marshall jcm 900 4100 blew a fuse, which tube is opv(2 or 3)?

[Helmholtz]

Schematic of post #1 calls for EL34s (see heater arrangement bottom left)?

[g1]

Schematic of post #1 calls for EL34s (see heater arrangement bottom left)?

That schematic is the wrong one for his version of the amp. That's why he was having trouble getting the grid voltage to the range that is specified for only the EL34 version.

[Helmholtz]

That schematic is the wrong one for his version of the amp. That's why he was having trouble getting the grid voltage to the range that is specified for only the EL34 version.

I'm sure you're right.
But I'm confused by the inconsistent info provided so far.
Not even sure what tubes are used here.
For 5881s the grid bias should be about 10V more negative than with EL34s.
Grid bias should be measured at the test point (bias source) and not at the grids.
Results with the bias probe(s) indicate that at least one of the meters might be bad/unsuitable at low voltages.
The bias probes should be measured for resistance.

[georage]

The -32 to -48 came from the schematic i found on the internet that is "very close". The test point is in between the 150k resistors that go to the control grid. Thanks everybody for your support and for putting up with my errors.

[g1]

Those numbers are for the EL34 version only. Here is the 5881 version which shows -50V. (*USA version mentioned in the notes refers to EL34 version)

[georage]

helmholtz! i measured the negative voltage at the grids, not the bias point. i blew it! i smoked opv 3 and 4. the bottles are blackened and the amp was reported to have hummed and started to smell like burning. its back in the shop now.

thanks for the new schematic G1. I will print that out and compare to the one i provided.

this amp came with 5881s and there is a plate on the amp that says use 5881s.

I biased the amp using the ot method to 78ma per pair, and played the amp for 30 minutes at low volume and then checked the bias again. I didnt have a powerfull enough speaker to test at high volume so i asked the owner to use the amp at practice.
he said he played the amp for 30 minutes with no issues, then left it on for 2 hours and then jammed at practice when half way through practice it started humming and he shut it off.

To review, originally, the amp came in with a bad opv3 tube, a desoldered opv1 screen resistor and blown opv2-3 and mains fuses.
i soldered the screen resistor and replaced the fuses with slow blow .5 and 4amp fuses and replaced the tubes with psvane 5881s.

[g1]

39mA idle current (78mA per pair) is quite hot. These are 23W tubes, not 30W.
The '5881' tubes these amps were equipped with when new were Sovtek russian tubes. They had a reputation for being basically bullet proof. I would suggest using russian 5881 tubes and biasing as a 23W tube.
The psvane are reportedly chinese tubes made by some people who left Shuguang. In general I would not expect them to be as reliable as the russian tubes.

[georage]

23 watts hmmm. so the amp did come in with sovtek tubes in this case. So if i could ask you guys an elementary school question; I figure that 23 watts at 470 volts would mean the bias should be set to 48 millivolts, or actually 70 percent of that, right? so 34 millivolts?

[georage]

i set the bias to 39 maybe 40 millivolts, is that enough to red plate the tube and cause it to get smokey inside and melt the tube base?

[The Dude]

While that bias is hot, I don't think it would melt the tube base. I'm wondering if you are losing bias supply as the amp warms up. Those amps do have a fairly common issue with the coupling cap that couples AC to the bias rectifier (C12 or C15 depending on model). I wonder if yours is intermittent. I would remove output tubes and monitor bias voltage as the amp warms up and see if it remains stable. If not, I'd either check or replace that cap. You could also have simply had a defective tube. It's not that uncommon these days for new tubes to be bad.

[g1]

These amps were not biased at 70%, that would be the maximum but not recommended. From the factory they were probably around 60% tops.
The hot bias did not directly make the tube smoke, but likely caused the tube to fail, which made it melt.
In general, true 5881's were not designed to handle these kind of plate and screen voltages. The russian 5881's worked because they were actually a russian numbered tube similar to a 5881 and re-labelled as such. They are generally capable of handling higher plate and screen voltages.

[The Dude]

I don't see it in the thread (may have missed it). What is the actual plate voltage in this amp?

[g1]

I don't see it in the thread (may have missed it). What is the actual plate voltage in this amp?

In post #23 he mentioned 470V.
The psvane are a 6L6GB, not GC.
For 5881, I would only use sovtek in these amps, otherwise any 6L6GC should do. It was one of my gripes about marshall labeling the amps '5881' in the first place. Should have just used a 6L6GC labeling with those kind of voltages.

[georage]

"The hot bias did not directly make the tube smoke, but likely caused the tube to fail, which made it melt." that comment g1, thats what i dont have experience with; and Dude, im listening, i will try to get a pic up on the melted base. Both opv 3 and 4 smoked.

[georage]

should i be thinking about using 6l6gc instead? or do these amp have to have sovtek tubes? 470 volts is way over the datasheet voltage for both tubes