Marshall Mosfet 100 Problem

HHwarner

I took my backup amp (Marshall Mosfet 100) to the studio tonight to put it through its paces. Its the first chance iv had to really try it out. All was fine for about an hour then it started to get a bit crackly and the volume dropped off. At this point I switched it off as I could smell burning. I got it home and opened it up to find the mess below, R46 & R47 fried. Is it worth just replacing the fried resistors or should I get it checked out by a tech. Bear in its a £100 amp so I don't want to chuck too much at it

http://i879.photobucket.com/albums/ab355/hhwarner/mosfet1.jpg

Misterg

Found this while I was looking for a schematic:

"I know I'm a little late in answering this query, but thought this might be of benefit to others with the problem of fried resistors in the Lead 100 Mosfet. The resistors R46 and R47, both of which are 270 Ohms 2 Watt, overheat because of a fault in the reverb driver circuit based around TR5 and TR6. Remove both transistors and test the small diodes D9 and D8, both of which should show infinite resistance in one direction. These diodes are prone to failure and can cause the failure of either TR5 or TR6 or both. In addition, failure of these transistors will almost inevitably burn out the input (send) coil in the reverb chamber, which will also need to be replaced. TR5 is an MPSA06 or BC538; TR6 is MPSA56 or BC528. Both are 80V transistors, so don't be tempted to use any lesser device. They should both measure infinite resistance between collector and emitter when removed. The diodes D9 and D8 (originally 1N914) can be replaced with 1N4148.

Best wishes, Ray"

:- /

HHwarner

Thanks for that, it makes perfect sense. The reverb has been intermittently working. I wasn't overly worried as I don't use the reverb

ICBM

Quite an odd one, that… those are the resistors which feed the second-level power rails (+/-34V) rather than the full voltage to the power transistors, so whatever has gone wrong is at that level - also not at the +/-15V level or R44 and R45 would have gone as well/instead.

There's no point in changing them - they're the symptom not the cause, new ones will just go the same way because something they feed is shorted. It doesn't look likely to be a filter cap failure since that would be unlikely to affect both halves at once, which leaves… probably the reverb driver circuit. That's essentially a small power amp which operates from the +/-34V rails. It's probably one or both of the two transistors connected to R31 or R33 (can't decipher the numbers on the schematic, it's too blurred). They connect across between the two rails with no common ground point so a short there would cook both the resistors equally.

At least I think that's the most logical explanation!

It's a shame these amps aren't worth much because they really are very good - this is almost the first serious fault I've ever heard of on one. It's worth fixing even if it costs close to the financial value - they're normally extremely reliable.

HHwarner

Cheers ICBM, I have had problems with the reverb working intermittently, the tank is ok so reverb driver is the likely culprit

ICBM

Ah! Misterg must have found that while I was trying to read the schematic and work out what could have caused it .

If you don't mind risking a bit of possibly unnecessary damage…

Find TR5 and TR6. See which leg of the transistor connects to the trace which runs all the way back to R46 or R47, and cut that one on both - carefully, leaving a small gap so you can resolder it if you want at a later date. That will disconnect both the transistors and the possibly shorted diodes from the supply rails. (They will still be connected via R29 and R30 but these are 100K so they can't draw enough current to be a problem.)

Now power the amp up again. If the R46 and 47 resistors start to smoke the fault is elsewhere! If they don't, it will probably work - even though they look cooked, they've probably survived and might not even be badly out of tolerance.

HHwarner

Ill give that a go, many thanks once again

HHwarner

Misterg said:

Found this while I was looking for a schematic:

"I know I'm a little late in answering this query, but thought this might be of benefit to others with the problem of fried resistors in the Lead 100 Mosfet. The resistors R46 and R47, both of which are 270 Ohms 2 Watt, overheat because of a fault in the reverb driver circuit based around TR5 and TR6. Remove both transistors and test the small diodes D9 and D8, both of which should show infinite resistance in one direction. These diodes are prone to failure and can cause the failure of either TR5 or TR6 or both. In addition, failure of these transistors will almost inevitably burn out the input (send) coil in the reverb chamber, which will also need to be replaced. TR5 is an MPSA06 or BC538; TR6 is MPSA56 or BC528. Both are 80V transistors, so don't be tempted to use any lesser device. They should both measure infinite resistance between collector and emitter when removed. The diodes D9 and D8 (originally 1N914) can be replaced with 1N4148.

Best wishes, Ray"

Looking at the schematic for mine I think Im correct in saying its TR4 & TR5, is that correct

Schematic

Misterg

HHwarner said:

Looking at the schematic for mine I think Im correct in saying its TR4 & TR5, is that correct

Schematic

Yes - I would have suggested taking them out, but ICBM's suggestion is neater.

HHwarner

Iv actually removed both transistors and tested those and the diodes as above Im getting infinite resistance in one direction on D8 and D9. Iv checked both TR4 & TR5 between collector & emitter. TR5 is infinite TR4 is not so I guess TR4 is caput. The amp runs fine without them. No smoking resistors so I guess that's where the problem is hopefully.

Misterg

If TR4 shows continuity between C and E, then it's dead and it would account for the symptoms - there aren't many other things that could (as ICBM said). I'm slightly surprised it didn't take out TR5, too, but hey ho...

The proof of the pudding is that the amp works OK without them. It's possible that voltage spikes from a bad connection to the reverb tank 'send' could have taken out TR4 - worth checking should you ever decide to replace it. (I'd leave TR5 out until you do).

HHwarner

Thanks for your help guys. I will probably be back with this one as I'm going to try and replace both transistors.

ICBM

HHwarner said:

I will probably be back with this one as I'm going to try and replace both transistors.

If you do, replace the two diodes with 1N4148s as suggested as well, since they're a known cause of failure even if they haven't done so in this case.

HHwarner

Cheers ICBM. Iv ordered the diodes also. Although I don't use the reverb I like things to be working correctly. Thanks once again

HHwarner

Well iv fitted the new transistors and diodes. Everything works fine except the reverb is still intermittent and crackly. When I bang the amp I can't hear the reverb springs coming through the speaker so I guess the tank is iffy. Iv disconnected it now. Funny thing though, the reverb pot has an effect on the eq when turned even with the tank disconnected. It's like a tone knob???

Misterg

Just checking you've disconnected the input to the tank as well as the output - a dodgy input connection could have been responsible for killing TR4.

The reverb pot could have some audible effect, but I would expect it to be pretty subtle. If it's very noticeable, try shorting out the phono plug on the end of the lead that plugs into the output of the tank in case there is some weird feedback / oscillation thing going on.

HHwarner

Yes Iv disconnected both input and output. The tonal effect of the reverb pot is quite subtle. Ill try shorting the output though and see what happens. Thanks, for your help

Another question Im afraid, does anyone know where I can get exact replacement resistors from or are there any alternatives I can use. When I measured them last night they were reading 370 ohms

ICBM

If you don't get reverb crash when you bang the amp it's the output end of the tank which is dodgy. Although it could be both.

The reverb knob will act as a tone control since it's rolling off treble to ground via C26 and R36 when the pot is at zero. It won't have a huge effect but it will probably be audible. It's not an ideal design really, it would be better if the pot was 'backwards' - rotor input, not output.

Any 2W 270-ohm wirewound resistors will be fine, although if you can get ones that fit I'd go for bigger power rating. You should be able to get at lest 3W in that space.

HHwarner

Will something like this be ok

http://uk.rs-online.com/web/p/through-hole-fixed-resistors/4851975/

ICBM

Ideal, in fact better than the ones it had.

HHwarner

Thanks once again for your help. Much appreciated. Iv a lot to learn but willing to try