove the board and trace it through.  The foot switch works which rules out that particular bit of the circuit so I'm probably down to about 8 components.  Help thus far is much appreciated.

Sounds like it's a LT supply problem. I don't have an accurate schematic either, but the one for the Subway Rocket (non reverb) shows a standard diode, resistor (unusually four in parallel, unlikely they've all gone), cap, resistor (again two in parallel), cap, Zener diode arrangement (two 15V, one for -15V and both for -30). Tracing the voltage through it to where it goes dead should pinpoint it. If I had to make a small bet, in an amp of this age I'd go for a shorted cap rather than an open circuit resistor, since the resistors are paralleled, and if one of the Zeners had shorted you would get -15V not zero.

uncledick

Spent a bit more time on this today.  In an effort to understand the circuit more clearly, I took the board out of the case so I could do a bit of reverse engineering.  Having drawn the layout on a piece of A4 and doing a search on Google, it would seem that the Subway Rocket Reverb uses the same low voltage supply as the Rectoverb 50 Series 2.  Without going into all the detail, this uses  a 560 Ohm resistor on the transformer, 2 1N4007s, 2 zeners, 4 100R resistors and 4 1000uF caps.  I removed all 4 caps and the zeners and they all appear ok. 

I then measured the resistance/load 'seen' by both circuits:  The +15V side is around 1Meg but the -15V is more like 80 Ohms! After much head scratching, I reckoned that no single component in the transistor (Q1-Q4) circuit - I think this is some kind of differential amplifier - could fail and give that kind of load.  Careful use of a sharp knife on the pcb isolated that circuit from the headphone out and reverb drive circuits which basically consists of 2 5532AP dual op-amps.  This is where the 80 Ohms still resides.  My assumption is that the resistance across pins 4 and 5 of these devices should be considerably higher.  I know I could remove them and compare but that's a bit of a faff.

Would anyone like to comment?

ICBM

Yes, sounds like a shorted IC.

DJH83004

To be 100% honest I think you are getting to the limits of remote advice the guys on the forum can give. Its getting to the point where you need the amp in front of you to diagnose further. I would be very wary of taking resistance / load measurements on dead circuits as anything meaningful. Components can show resistance values then break down under voltage, the only real exception being dead shorts (and even then cross checking by reversing polarity of the meter probes). Like IC I have repaired many Mesas over the years and before now have had to frustratingly remove chips, FETs and relays to eventually track down a fault. All that said good luck   

uncledick

DJH83004 said:

To be 100% honest I think you are getting to the limits of remote advice the guys on the forum can give. Its getting to the point where you need the amp in front of you to diagnose further. I would be very wary of taking resistance / load measurements on dead circuits as anything meaningful. Components can show resistance values then break down under voltage, the only real exception being dead shorts (and even then cross checking by reversing polarity of the meter probes). Like IC I have repaired many Mesas over the years and before now have had to frustratingly remove chips, FETs and relays to eventually track down a fault. All that said good luck   

Yeah, point taken, I like a challenge though.

As a final bit of advice, who would you recommend to buy small quantities of components from?  My late brother was a TV engineer by trade so I had access to RS for bits & pieces.  They were always a bit sniffy about selling to the public though.

DJH83004

Can't fault you for having a delve, you clearly know what you are doing regarding the dangers lurking inside valve amps, which is the main concern when giving advice. Regarding suppliers, yes you will be able to set up an account with RS or Farnell for components, the only downside being they often only sell multiples of components e.g. packs of 5 or 10. Believe it or not EBay is a good source of more obscure parts as well.  

uncledick

All sorted  After desoldering the two op-amps, one of them did - as expected - measure just over 80 Ohms between pins 3 & 4.  To get free carriage from Farnell, it was a 20 quid spend so I ordered enough caps to replace all the electrolytics, whilst it was on the kitchen table bench.  I also put in some 8 pin sockets for the new chips because a) it makes future testing easier and b) it saves any chance of me cooking the new bits with my soldering iron.

Everything is now back together and working just as it should.  Almost as an aside, I was digging around in the attic putting the Xmas decs away when I found a box of spare, used, valves from my old AC30 days.  Suffice to say a nice british made Mullard in the V2 position is as sweet as a sweet thing.  Lovely harmonics and just a hint of squishiness which the old EH didn't have.

Thanks to the ever-patient @ICBM and @DJH83004 for their help and encouragement.

tone1

Well done Chap.... it’s nice to know you’re just down the road when my amp blows up

ICBM

Excellent .

Yes, for things like this I prefer to socket the chips as well. It's funny that Peavey always did this with their 'budget' solid-state amps, but Mesa doesn't with its supposedly fairly high-end valve ones...