Marshall JCM900 Valve change and bias

[MARSHALL BUSTER]

Hi everyone
I'v picked up a Jcm 900 4100 100w high gain dr, and i'v ordered a new set of 5881's for her.

I'v recently revalved and biased my TSL122 that was very nice and easy to do, but the 900 is not so straight forward!! there is no conector on the rear of the amp to do this. After some reserch on tinternet i see there is a tool for my dvm to fit the valves into to do the job, but do i need four of them or can it be done with one??

Also there is a more round about way by conecting a resistor in the circut?? this way im very interested in knowing as it sounds like the cheapest way to bias the amp.

Plese can one of you 'tube pro's' explain step by step the process of biasing the amp? i want to be safe and not damage the amp so your advice is greatly appreciated.

Thanks guys take care
M

[Enzo]

Well, note that the factory schematic says - for the 5881 version - to set the bias VOLTAGE to about -50v. Much as guys love to set tubes by some current standard, pun intended, it is not necessary.

To set the bias by current, it is done the same as the other amp, just you have no haqndy test points here. Since ther is just one bias control, and presumably you are installing a set of relatively matched tubes, there is not a lot of point in checking all four at once. SO a single socket probe adaptor is enough. Pick one tube and do it there. Get the amp wher you want it to be, then if you like, you can move the probe to the other sockets to see that they all are working at about the same level.

There are basic socket adaptors you use with your meter. Therre are fancier biasing units that have meters built right in. Some of those have more than one adaptor so you can do it in pairs. I don;t remember, did someone make a four adaptor bias meter?


This amp has a couple of cathode fuses for the power tubes, one per pair of tubes. You COULD remove the fuse and connect a milliammeter in its place and take a direct current reading. You COULD remove the fuse and clip a 1 ohm resistor in its place and take millivolt readings across that. You COULD cut a trace or two and add a 1 ohm resistor in series with the fuse. Remember the readings you get that way will be for two tubes.

[booj]

I clip leaded a one ohm resistor between the cathode of one output tube and ground after removing the fuse for that tube set. Curious that t the fuses connected to either the two outside or the two inside tubes

[g1]

They do it that way so if one fuse blows, you can keep playing still having both "push & pull" halves of waveform.

[TimmyP1955]

Just put in 1 Ohm cathode resistors and leave them in.

(The JCM2000 bias setup is terrible IMHO.)

[TomCarlos]

Gents....

I know this is an old thread but it is related to my question. My neighbor's JCM 900 - 4100 has a set of Ruby 6L6GC tubes (not sure at this point if they are a matched set).

The pots and jacks need a good cleaning and I am giving the amp a look over. I was curious to see if I could find a spec sheet for the Ruby 6L6 tubes but no such luck. I would assume they come close to JJ or similar 6L6.

So the plan is to follow the instructions Enzo mentions in Post #2 - "set the bias VOLTAGE to about -50v" (intersection of R26 and R27 - that is shown on the schematic). I guess this is ok since 5881s and 6L6 tubes are fairly identical? I'll measure the plate voltage and bias current too just to make sure the bias current is within spec (will put a meter in the place of a fuse). I do not see the expected plate voltage on teh schematic? One other thing - I guess the measurements should be taken with the "Output Power Selector" switch set to "Full."

Anything else I should look for when using a 6L6 in lieu of the 5881?

Thanks, Tom

[Enzo]

5881 is a variant of 6L6, I wouldn't really look for anything different.

Ruby sells several 6L6 types, there is no one Ruby 6L6. They sell Sovtek, and JJ, and Chinese.

[Justin Thomas]

And in the case of these new Marshalls/Fenders/anything else, the "5881" they specify IS a 6L6GC, for all intents and purposes. Not to be confused with the NOS Tung Sol "genuine" 5881 of decades ago OR the reissues - those are 23W tubes. But anything Sovtek 5881 - generally marked 5881WXT, 5881/6L6WXT, is just a 6L6GC equivalent - 500V, 30W.

Part of me wants to ask to anyone with a tube question to always please include any suffixes, letters, numbers, etc... There's a healthy difference between a 5881 & a 5881WXT, or a 6L6 & a 6L6GC...

Justin

[TomCarlos]

Thanks Enzo and Justin....

This amp has Ruby Tubes 6L6GC STR.

I am adjusting the bias control, VR2, Full clockwise. The lowest I can go is -53.8 volts.

Tom

[Leo_Gnardo]

There was that period in the 90's when Marshall installed Sovtek wafer base 5881's in their amps. It's possible this 900 was built around then. Some of those Marshalls have cages around the output tube section and won't accept longer tubes but the wafer 5881 fits perfectly. They're still available from several sources including Ruby & are good tough tubes.

FWIW JJ have a 5881 as well, and I've found them to be an excellent tube as well. I treat them as "real" 5881 with a plate dissipation of 23 watts. They look smaller than stock 6L6's but I haven't tried them in space-limited Marshalls so can't be sure if they'll fit.

[TomCarlos]

Thanks Mates....

This amp has some issues and I will try to deal with them one at a time. When it was brought to me, it sounded ok for a few minutes but then you'd hear a crackling sound and then the output would cut out. Now, I have no output but do have signal at the Line Out Direct. However, that signal is extremely noisy.

First things first... I do have 485 volts at the Plates. I have -53.8 volts at the Grids. That seems to be a tad high. I cannot adjust it go get any lower. So I am hoping to use this thread to solve that problem.

VR2 on the schematic looks like a 47K pot. But the pot itself shows 22K on the side. I think I want to pull that board and the pot and check it. Let me do check that pot by lifting R28 and we'll go from there.

But assuming the pot is the correct value, would the fix to lowering the Bias voltage be to raise the value of R28?

Tom

[TomCarlos]

I think I need a Time Out. This amp is the JCM 900 100W Hi Gain Dual Reverb. I check a couple web sites for the schematic. But the PC board that is in the amp (PC0112) does not have the same value R28 and VR2 as shown on the schematic. R28 has a 68K and VR2 measures out to be 17K. I am searching other Marshall schematics but do not see anything else out there.

[pdf64]

...would the fix to lowering the Bias voltage be to raise the value of R28?

No, I think that increasing the value of R28 would tend to increase the magnitude of the bias voltage (and also reduce the range of the trim pot control).

[Enzo]

Think about it, -53v is too high? All it does it cool the tubes a little. Your problem is not distortion or lackluster sound, your issue is that it crackles and goes away. That won't be bias limits.

I recommend you find out what is wrong with the amp before modifying the bias supply.

The pot will only measure rated value when set to the end. TUrn it all the way CCW and see if it comes up to the 22k or close.

Don't worry that your amp and whatever drawing you have do not agree 100% on all part values. The power amp drawing you linked was ISSUE 14, that means at least 13 other versions of the circuit exist somewhere. And who says 14 was the last revision? This is just a bias circuit, a power supply, all you need is resistor values that provide a range of voltage that suits you. I have no idea if all my 14 are available online anyway. The circuits are likely all the same, just different resistor values.

You could lower R28 or raise R29.

[TomCarlos]

Understood Enzo.. I am just wondering how it is that I have different Resistor and Pot in this amp. I do not see any indication that these have been replaced.

For now, I cleaned the pots, jacks, tube sockets,swapped the 6L6 tubes, and the crackle is gone.

But there is still an issue to contend with. I am going to read the other MEF threads on this amp tonight and if needed, I will either piggy back on an old thread (if related problem) or open a new one.

Tom