Marshall Mosfet Lead Output Stage Problem

[J M Fahey]

Yes, the current source supplies 6,5 mA; which translate into 650mV across 100 ohms, which I find *very* low even for audio type MosFets.
I would put a 1K trimmer there (set to 0 ohms) and *very carefully* increase its setting until I got, say, 33mV across each .33 ohm source resistor, equal to 100mA bias current.
I'd play until heatsinks get warm and check again .
If bias current kept the same or went down, I'd leave it there; if it went up, I'd lower it to said 100mA.
You can solder a 10uF cap across the trimmer legs.
If and when you re-rewind your power transformer to original spec, re-bias the power stage.
Good luck.
Pobre Chile, buena suerte es lo que le falta últimamente.
Un abrazo.
Juan Manuel Fahey.
PD: si alguna vez andás por Buenos Aires y querés hacerlo, traete la cabeza que le pegamos una revisada en mi taller, sin compromiso.

[Mars Amp Repair]

Wow,
Great info here about the parasitic oscillation issue with the Mosfets...gate stopper resistors (why not? I guess it would be a similar issue as used with internal capacitance like a tube) , how to build a Zobel...this will be very helpful with some issues I've had on a few Trace Elliot & other mosfet amps.

Also biasing the mosfets in the Mosfet power stages, great stuff. Thanx for all the info in for an ancillitory issue...glen

[MarkusBass]

Wow,
Great info here about the parasitic oscillation issue with the Mosfets...gate stopper resistors (why not? I guess it would be a similar issue as used with internal capacitance like a tube)

In the amp that I build (based on Marshall power amp), I used 2SK1058 and 2SJ162. They require much higher voltage on gates then original 2SK135 and 2SJ50 transistors - at least 2 Volts. This can be achieved by increasing R14 value. You may also encounter oscillations - I added 5,6pF capacitors between gate and drain of 2SK1058 (although some people advise to add even 47pF). I think that I tried to increase gate resistors but it did not help.

this will be very helpful with some issues I've had on a few Trace Elliot & other mosfet amps.

If you look carefully at Marshall and Trace Elliot schematics, you may find out that they could be designed by the same person (or copied). I think that TE uses BUZ9xx transistors which are still available. I fixed lately a rare TE 1225V bass combo (has anyone heard about it?) - very loud and great sound, probably partly due to MOSFET power module (and high voltage power supply).

Mark

[olddawg]

Marshall Lead 100 MOSFET? I own one of these. I picked it up with the cab and a pile of other stuff for $125 a while back. It's red. I had to fix it but the problem wasn't the output stage. I don't like it much but it is loud. It seems like you are going to a lot of trouble. These amps are not that valuable or even desireable for most people except to gut and build a smaller tube amp in the cab. It seems like it's really butchered. Here's what I would do. Look on eBay for a doner amp. You can get a working ratty one for under $200. Then swap out the chassis. Actually I would gut it and build a plexi in the box. Either way you will be spending a lot less time and frustration unless you have made it a mission from god. BTW, do not crank these amps on a 4 ohm load. I know the specs say you can, but they originally came with one or two 16 ohm cabs. Anyone I know that ran them at 4 ohms killed them.

[J M Fahey]

If you look carefully at Marshall and Trace Elliot schematics, you may find out that they could be designed by the same person (or copied).

You bet. It's basically the Hitachi datasheet example or its sibling, the Exicon "simple" version (They also offer a *much* more complex version, which nobody uses)

These amps are not that valuable or even desireable for most people except to gut and build a smaller tube amp in the cab.

Well, not to tubeheads
Yet, they are LOUD , and a good platform to power good pedalboards, PODs and the like.
I agree on the 4 ohm load being stressing, a common current capability in "Audio" MosFets is skimpy 6.5A, overheating a lot beyond that.

[olddawg]

Well, not to tubeheads
Yet, they are LOUD , and a good platform to power good pedalboards, PODs and the like. I agree on the 4 ohm load being stressing, a common current capability in "Audio" MosFets is skimpy 6.5A, overheating a lot beyond that.

Like I said I own one. And actually, it is one of the better SS heads I have used. I also said they very are loud. It's my backup head. But yeah, tube heads convert them, especially if you get a blown one with the factory speaker cab to make a nice looking set. An 18 watt clone in that head cab and the 1965A 4X10 speaker cab is a great setup and an easy conversion. Done right it is a lower volume JCM800 and even looks like one. Seriously, check eBay every day for a couple of weeks. You will eventually find a ratty but working amp that you can swap parts from or a naked chassis left over from a conversion for $150 or so. I've seen them go for $100. Compared to the headache of trying to straighten that one out it's money well spent IMHO. I can't believe that someone actually rewound the PT in the thing. You might even find just a replacement PT online for cheap. I wouldn't re-engineer it. Get the right parts to start with. Get a doner. BTW, on another forum some kid was asking how to make one of these even louder. Some hotshot genious told him to rewire his cab for 4 ohms. I advised him not to do it. He posted a couple of days later that the amp was a little louder but it blew up after a couple of hour of play. If you post over at Guitar Geek someone might have a doner amp for cheap. There were a couple available a few weeks ago. Post in the guitar amp section.

[MarkusBass]

You will eventually find a ratty but working amp that you can swap parts from or a naked chassis left over from a conversion for $150 or so. I've seen them go for $100.

I see that the author of the topic is from Chile and I don't think that he can buy such an amp for 100$ (how far is Chile from the States?).
He wants to fix it and in my opinion this is a good idea.

Compared to the headache of trying to straighten that one out it's money well spent IMHO.

Well, it depends on someone's experience. For me it's no more than half an hour to fix the amp (as I wrote I build such an amp - from scratch). I wouldn't spend 100$ + delivery fee (another 100$ ?) on another amp to be fixed. Let Lee learn how to fix the amp.

Olddawg, you assume that Lee lives in the States and can easily spend 100$ on another amp. But this is a false assumption. Have you heard lately about what happend in Chile? I bet you had.

Mark

[Steve Conner]

Well, let's fix it then!

You never answered my questions and I think they're important.

Apart from the hum, does the amp work, can you play through it?

If so then it could be a grounding problem, the power amp could be working fine.

I've built a few amps using these lateral MOSFETs. I had one that would just burst into oscillation and blow its fuses at power up. I used different FETs to the design I was copying, and it turned out that it now needed a Zobel network, whereas the original was happy without one.

I don't know about Markus' idea of using capacitors to snub the oscillation. To me, capacitances are what cause the parasitics, and resistors are what you need to damp them out.

[Lee_ranaldo]

Thanks guys for ur replies
!

Although it's not a collectible amp i do like it since i like to restore vintage equipment. Althought i have the money to spend it for another amp in the same price range my goal is not buying another just restoring this amp into his original status (or let's say close to it)

The chassi is indeed big ( as big as the JCM800 2204 i have laying around)
sounds like a good candidate for a tube amp.

I have no idea about the zobell network thing. I tried to resolve parasitic oscilations in tube amps with the cap solution and sometimes it did the job, sometimes didn't.

Give me a couple of hours and i'll update the voltage readings witht he fixed board that i'm working on now.

Thanks guys for ur replies! and i hope this thread is usefull to some other guy restoring another Mosfet Lead

Cheers!

PD: muchas gracias por tus comentarios de apoyo, por suerte aca no nos ocurrio nada solo figuras de adorno quebradas pero todos bien.
Ojala sirva para unirnos mas por que tenemos estos conflictos heredados que generan desunion

Un abrazo hermano argentino y salud!

[Mars Amp Repair]

yep,
In the Trace GP7 SM bass amp, I changed the BUZ9xxx mosfets to 2SJ50 & 2sk135. I had to add source ballast resistors to balance out the current sharing as well as snubber discs between the G-D to keep that bad boy from oscillating.
Worked like a champ. glen

[J M Fahey]

Boring International Economics.

Well, sort of.
Just to put things in perspective, let me explain why this old workhorse is worth repairing, rewinding, whatever in Chile (or Argentina, for that matter).
Beautiful Santiago and my own Buenos Aires are 1000Km apart (or close, take your pick); 10000Km from USA; 14000Km from Europe.
Really distance by itself is not that important, Customs are.
Freight increases too, way over the part cost.
That's why a U$S 4 MosFet, "free delivery anywhere in Continental US", becomes a U$S 12 part, considering airfreight 7000 miles away, plus a trip downtown to pick it up at some Customs office, bring a good book with you to stand the long, slow queue.
About the MosFet 100: here it's not a $100 pawnshop special or a $10 garage sale item (if dead) but :
WOW !!!!!!!!!!!!!! A MARSHALL !!!!!!!!!!!!!!!! ......... "vintage" (25 Y.O.) ............. Made in ENGLAND !!!!!!!!!!!!
all of which is true, and, besides, the quality of construction is 80's England, not 2010's India or China.
I'm not demeaning India or China at all, but the bean counter mentality that usually surrounds modern cut-throating , price based competition.
I searched on Chilean EBay (Mercado Libre) and found:
Cabezal Marshall Mosfet Lead 100 - $ 300000.00 - MercadoLibre
the asking price is 300000 Chilean Pesos, or U$S 600.
To make things even clearer, an Argentine or Chilean minimum salary is around 300 U$S, so this used, 25 Y.O. head, on "EBAY", is worth two minimum salaries (the real yardstick); translate this into your local minimum salaries, whatever they are.
So let's help this fine chap, Lee Ranaldo, eh?
>>>>>>>>>>>>>>>>>>>>>>>>>>
UPDATE: just for kicks, I checked Argentine "EBay", I found
MARSHALL LEAD 100 MOSFET 100W + CAJA VALVESTATE 4X12 ENGLAND - U$S 1750.00 en MercadoLibre
where this CRAZY guy asks U$S 1750 for his Mosfet100 + 412 Valvestate cabinet.
Unbelievable? Just click on those links.
PS2: building a Plexi there will LOWER its market price .... because it will no longer be an original British Marshall but some homemade kludge.
Live and learn.

[MarkusBass]

Well, let's fix it then!

Yes, I also suggest to fix it. It should be quite easy under the condition that Lee is going to help. Since I still don't understand the issue with additional board, could you post a photo of the power amp board (top and bottom)?

... and it turned out that it now needed a Zobel network, whereas the original was happy without one.

I don't know about Markus' idea of using capacitors to snub the oscillation. To me, capacitances are what cause the parasitics, and resistors are what you need to damp them out.

It depends how a capacitor is connected (possibly with a resistor in series). A good example is the Zobel network - it would be hard to say that it causes oscillations. Another example are capacitors in a power supply (after a bridge).
I think that I tried adding a Zobel network and it did not help. Only capacitors solved the problem. But we may try several solutions.

Mark

[MarkusBass]

Three infos:
It seems that EXICON MOSFETs can be a direct replacement for HITACHI lateral MOSFETs: ECF20N25A - this is the same package and higher power.
It seems that the value of the R14 resistor I mentioned earlier was not correct. Value +200 Ohms was for EXICON transistors. In my amp (2SK1058) I have 33 Ohms which makes me wonder - I thought that the resistor should be increased. Anyway, it's easy to find its value. You have to get reasonable current through balast resistors - at least 5-10 mA. So you can replace the resistor with a pot, set the current and replace it with a fixed resistor.
It is possible (it was mentioned here by someone) that the amp shouldn't be used with 4 Ohms cabinet. "Older brother" (used in bass combo) which could be used with 4 Ohms had two pairs of output transistors. I would use this amp with 8 Ohms for the start.

Mark

[Steve Conner]

Hey Markus,

Yes, I've built amps using the 2SK135/2SJ50 and also the Exicon MOSFETs, and they perform much the same. The Hitachi MOSFETs do bias up with a very low gate voltage, only about 0.6V. I've worked on hi-fi amps using these devices where the markings were removed from the transistors to prevent copying, and I mistook them for BJTs.

The Exicons are available in a dual die version: two chips paralleled inside a single package. I think this is a worthwhile upgrade for any lateral MOSFET amp. The Magnatec BUZ90-whatever are also available in dual die versions. The one you recommended has a 16A rating, so it must be one of the dual die types.

Exicon, Magnatec, all lateral MOSFETS now available except Hitachi, all use the same dice inside.

Of course getting hold of any of these in Chile may be a problem, which is probably why the amp got bodged in the first place.

[MarkusBass]

Of course getting hold of any of these in Chile may be a problem, which is probably why the amp got bodged in the first place.

I think that Lee has the same option as I have - buy them for Profusion Plc . Maybe the delivery fee will be slightly higher. But, of course, since Lee has already 2SK1058/2SJ162, I would use them in the amp. He could use 2 pairs to allow the amp to work wth 4 Ohms speakers.

Mark