1982 JCM 800 2203 runaway bias? (picture heavy)

[GuitarIV]

So, where do I start?

I got myself a 1982 Marshall JCM 800 2203 about 3 years ago, fulfilling myself my teenage dream of owning one.

The amp was totally mint and serviced, had all the original tubes and filter caps and I got to open it up once with the preamp on 8 and the master on 6. Unbelievable experience. Sexual.

About half a year ago I was playing it at home and after half an hour the sound faded and a low frequency hum started fading in. I turned it off and phoned up my amp tech.

He came over, we fired it up and noticed that one of the EL34s was red plating.

Recently I finally had the cash to order a quartett of Tung Sol EL34Bs and a Bias-Master from TAD.

I am still a newbie in regards to tube amp electronics but I have my amp tech who helps me, I ordered some books he recommended and after I talked to him today on the phone and made sure I understood the biasing procedure I cleaned the table in my kitchen, took the amp out of the chassis and went for it.

Some pictures:




I was a bit nervous as I have never done this before, but my tech told me to take my time, not rush things, follow the "one hand in your pocket" rule and go for it.

After spending the better half of the day on this I can tell you it's not that hard if you're not an idiot. Respect electricity and be patient, then you're golden.



Now to get to the techy part of the thread where I need your input guys:

When I fired the amp up with the new tubes in the bias pot was turned up about 2 thirds of the adjustment range. Going with the TAD biasing guide the EL34s go from 25mA cold to 40mAs hot. That's the range.

Now I let the amp sit for 10 minutes, all the pots were turned down, I noticed one of the tubes in the middle was running quite hot. Somewhere around 46mAs. I turned the bias pot down a bit and got it to 40, after a minute or two it started rising up again untill I eventually ended up turning the bias pot all the way down to the end of its adjustment range. When that specific socket reached 47mAs I turned the amp off. The socket in question was one of the middle ones, I think V5 (when counting starting with the first preamp tube being V1 and the last poweramp tube being V7).

I turn the amp on again after a couple of minutes and suddenly the middle tube next to V5, so V6, starts rising up the most. The Bias pot was still all the way down.


After I let the amp cool down and turned it around so the electronics were facing the table downwards again (bias pot still turned all the way down) I fired it up and everything stabilized.

I had readings of 18, 21, 22 and 19mAs. It stayed stable.

I plugged the amp in, played it, it sounded good although a bit sterile which is to be expected with such a cold bias and my amp tech wrote me he'll call me tomorrow after I messaged him with my results.


Now I didn't want to call it a day and fired the amp up again. Bias was still stable with the same readings of 18, 21, 22 and 19. I started to adjust the bias pot up again and suddenly I had the middle tubes run away again!

Turned it off, let it sit, fired it up and this time I managed to slowly turn up the pot till I had something like 28, 32, 33, 29.

This was steady, after half an hour it stabilized somewhere around 32, 35, 36 and 34. (I'm giving numbers here from memory)

I plugged the amp into my Two Notes Captor and played it that way into my interface and my DAW and it sounded glorious! I did notice that the values would rise up to around 40mAs after having played the amp, but they would come back to the previous 32-36ish values after 5 minutes or so when I let the amp sit.


Now again I am no expert, but from my observations the following things come to mind:


- When the tubes run too hot the heat causes a resistor or cap to leak which then in turn increases the mAs and that could have caused the redplating of the original tube in the first place.

- it might be a mechanical issue (cold solder joint?) as the problem seems to be less severe or go away when I turn the amp around the "right way" as opposed to upside down

- it seems to be centered around the inner pair of the tube sockets as I have higher readings in the middle than the outside tubes. I might also just have a not so well matched quartett of tubes.


Now as of tonight the amp was stable with the readings of 32, 35, 36 and 34. I am tempted to fire it up tomorrow again, adjust the bias pot a wee bit down till the lowest reading tube is smack at 30 and call it a day.

However my instinct tells me there is an internal issue and if I don't fix that my bias will eventually run away again and it'll fry some of the new tubes. I don't want that.

I will be talking to my amp tech tomorrow anyway, but I figured some of you knowledgeable folks here might have a clue what might be causing this. The amp sounds great, I wanna have it running smoothly.



Sorry for the long thread, but I wanna be as accurate as possible. If you have any ideas, your input is very appreciated!

Cheers

p.s.: my amp tech developed a tremor in his hands, he can't work on amps anymore. I t would be a no brainer to have him fix the amp otherwise. This way he's helping me out and I might become the apprentice.

 

[william vogel]

Mark the tubes for their positions and write it down. Move the inside tubes to the outside positions, write this down. Power up and record the measurements. Be very careful to always record your changes. Sometimes new tubes have to burn in a bit to find stability. I’ve gotten crap tubes also and it stinks because not everyone can afford to buy 8 matched tubes at a time. I’ve had this happen before though. What is your plate voltage? I’ve found that high plate voltage 450-460+ is tough on new production Russian EL34’s.

 

[Chris-in-LA]

I had the same problem with my 2204, it’s not the tubes.

 

[Chris-in-LA]

I had the same problem with my 2204 and I took it to my tech so that he could diagnose it. He ended up replacing a couple of capacitors on the board and it has never had a problem since. You need to have someone look at this and check the values of the components, not just expect you to put new tubes in and bias it. I think your instincts are correct, you have a bad component somewhere on that board, it is not your tubes. Also, when you let the tubes red-plate, it is not good for them and you are ruining a set of brand new tubes.

calling @neikeel @mickeydg5

 

[Chris-in-LA]

I started looking at my amp. The last two red caps on the board let your signal pass through to the control grid on pin 5 of your power tubes. If those caps go bad they can leak power to the control grids causing the tubes to run away. You are probably better off running the amp without power tubes and monitoring the voltage at all of the pin 5’s. If that voltage doesn’t remain stable you will not be able to bias your amp. But I’m also hoping that some of our experts weigh in here.

There’s also the possibility that one of those black electrolytic caps went bad, they also work to control bias levels.

 

[South Park]

If those filter caps are the stock ones you most likely need replacing then are the tops pushing out if so replace them all . The bias caps might be leaking by .

 

[coolidge56]

Well there's definitely a problem, bias should not be stable then run away, first on 1 tube, then another, then both middle tubes. Swap the middle tubes with the two outside tubes. See if the problem follows the inner tubes to their new location on the outer tube sockets, if so its the tubes. If not then its not a tube issue.

The power tubes are fed signal in pairs, V4/V5 and V6/V7. Which might help narrow down where to look except you are seeing runaway in V5/V6 one of the tubes from each pair.

The negative bias voltage however is fed to all 4 power tubes from a single source. I'd pull all four power tubes and measure your bias supply voltage from cold to hot to see what range you get. I'd probably play around with that pot adjusting it while watching the voltage to see if there's any evidence the pot is gone wonky, bad spots on the wiper.

That said at 38 years old you are overdue for a filter cap refresh. The main can filter caps and those bias supply can caps are wear parts that will need to be replaced at some point.

 

[mixn4him]

At the least the bias caps should be replaced..

 

[neikeel]

The things suggested above are all relevant but you have to study and interpret your findings. We should assume that you bought a matched and pre-burned in set. Otherwise you may get drift at different rates as they burn in.
If the whole quad drifts you probably need new bias caps.
If a pair (v4&5 or v6&v7) drift it is probably a leaky pi output coupler allowing dc onto that grid (probably what Chris was describing).
If one tube only red plates in a quad you need to check that tube, typically label them and then just swap that tube with one of the lower reading ones. See if the drift follows the tube or just that socket. If it stays on the socket - could be cold joint (pin 5, or the 5k6 grid stopper).
I would also check the 1k screen resistors and the grid stoppers too, that is kind of routine before you even start to bias.
A good place to start is to measure the negative grid volts to ground on Pin 5 of each socket before installing new tubes and note the range you get. Then adjust for most negative before dropping the new tubes in.
The inner tubes do not share any features that the outer pair do not. Bear that in mind.

 

[Seanxk]

Bias caps as mentioned for sure.
As Neil mentions, the middle two EL34 are not a pair, v4 and v5 are a pair, v6 and v7 are the other pair, they should be matched to the OT sides.
Measure the ohms of the OT centre tap to pin 3 on v5 and write down, then v6 and write down.
The higher ohm reading, put your 35 and 36 valves in that side, that should even things out more.

 

[GuitarIV]

Mark the tubes for their positions and write it down. Move the inside tubes to the outside positions, write this down. Power up and record the measurements. Be very careful to always record your changes. Sometimes new tubes have to burn in a bit to find stability. I’ve gotten crap tubes also and it stinks because not everyone can afford to buy 8 matched tubes at a time. I’ve had this happen before though. What is your plate voltage? I’ve found that high plate voltage 450-460+ is tough on new production Russian EL34’s.

I had the same problem with my 2204 and I took it to my tech so that he could diagnose it. He ended up replacing a couple of capacitors on the board and it has never had a problem since. You need to have someone look at this and check the values of the components, not just expect you to put new tubes in and bias it. I think your instincts are correct, you have a bad component somewhere on that board, it is not your tubes. Also, when you let the tubes red-plate, it is not good for them and you are ruining a set of brand new tubes.

calling @neikeel @mickeydg5

If those filter caps are the stock ones you most likely need replacing then are the tops pushing out if so replace them all . The bias caps might be leaking by .

Well there's definitely a problem, bias should not be stable then run away, first on 1 tube, then another, then both middle tubes. Swap the middle tubes with the two outside tubes. See if the problem follows the inner tubes to their new location on the outer tube sockets, if so its the tubes. If not then its not a tube issue.

The power tubes are fed signal in pairs, V4/V5 and V6/V7. Which might help narrow down where to look except you are seeing runaway in V5/V6 one of the tubes from each pair.

The negative bias voltage however is fed to all 4 power tubes from a single source. I'd pull all four power tubes and measure your bias supply voltage from cold to hot to see what range you get. I'd probably play around with that pot adjusting it while watching the voltage to see if there's any evidence the pot is gone wonky, bad spots on the wiper.

That said at 38 years old you are overdue for a filter cap refresh. The main can filter caps and those bias supply can caps are wear parts that will need to be replaced at some point.

At the least the bias caps should be replaced..

The things suggested above are all relevant but you have to study and interpret your findings. We should assume that you bought a matched and pre-burned in set. Otherwise you may get drift at different rates as they burn in.
If the whole quad drifts you probably need new bias caps.
If a pair (v4&5 or v6&v7) drift it is probably a leaky pi output coupler allowing dc onto that grid (probably what Chris was describing).
If one tube only red plates in a quad you need to check that tube, typically label them and then just swap that tube with one of the lower reading ones. See if the drift follows the tube or just that socket. If it stays on the socket - could be cold joint (pin 5, or the 5k6 grid stopper).
I would also check the 1k screen resistors and the grid stoppers too, that is kind of routine before you even start to bias.
A good place to start is to measure the negative grid volts to ground on Pin 5 of each socket before installing new tubes and note the range you get. Then adjust for most negative before dropping the new tubes in.
The inner tubes do not share any features that the outer pair do not. Bear that in mind.

Bias caps as mentioned for sure.
As Neil mentions, the middle two EL34 are not a pair, v4 and v5 are a pair, v6 and v7 are the other pair, they should be matched to the OT sides.
Measure the ohms of the OT centre tap to pin 3 on v5 and write down, then v6 and write down.
The higher ohm reading, put your 35 and 36 valves in that side, that should even things out more.

Thank you everyone for your ideas and suggestions!


I should give you an update: I talked to my amp tech and we came to the conclusion that the bias pot might have oxidized, since I moved it around a few times and the amp stayed stable after I suppose that got rid of some of the oxidation film.

I took it to rehearsal with me on thursday and it sounded glorious, it was unbelievably loud with the master volume set to slightly above 1, I'm not sure if the new tubes are to blame but maaaan it simply leaves me awestruck every time I play one of these monsters.


My next step is to get some DeOxit, clean a few contacts and move the bias pot around a few times through the whole adjustment range. I will also be moving the tubes around to see if I get better readings and take some measurements.


I am new to this whole thing but I am eager to learn, I have enough books and materials and my tech who assists me.

I wanna do a cap job somewhere in the future when funds allow it and I am confident enough to do it; what brand would you recommend? I see a lot of people saying how you should keep the original caps in as long as they don't cause any problems and that new caps will change the sound; I also read about how the old ones are ticking timebombs if you leave em in.

When I bought the amp used the tech who sold it told me he reformed the caps and they do look good and the amp is quiet, however I'd rather not lose a transformer.


I wanna make sure the amp lasts another 40 years, these are a thing of beauty and they sure as hell sound amazing, especially with a proper Les Paul

 

[Seanxk]

Great news, the bias pot can be a problem to clean because of the closeness of its sealing, thus also they don't really get dirty. But I have had the same problem in the past with a pot that seemed to skip in places, I replaced it ( Piher 25k seems all that's available not 22k) and took the old one apart to find a burn dent on the track, just where bias was mostly set.

 

[Lance Chambers]

...........especially with a proper Les Paul

 

[coolidge56]

I wanna make sure the amp lasts another 40 years

Then have the re-cap done. German F&T 50 uF + 50 uF can caps are the gold standard for the main filter caps at $12 each. For the bias caps Sprague Atom Electrolytic Capacitor 10µF / 150V. Both available from Tube Depot.

For that wonky bias pot consider replacing it with a remote pot. On my JCM800 2204 I have relocated the pot from the board to the chassis so I don't have to remove the chassis to bias the tubes. That's a PEC locking 25k pot but less expensive locking pots are available. Those blue caps to the left are the F&T's.

 

[neikeel]

I still struggle.

The amp sounds great.

Your tech reformed and tested them properly.



Your amp, your money. Wait til you have a new cap go pop on you