Marshall TSL 122 positive bias voltage

[Danglin' Fury]

So Leo Gnardo, I did the "pin 5 isolation" technique that you sent in that article.... yyyyyeah.... I have "OL" on my meter from pin 4 to pin 5 on all of my power tubes, but I still have +150 volts DC on pin 5 of V5 and V6. What the...? Any other ideas? That is with "CON1" and "CON12" disconnected. And when I flip the "Standby" switch, there is a full plate voltage on pin 5.

[Helmholtz]

Did you perform the complete mod as described in the article and shown in the pics?

Can you post pics of your mod?

[Helmholtz]

I have "OL" on my meter from pin 4 to pin 5 on all of my power tubes, but I still have +150 volts DC on pin 5 of V5 and V6.

Does that mean measured between socket pins 4 and 5?
What resistance do you measure between socket pins 3 and 5?
What is the max. resistance your meter can read?

[Danglin' Fury]

I only use Flukes.... It can read Meg ohms for sure. On V5-V8, they all measure 1.295 k ohms between pins 3 and 4 on the power tube sockets with tubes removed. Is it possible that the 82k (R3) in the PI is going down in resistance as it heats up? It measures 82k when cold... Hang on... I noticed after I turn it on, the voltage starts out around 50 volts DC on pin 5 of V5 and V6 and it climbs up to 138 volts.... I turned the amp off and immediately measured R3 (82k) and it is reading 90 ohms and climbs back up to 82k. I think we have a winner.

[Helmholtz]

PI plate voltages don't affect power tube grid voltage as the coupling caps C6 and C7 block DC (i.e. if not leaky).
There is a possibility that the board leakage only develops at high voltage and elevated temperature, so won't show with an Ohmmeter.

[Danglin' Fury]

Well shit. I would think that the bias board might be the culprit, but the high voltage is on pins 5 of V5 and V6 even when it's disconnected... Hell, maybe C6 is leaky... I have one of those laying around, I think I'll change it just because.

[g1]

If pin5 is disconnected from everything, coupling caps from PI will not make any difference.
When you say pin 5 is disconnected, you mean the socket pin has been cut free from the board, and you are measuring that 150V at the socket pin 5?

[Danglin' Fury]

Yes. They are completely isolated from the board. I Dremeled a hole around them and soldered the 5k6 resistors directly to pin 5 of the socket per the mod Leo Gnardo had as a link in his post. I followed the mod to the letter. After I changed R3 and C6, I now get on pin 5 of the V5 and V6 sockets, at power up, the voltage starts out as -17.45 and climbs to -8.7 voltage, but now , at least, the negative voltage is adjustable via the bias pot, only it doesnt get anywhere near what it needs to be.

[g1]

Sorry I wasn't more clear. I was trying to ask whether you had reconnected any of the parts or if socket pin 5 was completely disconnected. That was to eliminate any possible socket problem.
So I would suggest you disconnect the 5K6's and recheck the socket pin 5's.

[The Dude]

......There is a possibility that the board leakage only develops at high voltage and elevated temperature, so won't show with an Ohmmeter.

That is usually the case with these amps.

[Greg Robinson]

My experience is that the mods involving cutting slots are only sometimes effective, and then not for the long term. It's not just the grid circuit where the board leakage is a concern, but also the bias rectifier circuit etc. I liken it to the whole board sitting in salty water - drying up just the area around pin 5 isn't going to fix the rest of the board being conductive...
Don't know what the source of the information that "the boards are out of production" is - they're still available from CEDist at least, I haven't checked stock elsewhere.

[Danglin' Fury]

Ok G1. I will disconnect the 5.6k resistors and check the voltage.

[Danglin' Fury]

Hey Greg. I hear ya about more than that portion being conductive, but the thing that has me buggered, is V7 and V8 have a pretty solid bias. V5 and V6 did too before the incident that caused R63 to smoke and the HT fuse to blow. V7 and V8, those voltages are good and solid. It's only V5 and V6 that are having the problem. I was experiencing+138 volts on pin 5 of V5 and V6 until I changed R3 and C6 yesterday. Now, upon power up, I'm showing -17 volts, and it goes up to -8 volts. I can take it higher and lower with the bias pot but it is not where it needs to be, obviously. If the board around the rectifier was conductive, would it not also affect the bias voltage on V7 and V8?

[Greg Robinson]

Sorry, don't have time at the minute to re-read the whole thread to check what's been done already, but I'd be checking for voltage drop across: the grid stoppers R66 and R7, the grid leak resistor R67, the voltage across the bias filter cap for V5&6 C36, the voltage at pin 1/7 of con1, and double checking the wiper of PR1. If you have good bias voltage range for V7&8, it must either be a bad wiper for the bias pot, or leakage at one of the other points listed.

[Leo_Gnardo]

I'm pleased to see you gave the Hullerum method a go. AND a bit baffled why bias voltage hasn't been entirely sorted out yet, though it's a big step in the right direction.

I just recalled a dim memory from having dealt with this - a 7 picofarad cap spanning plate to grid that is known to also fail and become an unwanted current path between those elements. If that is still there, prime suspect. Nip it out and let's see if things don't improve a bit more, maybe to where they should be. If it isn't replaced, and I haven't replaced only removed 'em, it seems to have no negative effect on the amps operation to have it absent. Besides, I came up short seeking 3000V 7pF caps.

The bias board is no winner either. I've found trimpots to be defective in some units.

Seems this amp is snakebit right there in the output amp. Once sorted, actually a quite versatile and good sounding amp although I'm not a fan of its overcomplexity.

Having nothing to do with the annoying bias problem, another bugaboo of this series is a ground connection between output jacks that depends on the jack normalling contacts to deliver that ground to one set of outputs and not the others. Set up for failure once again Marshall, how thoughtful. That ground isn't made, and your amp's not connected to a load and it will object as tube amps do. Not a good thing. I take that long board out, with the output jacks & all the rest of the battle rattle, solder a sturdy jumper over the break in speaker jack ground path, and reinstall. No dodgy connections to speakers after that.

OK enough memory ransacking for one day. There's a nice easy Deluxe Reverb to work on next here, about time I got to it.