JCM 2000 DSL 100 ISS 5 - That Won't Hold A Bias - Change Diodes?

[Jecarroll7]

I couldn't get this amp to bias properly, been at it for a while.

After poking around tonight, I found that the PI plate voltage was randomly jumping (rising, holding then dropping) as well as spiking - both in the range of 25 to 50v. This circuit also feeds the screen grids (heavy watt resistors) of the output tubes.

The only thing in the circuit I see that might cause this is the bridge rectifier feeding the circuit (1n4007 diodes) so I thought I'll replace those and see.

Since I need to tear down the amp to get to them, I plan to replace them with the 3A 1000 PIV 1n5408 diodes since I have a bunch on hand.

I'll report back, just wondered if anyone else has seen this?

[52 Bill]

I found that the PI plate voltage was randomly jumping (rising, holding then dropping) as well as spiking - both in the range of 25 to 50v.

And what is happening to the bias voltage when the B+ is fluctuating?

[Jecarroll7]

Hey Bill,

Issue I'm having is tube runaway when biased above 20ma per side. Bias quickly jumps above 150ma per side and climbs with a loud screech through the speaker and I have to shut the amp down quickly.

After stepping through the screen and bias circuit resistors, I realized pulling the PI let me leave it up long enough to troubleshoot. A quick check of bias, grid and plate voltages looked okay with a meter. It wasn't until I put a scope on the PI plate supply that I noticed the problem.

Note that the amp originally came in with one of the 5watt grid resistors and a small ceramic cap toasted.

[Wes]

What about the possibility of the coupling caps on the PI plates causing this problem? I see a lot of (older) amps
that have leaking coupling caps that drop PI plate voltage onto the grids of the power tubes. Where you should
see negative voltage for your bias, you end up with positive voltage that causes situations like what you're de-
scribing.

It is something else to consider.

BTW- what is ISS 5 in the title JCM 2000 DSL 100 ISS 5? Is that issue 5?

I'm not familiar with the version numbering of those amps, but I know there has been many, many issues with the
bias supplies in those amps. I'm thinking it was mostly the earlier ones, but I wouldn't rule it out in newer stuff.

However, I don't think your problem is with your rectifier. I'd look at the bias supply and PI coupling caps.

Good luck with it, man...

[Jecarroll7]

I hadn't thought about that, will definitely check it out.

It's my guess it's ISS 5, next to the board number it has 5 squares marked ISS 6 through 10, which I thought might be set aside for future rework upgrades? I'm using an ISS 7 schematic which seems to be following the board pretty close.

The older ISS schematics are quite different.

[52 Bill]

The point of my question was to make you think about what would cause the B+ to reduce and the output tubes to redplate. Any loss or reduction of bias voltage would cause both things to happen. If only the B+ voltage dropped, I don't think that the tubes would redplate.

The additional information of the loud screech through the speaker points in the direction of oscillation, which could cause the current draw to be high enough for the outputs to redplate.

[Leo_Gnardo]

I couldn't get this amp to bias properly, been at it for a while. <snip> I'll report back, just wondered if anyone else has seen this?

Bias failure a major problem with this JCM2000 DSL & TSL. About 2006 they fixed the factory-install boards but up 'til then you have to replace big $$$ & time, OR be prepared to do some board surgery. (I could be wrong about the date. If anyone knows when they improved, let us know please.) If you go the board surgery route some very good techs in Germany will show you how in this link:

The Marshall TSL122 TSL100 thermal bias drift repair page

[Alex R]

As above, you may well have the heat-related bias drift problem. It is easy to test for. Clip a vm to pin 5 on a power valve and blow hot at at the board, in the power valve area. You will soon see if the bias voltage starts to drop.

The sure way is a new board.

The older boards that have this problem are a pale green colour with no weave visible through the surface. The new boards are fibreglass, a darker translucent green, and you can see the weave through the surface.

[Jecarroll7]

It's the darker green - newer type board.

I may not have mentioned this earlier, I pulled both the PI tube and output tubes and B+ was still jumping around.

Checking it tonight with those same tubes pulled, B+ is spot on and stable after letting it sit turned off for a couple days.

[Audiotexan]

Bias failure a major problem with this JCM2000 DSL & TSL. About 2006 they fixed the factory-install boards but up 'til then you have to replace big $$$ & time, OR be prepared to do some board surgery. (I could be wrong about the date. If anyone knows when they improved, let us know please.) If you go the board surgery route some very good techs in Germany will show you how in this link:

The Marshall TSL122 TSL100 thermal bias drift repair page

Excellent read, and thanks for sharing =D

[Enzo]

B+ jumping? If your mains are 120vAC, and your B+ is roughly 480v, that means there is a four to one ratio. For every one volt your mains change, your B+ will change 4v. If your mains are anything like mine, they jump around all the time. So check the mains, if they are moving around, then so will your B+. A 5v mains change becomes a 20v B+ change.


But it could be in the amp too. Monitor the current draw the amp makes from the mains, is that bouncing too, or is it only the B+ voltage?

B+ goes up and down easily as current draw on it changes. With the PI and power tubes gone, there are few things left to draw current. Disconnect the OT primary CT from the B+. ANy change? If that left the B+ steady, then you either have OT trouble or maybe arcing.

[davetutt]

Morning

Lots of problems with this amp many due to leaky printed circuit boards which you can either fix by removing parts of the board and hard wiring or paying for a replacement board from Marshall. There is no quick fix here. All the TSL100 and 50 amps suffer with it so it includes all the amps with the raised metal panel in the centre of the chassis that is removable carrying the output valves and the ecc83's. If you dont know what you are doing a new board is required otherwise its about a days work using a dremel and removing the pcb around pins 4 and 5 of all the el34's, hard wiring away from the bases and if possible removing the long tracks going back to the driver coupling capacitors as well as those that run to the biasing network in both directions. You should also replace with higher wattage metal film resistors all those in the bias network and the coupling caps with ideally 22nf 1600volt film caps. This applies to all boards from rev 3 - 7 from what I can tell the 5 being the worst. The amp will get worse as it gets older. The epoxy in the pcb is almost fully conductive (I've had down to 10k cm) below the surface and in a very hot position under say the el34 bases it causes drift and reliability issues and is very thermally sensitive. Many arrive at my workshop with full HT on the bias pins of the output valves so what should be around -40 volts end up with +450 or more and the valves totally collapsed. There is no pcb burning or visible damage except where the heater current for the valves enters the board which is not the problem with the amp. The total effect of these problems tends to be mains transformer failure if left (the owner usually changes the fuse to a higher value until it stops blowing!) all output valves too damaged to be recoverable and a stressed and noisy ecc83 driver which also has to go. When the transformer fails it can take out other things too like the channel switching as an example.

If you do this work you will notice a change for the better in the sound on the clean channel and far more reliable bias setting.

Hope this helps. There are write ups on what to do with pictures at any number of web sites but if you have one of these amps I would suggest you check the rev version of the valve board and if it a 3 or 5 have the work done soon. The mains transformer from Marshall is 86 UK pounds and a set of output valves around 70 so you won't see much change from £280 if you need these changed as well. These are a heat sinking amp to work on as you never know if you have got it all!

[davetutt]

One further thing. The clean tone is due to incorrect component values and poor components generally. The resistors in the output stage all drift and should be changed to at least 1watt metal film types in the bias network. The grid resistors on all the output valves have up until the much modified replacement boards been 220k half watt carbon types. These should be 5k6 1 or 2 watt types located off board on long legs ideally. The circuit diagram is correct the auto insertion machine in the factory totally wrong. You will need the long legs if you do the pcb work as you can solder the resistors directly to the valve base pin.

If your amp looks much the same as the one in this link then it will need the work carried out soon.
http://www.lynx.bc.ca/~jc/TSL122.html

[davetutt]

Morning

Lots of problems with this amp many due to leaky printed circuit boards which you can either fix by removing parts of the board and hard wiring or paying for a replacement board from Marshall. There is no quick fix here. All the TSL100 and 50 amps suffer with it so it includes all the amps with the raised metal panel in the centre of the chassis that is removable carrying the output valves and the ecc83's. If you dont know what you are doing a new board is required otherwise its about a days work using a dremel and removing the pcb around pins 4 and 5 of all the el34's, hard wiring away from the bases and if possible removing the long tracks going back to the driver coupling capacitors as well as those that run to the biasing network in both directions. You should also replace with higher wattage metal film resistors all those in the bias network and the coupling caps with ideally 22nf 1600volt film caps. This applies to all boards from rev 3 - 7 from what I can tell the 5 being the worst. The amp will get worse as it gets older. The epoxy in the pcb is almost fully conductive (I've had down to 10k cm) below the surface and in a very hot position under say the el34 bases it causes drift and reliability issues and is very thermally sensitive. Many arrive at my workshop with full HT on the bias pins of the output valves so what should be around -40 volts end up with +450 or more and the valves totally collapsed. There is no pcb burning or visible damage except where the heater current for the valves enters the board which is not the problem with the amp. The total effect of these problems tends to be mains transformer failure if left (the owner usually changes the fuse to a higher value until it stops blowing!) all output valves too damaged to be recoverable and a stressed and noisy ecc83 driver which also has to go. When the transformer fails it can take out other things too like the channel switching as an example.

If you do this work you will notice a change for the better in the sound on the clean channel and far more reliable bias setting.

Hope this helps. There are write ups on what to do with pictures at any number of web sites but if you have one of these amps I would suggest you check the rev version of the valve board and if it a 3 or 5 have the work done soon. The mains transformer from Marshall is 86 UK pounds and a set of output valves around 70 so you won't see much change from £280 if you need these changed as well. These are a heart sinking amp to work on as you never know if you have got it all!

I would question if they are financially viable to repair given that a new reliable valve head amp might be say £600 800 and given their age and the fact that you might want to change all the valves at the same time and that I am a cheap repair person where others would charge probably £400 for all the work and parts involve.

[Jecarroll7]

Fixed it, but didn't expect the OT to be the problem.

As it turned out, the B+ cleaned up after I replaced C38+C39 and the 4 rectifiers that feed them. But that didn't solve the problem of a loud fog-horn sounding noise from the speaker as I turned up the bias voltage past 25mv per side. As suggested I also replaced the 22nf 630 coupling caps with 1600v versions, but no change.

I built an 8 ohm dummy load and monitored the output voltage at the load. With the dummy load I cranked the bias to 65mv per side (B+515vdc) - no DC present across the load, and the ac signal would zero out with the volume controls. Scoping the entire circuit from input to the PI to the load - I didn't see anything out of the ordinary. I pumped in a 100hz and 1000hz audio signal into the input jack, and it looked fine at the load. I run guitar tuning frequencies into the input and cranked the volume up to 25vac - signals still looks fine at the load. So at this point bias is set and the volume is up, everything looks good.

So I disconnect the dummy load, and connect the speaker cab - wham I get that fog horn sound again out of the speaker cabinet. Different cab, same thing. I was able to put my scope on it quickly at the output jack. looked like a steady 100volts AC at about 600-800hz and when it kicks in the volume control is useless, only way to get rid of the noise is to kill the power.

After chasing this thing for months off and on, I finally tried replacing the output transformer this weekend. It's been playing a couple days without issue.

I've never had an output trans fail this way. Any insight as to why? or what was happening internally?