New amp… sort of… er…

ICBM

Not sure whether I should admit this, but here goes! Long story but I have to explain myself .

I seem to have accidentally acquired a Marshall TSL122 - the combo version of the TSL100 head. Yes, I hate these amps and I never have a good word to say about them.

Late last year an old friend got in touch about it - he'd taken it out of storage with the intention of selling it, tested it and it promptly melted the power valves. (They do that, occasionally.) So he asked me to fix it. I fairly quickly worked out that it wasn't the usual bias connector problem but was the more serious board leakage and bias runaway problem. The only real way to fix this normally is to replace the main PCB - which you can get from Marshall but isn't cheap. Add a new set of valves and labour (including fixing the footswitch, which had a broken cable) and it was going to cost him well over £200 to repair, which was not ideal given that he just wanted to sell it. So I had a bright idea - by a slightly complicated swap of some other stuff, I bought it off him for the scrap parts value, since I'm sure to need them sooner or later for other ones.

But somehow I just couldn't let it lie, could I… it was actually in very nice physical condition and seemed a shame to scrap it, so I thought I would use some spare time over the Christmas holiday and see if I could fix it *without* changing the PCB - just as an experiment, really.

So having cut out and bridged all the areas I could find which might be causing the problem, changed the bias supply components and still not cured it, I had a bright idea - why not convert it to cathode bias? That should automatically stabilise it. Or more accurately, combination fixed/cathode with enough cathode bias to compensate for the runaway. Fender did the same thing in the late 60s to cure a similar problem, although the result doesn't have the best reputation. So I did - I replaced the 1-ohm bias sensing resistors with cathode resistors and bypass caps, which has made it stable no matter how long it's on.

The real surprise it that it now sounds *much* better - much fuller and chunkier, particularly on the crunch channel. So good, that I actually like it… a lot. I'm not sure exactly why. So I have decided to keep it, at least for now!

Partly because I can't actually sell it yet either - in the mean time another fault has shown up, which is that the clean channel starts to hum when it's been on for a while, but I think that's probably just a filter cap on the way out. Grrr. So I still hate it too .

ecc83

IC, I Have long thought that there should be a simple, cheap way to simulate "auto bias" with an op amp and MOSFET and have even done some back-of-envelope schematics but it always turns out vastly more complicated than I am sure it need be!

What value is the cathode resistor and what is the voltage across it? Or, are you seeking a patent?!!

Dave.

sgosden

you'll have a DSL100 as your 'go to' amp before 2017 ends

ICBM

ecc83 said:

What value is the cathode resistor and what is the voltage across it?

50R, 5W. Cathode cap is 22uF/25V. At idle the resistors develop 3V (at 30mA per valve) and at full tilt 13V which is a dissipation of just over 3W. I have of course set them about half an inch above the caps! This is enough to prevent any rise in valve current while still allowing the trimmers to set the bias in the right range and avoid the need to mod that part of it as well.

ecc83 said:

Or, are you seeking a patent?!!

Hardly, I nicked it from the Vox AC50, the '68 Fender Twin and you know where else .

sgosden said:

you'll have a DSL100 as your 'go to' amp before 2017 ends

Not unless someone gives me one in a way I can write it off as scrap parts .

Actually both amps share the same main PCB so the fix would be exactly the same.

MtB

Cue the sudden rush for, and subsequent price hike in used TSL122s and DSL100s 

Voxman

ICBM said:

I seem to have accidentally acquired a Marshall TSL122 - the combo version of the TSL100 head. Yes, I hate these amps and I never have a good word to say about them.

So having cut out and bridged all the areas I could find which might be causing the problem, changed the bias supply components and still not cured it, I had a bright idea - why not convert it to cathode bias? That should automatically stabilise it. Or more accurately, combination fixed/cathode with enough cathode bias to compensate for the runaway. Fender did the same thing in the late 60s to cure a similar problem, although the result doesn't have the best reputation. So I did - I replaced the 1-ohm bias sensing resistors with cathode resistors and bypass caps, which has made it stable no matter how long it's on.

The real surprise it that it now sounds *much* better - much fuller and chunkier, particularly on the crunch channel. So good, that I actually like it… a lot. I'm not sure exactly why. So I have decided to keep it, at least for now!

Mate, I'm so sorry to hear about your breakdown culminating in the delusion that a TSL can sound good - I had no idea you were under so much pressure from work.  I'm sending the men in the white coats round, & they'll take good care of you, I promise.   

Bridgehouse

I don't have the faintest idea what any of you guys are talking about - but it sounds brilliant so  

jellyroll

TLDR. "Bought a broken amp. Fixed it."

ICBM

Voxman said:

Mate, I'm so sorry to hear about your breakdown culminating in the delusion that a TSL can sound good - I had no idea you were under so much pressure from work.  I'm sending the men in the white coats round, & they'll take good care of you, I promise.   

lol… trust me, it's quite different now. A friend came round and said he'd never heard one sound like that before either!

I'm puzzled as to why though - it sounds like it's got a lot more resonance at the bottom end, as if the NFB has been reduced or removed, but I don't see how combination-biasing would do that, by itself. The presence controls still work so I don't think the NFB feed has got broken in the dozen or so times I must have had the PCB out.

I did also solder up the speaker connections to get rid of the crappy push-connectors Marshall use (and so I could be really OCD and turn the speakers round so the logos are the right way up ), but I can't see that making any difference really, they were actually making contact before.

Bridgehouse

jellyroll said:

TLDR. "Bought a broken amp. Fixed it."

Ah! Now I get it.


Carry on  

Gassage

Mods,

Someone's hacked @icbm 's log in.

ICBM

jellyroll said:

TLDR. "Bought a broken amp. Fixed it."

Nah, that would be boring. This is "Bought a broken amp. Modded it and accidentally made it sound quite different. Now need to work out why."

Bridgehouse

ICBM said:

jellyroll said:

TLDR. "Bought a broken amp. Fixed it."

Nah, that would be boring. This is "Bought a broken amp. Modded it and accidentally made it sound quite different. Now need to work out why."

Sometimes I realise that I'm actually an inadequate average guy...

jpfamps

ICBM said:

Voxman said:

Mate, I'm so sorry to hear about your breakdown culminating in the delusion that a TSL can sound good - I had no idea you were under so much pressure from work.  I'm sending the men in the white coats round, & they'll take good care of you, I promise.   

lol… trust me, it's quite different now. A friend came round and said he'd never heard one sound like that before either!

I'm puzzled as to why though - it sounds like it's got a lot more resonance at the bottom end, as if the NFB has been reduced or removed, but I don't see how combination-biasing would do that, by itself. The presence controls still work so I don't think the NFB feed has got broken in the dozen or so times I must have had the PCB out.

I did also solder up the speaker connections to get rid of the crappy push-connectors Marshall use (and so I could be really OCD and turn the speakers round so the logos are the right way up ), but I can't see that making any difference really, they were actually making contact before.

Degenerate feedback will reduce the transconductance of the output stage and hence reduce negative feedback.

Also cathode biasing increases the output impedance from the valves which will reduce damping.

Richardj

Well that ^^, obviously.  

Time for more, or less, 'heavy' methinks..... (not going on any ill founded stereotyping of course).

cbilly22

Yes yes, but does this mean that your Peavey is up for sale now?

ThePrettyDamned

So...

Is this a repeatable and reliable mod? Or rather, could it be repeatable assuming you find it reliable? 

A high gain, reliable small Marshall would be an awesome home fun amp... 

clarkefan

Ok, sounds better, but would you describe the improvement as, "more Marshally"?

Kalimna

Simillar to one of the other posts, if this fix/mod turns out to be stable, would you offer it as a 'standard' mod?
Having said that, i dont currently own a TSL122....

Adam

ICBM

jpfamps said:

Degenerate feedback will reduce the transconductance of the output stage and hence reduce negative feedback.

Also cathode biasing increases the output impedance from the valves which will reduce damping.

Interesting… I need to learn more about theory for this sort of stuff.

I understand what you're saying there, but would it have that much effect with only fairly minimal cathode-biasing? I've only added enough to provide 7% of the total bias voltage at idle, up to about 25% at full output.

cbilly22 said:

Yes yes, but does this mean that your Peavey is up for sale now?

No, but it may mean that my Trem-o-verb is up for sale. Yes, I am serious. I love the T-verb but it's really more amp than even I need, it's worth a lot more to sell, and actually the Marshall sounds better at low volume, I think. And it really is heavy, even for me. I don't need more than one 100W valve amp. (Actually I don't even need one…)

ThePrettyDamned said:

Is this a repeatable and reliable mod? Or rather, could it be repeatable assuming you find it reliable?

In theory, yes. The actual mod is easy, it's just removing two resistors and replacing them each with a cap and a much larger resistor which is piggybacked above it.

My concerns are - 1, that the resistors do dissipate a fair bit of heat when the amp is cranked up, which given where they are in relation to other parts may be an issue. I did wonder about chassis-mounting some higher-rated ones but that was too much of a faff at this stage. And 2, that the HT fusing is currently inadequate to stop the resistor burning out if a valve shorts - the 1A fuse would allow each 5W resistor to receive 50W before the fuse blew if a valve developed a full short. It would also exceed the voltage rating of the cathode cap I used, but that's easier to fix.

ThePrettyDamned said:

A high gain, reliable small Marshall would be an awesome home fun amp... 

It's not that small or that high-gain, actually! And the most-improved sounds are the clean and the crunch, the lead is still a bit buzzy (although still better).

clarkefan said:

Ok, sounds better, but would you describe the improvement as, "more Marshally"?

If you mean that big chunky open roar of a real old Marshall, like a 50W JMP into a 4x12", yes. It also sounds somewhat more 'Mesa-ish' - I've actually always loved old Marshalls, but always wanted them to do things that Marshalls can't quite - I run them with the bass up full, and the mids usually fairly low… like a Mesa. In many ways, for me the Trem-o-verb is the Marshall that Marshall never made.

Kalimna said:

Simillar to one of the other posts, if this fix/mod turns out to be stable, would you offer it as a 'standard' mod?
Having said that, i dont currently own a TSL122....

Definitely not without further testing and better protection for the new components, as above. I'm happy enough to run it myself for the moment, but I'm not actually convinced I would gig it, yet. It is still a TSL after all .

chromatuna

For some reason I am finding this thread hugely entertaining. ICBMs surprise at what he has achieved plus the somewhat mysterious nature of it I think. :-)

Edit: there really should be a punk band called Degenerate Feedback