JCM2000 DSL100 Problem

[Ant]

Hi all,

Mate of mine has just dropped me off his faulty Marshall DSL100 head.

Had a quick check and the pre-amp looks to be ok (took send output to a seperate amp and the signal is fine). Just after taking the chassis out and found that both R6 and R9 (1R 1Watt) looked suspect as the paint looked to have peeled off these resistors (must have lit up). Meter checked proved both are O/C. I'm not up to speed on valve theory as I'm a TV repair guy. Both these resistors feed through to pin 7 (K) and pin 1 (G3) of the 4 power tubes. Any of you guys seen this fault before? Appreciate any checks you would suggest before I go diving in - something must have caused these boys to go high.

[Amp Kat]

Once again looks like another victom of a tube short and the 100 watt schematic is not the right one. Think you need the 5o watt schemo but D5r.Tube site is down right now. Those resistors are the current sense resistors going to pins 1 & 8 of the cathodes on the power tubes to ground. They measure the current that biases the power tubes using the pots PR1 & 2 on the back bias board to set the neagative grid voltage. Each side has it's own bias voltage so they can be different. The last amp I did that had those blown also blew open the plate voltage trace underneath the board. Is there spare sockets on each side of the power tubes. If so measure from pin 3 to other pin 3 on each side and see if you have continuity.

[Ant]

Many thanks AmpKat.

This is a 100watt head so there are no spare sockets either side of the tubes. I did check for continuity between pin3 of 1st and 2nd tubes and its ok. Also checked for continuity of pin 3 between 3 & 4 power tubes and they're ok too (Output tx is disconnected so I didnt check for continuity between all 4 but I'm sure they'll be ok.

I take it it's a new set of power tubes that are required (along with replacing R6/R9)? Out of curiosity, is there any way to check for a tube short just to see if I can find the offending tube?

Finally can you remind me of the bias setting procedure for fitting new tubes?

[Amp Kat]

Many thanks AmpKat.

This is a 100watt head so there are no spare sockets either side of the tubes.?

Somehow I keep forgetting that when it's 100 watts it's 4 tubes so kick me in the head.

I take it it's a new set of power tubes that are required (along with replacing R6/R9)? Out of curiosity, is there any way to check for a tube short just to see if I can find the offending tube?

Finally can you remind me of the bias setting procedure for fitting new tubes?

If you know anyone who has a tube tester or maybe take it to a local repair shop and ask them to test them for shorts.

There should be a board in the back that has three pins sticking out of it and two very small trimmer pots pr1 and pr2. First read your plate voltage on pin 3 then put your meter on DC mili-volts and put the black lead on the middle pin and red lead on the left pin. then get your wattage of the tube and calculate % dissipation. I use 70% if it's out of crossover so lets say 25 watts X 70 % = 17.5 watts so 17.5 / 475 (plate volts) = 36 ma's or .036 amps. X 2 tubes =72 milliamps. So across the 1 ohm resistor should be 72 milli volts or .072 volts. Tweek pr1 to 72 milli volts the put the red meter lead on the other side leaving the black where it is (ground) then do the same using pr2 and your done. I would go back after 20 minutes of warm up and do the procedure again as they may drift a hair after burn in. If a tube goes into thermal runaway (red plating) at start-up you still have a problem and I'd shut it down then post back. Good luck!

[Ant]

Thanks for the feedback Amp Kat.

I've a matched quad set of Mullard EL34's on order and should have them by tomorrow. Will get back to you with the results.

[drewl]

[QUOTE=Amp Kat;28242]and the 100 watt schematic is not the right one. QUOTE]

...What do you mean not the right one?
Is there a wrong one?
Got one with the same problem....tubes died and the resistors fried....

[Alex R]

I'd put in new tubes anyway. Marshalls in particular & EL34 amps in general do this kind of stuff - every sign of thermal runaway having occurred, fuse blown, burned cathode sensing or screen resistors etc, yet the valves still check out ok on the tester. Clearly they're not ok so I change them.

On older Marshalls that do this persistently (and some do) I do the raw-bias-to-suppressor grid thing to stop it happening again. Does anyone do this to PCB Marshalls?

[Amp Kat]

Drew, Some of these 100 watters use the 50 watt schematice so if your mapping out components and the numbers on the scematic don't jive with the PCB then you'll have to go to the 50 watter schemos. If they match your cool. I've seen some 50 watters that use the 100 watt pcb and some don't.

Alex, On another thread on this forum someone mentioned that there is an underated cap and increasing it from 500vdc to 1kvdc is recommended.
http://music-electronics-forum.com/s...27435#poststop

I would use flyback diodes on the plates to keep it from smoking everything that it does and of course what your doing is a good thing too. The tube I checked was bad but then again this was a major short that opened the trace up and that's not a thin trace so there could be another problem causing this. It is happening alot to these amps but aren't they a few years old ? We have a brand new JCM 2000 in the shop and I haven't gone over it yet but I did notice the bias jacks are gone from the back and the pots. You know and I'm not blaming Mr.End-user here but these problems seem to appear more frequently with exterior bias jacks than interior bias jacks. If I were an amp builder for the public I'm not sure I'd give that option to someone that doesn't know how to correctly bias an amp. But on the other hand it's extremely beneficial to someone who does so your kinda damned if you do and damned if you don't.

[Dave Curtis, dB AudioTech]

Somehow I keep forgetting that when it's 100 watts it's 4 tubes so kick me in the head.


If you know anyone who has a tube tester or maybe take it to a local repair shop and ask them to test them for shorts.

There should be a board in the back that has three pins sticking out of it and two very small trimmer pots pr1 and pr2. First read your plate voltage on pin 3 then put your meter on DC mili-volts and put the black lead on the middle pin and red lead on the left pin.

The black lead can just stay clipped to ground...

-DC

[Alex R]

Oh yeah one of those little caps often goes... not thinking of the possibility of flyback voltages so much higher than normal plate I usually replace them with what is in the drawer, in this case 22p 600v silver mica, rather wastefully... must get me a range of cheap 1kv ceramics.

Isn't there a modern production amp, not a Marshall that has the anti-flyback diodes in as standard? Fender? I seem to remember finding one with a diode burned out... no idea why.

No I was just idly wondering if anyone had found a way of doing the raw-bias-to-pin8 thing on PCB Marshalls without too much hassle, lifting of too many traces etc. I guess it can be done with a jumper and a trace-scrape. I've never bothered trying but it works so well for me on p-to-p Marshalls so maybe I should. I had a couple of old Marshalls with the old OTs still in that kept on coming back. First I put in screen resistors (they were both old enough not to have any at all), then cut the currents down to 25-30ma; and still I got 'third time I used it it hummed and the plates glowed after 3 hours playing'. Bias to pin 8 stopped it so I have a naive, primitive, experientially-grounded faith in the procedure. But then I think lightning means the clouds are angry.

[drewl]

Where is that little cap (which c number) it's fried on the 2000 I have on the bench, but the amp's at home and I'm at work!
R9 is also fried, but so is R8....what value is that please, I can't find it on some of the schems.

And finally, how about fusing pin 1/8 to prevent future meltdowns?

Thanks all.

[Alex R]

on a tsl60 it's c26 and c27. Goes from plate to screen. Near the screen resistors on the board.

[Ant]

Hi all,

I replaced R6 and R9 (1ohm 1watt) and fitted a new set of matched power tubes. Switched on and everything looks good so it looks like amp cat was right (no surprises there) that there was a tube short.

As I have said before, I'm a novice on working on amps and still finding my feet so I have a few questions:

1) the guy who owns the amp bought the new set of EL34's and gave them to me. He bought new Mullard EL34's and they came with a "79mA" sticker stuck on the lid of each box. Surely that can't mean 79mA per tube? I checked the pin 3 volts and it was 465 volts and that works out about 38mA per tube times 2 = 76mA per side which isn't too far off the "79 mA" on the box. Can anybody enlighten me as to what the bias should be?

2) to me these tubes seem to be running really hot but I'm comparing this finding on my own amp which is a DSL 50 watt. Maybe I'm being a bit paranoid due to the biasing uncertainty I have with item one above, after all the DSL 50's don't have vents on the top of the amp housing but the DSL 100's do. So is this normal?

[Alex R]

Hi Ant. The 79mA sticker is just a tester result. Valve dealers like to stick these on to show that valves are matched etc etc. It is very misleading (I mean to thoughtful purchasers like yourself) to add something like 'mA' to the numbers though, as this only relates (and probably tenuously) to some artefact of the tester, and is in truth done just to make it all look more technical. It's a way of 'adding value' as retailer gurus say. Like our milk is better than the other guys' milk because it has additives to stop your kids bones from crumbling, and so on. Or our chocolate is good because it's sexy chocolate. Maybe I'll start selling valves and get some stickers saying 'tone factor: 2 megs'.

Marshall EL34 amps usually bias between 30 and 40mA per valve as a rule of thumb.

[Amp Kat]

If you want to bias to about 70 % than 25 (if 25 watt EL-34) watts x .7 = 17.5/465 vdc= 37 ma if you want it a little hotter or more out of crossover than go to 80% or 43 ma's. The 1 ohm resistors are basically fuses for pins 1 &8. That's why they don't use 2 to 5 watters and they also use metal film flameproof resistors.