Marshall JCM800 Bias Problem

[gibson17]

Hello all,

Thank you for taking the time to read my post. I have a Marshall JCM800 Lead Series 50W head model 2205. It is having a bias runaway problem. When I set the bias, the amp seems stable, but after a few minutes it starts to climb at about 1 mv every minute or two. A few times, I tapped on the bias pot and it seemed to come back into spec, but I have been unable to replicate that as of late. I checked the solder connections and they all seem fine. I know this amp is getting old, perhaps some of resistors are going out of spec and need to be replaced? Thoughts?

[Jazz P Bass]

If you have a another pot at hand, replace it.
If not, when it is working, measure the resistance that the pot is applying to the circuit & temporarily replace the pot with an equal value resistor.

[big_teee]

If you have a another pot at hand, replace it.
If not, when it is working, measure the resistance that the pot is applying to the circuit & temporarily replace the pot with an equal value resistor.

+1.
Change the Pot.

[gibson17]

Thanks for the replies. I've ordered a new trim pot and will let you know the outcome. Have a great day!

[52 Bill]

As a temporary fix, you can try cleaning the trimmer with a quick shot of Deoxit.

[gibson17]

Thanks Bill, I did already try that. I've since removed the pot since the iron was hot. Now I wait =)

[Jazz P Bass]

Did you not measure the pot where the bias was set good?
With a resistor as the pot you could have verified the fix.
As it is you are guessing, no?

[gibson17]

Jazz, I am guessing. I was unable to get a stable reading from the pot, that coupled with the fact that it changed values while setting bias when I tapped it, I figured the pot was bad. Too late now as I removed it. When I was removing the pot, the side with the single leg seemed to just fall apart... perhaps this was part of the cause?

[Tom Phillips]

...When I was removing the pot, the side with the single leg seemed to just fall apart... perhaps this was part of the cause?

Interesting. The way the Marshall bias circuit is implemented the bias voltage would become more negative if the bias adjust pot wiper or the other legs lost good contact. This would have the effect of reducing the bias current which is the opposite of what you observed.

[gibson17]

Tom, for clarification purposes, when watching the bias with the amp at idle, once the bias began to climb if I tapped on the pot, the bias would drop back down into range. I was later unable to reproduce the same phenomenon by tapping on the pot. When I say fell apart, the leg broke as soon as I touched it with heat applied. I guess I'll find out what's what when the new pot arrives. Thanks again for the help!

[Chuck H]

FWIW, and if it hasn't been done, replace the bias supply filter caps too. If they haven't been replaced they're old and feeble. Not to mention that this facilitates resoldering more stuff. Which brings me to my second suggestion... When installing the new pot, go ahead and check component values, replace as needed and re-flux and meld every joint in the bias supply. Right down to the feed resistors and grid stops. No harm can come of it. The bias supply may be the most important circuit in an amp. And they're not hard to maintain.

[gibson17]

The bias pot arrived last night. I installed it, let the amp warm for a few minutes and set the bias to 38mv and walked away as I was packing up other equipment for a gig. When I came back, I noticed the bias was rising, so I turned the bias as low as it would go, which was around 30-31mv. I then set a timer for 20 min and continued getting ready. When I returned it was at 34.5mv. Before the pot change the bias was getting up to 50mv at idle from 35mv before I would shut it down. I decided to see if that trend continued and set the timer again for 20 min. When I came back the bias had settled to the original 30-31mv and was stable. I then set the bias back at 38mv and it's stable after another 20min. Thoughts?

Chuck, no work has been done on this amp since factory, save for my fiddling, which isn't much, an input jack and now the trim pot. Someone had said to me that I might need to replace some caps or resistors in the bias section because of age, that's why I mentioned it in my original post. I don't mind replacing anything, but my knowledge is limited. Which ones are the bias supply filter caps? I haven't reflowed any of the solder joints, I took a look at them with a magnifying glass and didn't see any broken or cold joints. I'll reflow while changing the caps. Thank you for the help to you, Jazz, Tom, Bill, and Big.

[Chuck H]

On most Marshall amps the bias supply filters are the two 10uf, 50V or 100V caps very near the trim pot. They will be parallel to each other and have the "+" side grounded. It's possible that many caps should be replaced in that amp. But at the very least, since you were working on the bias supply anyhoo, the bias supply caps should have been replaced. Electrolytic caps go bad with age and not always analogous to usage. The average life of an electrolytic cap is ten to twenty years I suppose. Though some perform poorly after a few years and others might last for fifty years. Still, with something like a bias supply I would be inclined to replace any electrolytic cap over fifteen years old "just because".

[km6xz]

Just a point of interest. Reflowing solder is not a good idea. The flux is gone, there is a high possibility of a worse connection reflowing old oxidized solder so remove the old solder and apply new rosin core solder or use flux with non-rosin core solder. When the solder is removed inspect the pad and the lead to see if there is any corrosion before re-soldering. When soldering be sure to heat the joint to solder melting temperature before applying solder to the joint.
It is common practice to "wet" the tip of the iron with a little fluxed solder to aid in heat conduction and reducing the corrosion rate of the tip. When done the joint should look shiny and bright and not dull grey, and it will look very smooth and the solder concave, not bulging with a convex surface. If it is convex, there is a good chance the lead and pad were not fully wetted by the solder. That results in weaker bonds between the lead and solder, air pockets or gaps and corrosion.
To see tiny cracks in joints(the big cracks are easy to see with the unaided eye), you can put a drop of Isopropyl alcohol on the joint and touch the lead from the component side. Even the tiniest crack can be seen if you watch for ripples in the reflection of a bright light. Cracks much smaller than is visible directly clearly show as the surface scattering on the alcohol. Another way is to lightly lay the pads of your fingers on the foil side of the board while touching components on the other side. You can detect slight movement that you could never see even with an inspection magnifier.
There was some confusion when you were using "bias" and cathode current interchangeably. Measuring bias is a voltage measurement of the grid voltage. Measuring idle current is a current measurement of the current flow through the cathode.
Good luck with the amp

[gibson17]

Thanks Chuck and KM. I have a bias probe in the tube socket, so that's what I mean when I say MV at Idle. I know just enough to be dangerous. That being said, I do know that the voltage inside can kill me and I know where not to touch, not to mention I made a power bleed with resistors on a drum stick to ground, so don't worry too much about me Thanks for the soldering tips, I thought reflowing was a good thing for a time or two. I understood it is not a good thing when doing it over and over again, making the joint brittle. I'll apply the tips you shared, again thank you!

Even though the amp remains stable, I think I'll go ahead and replace all the caps as Chuck has recommended. Do I need to change the giant caps that hang down next to the power tubes by the tranny?

Thanks again and have a great day!