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

[georage]

A guy brought in a jcm 900 4100 with a blown main fuse and i might really need some help on this one. i replaced the main fuse, found that one tube would cause the dimbulb to grow bright and start to strobe, and revealed a blown power valve led(and blown 500ma fuse) in OPV(2-3) position.....but i have no idea what tubes that refers to. anyone? do you know? I think i can use this schematic to determine that the middle two tubes with the green and orange wires and no 1k5 resistors are( v4 and v7), and those have their cathodes tied together. Would that and v5 and v6 be a pair? Each tied cathode pair connects to one of the OPV failure lights. once i can name the correct tube position I can ask about a few other issues. Thanks

[g1]

Each fuse controls one 'push' and one 'pull' power tube. Of the 4 power tubes, opv 1&4 are the outer pair, opv 2&3 are the inner pair (as physically located on chassis).
So V4 and V7 would be OPV 1&4, V5 and V6 would be OPV 2&3.

[Jon Snell]

Always replace the complete set of output valves and bias accordingly when one valve fails, the others are not far behind.

[georage]

great! that is confirmed except that 2&3 are the middle....it looks like v4-opv2, v7-opv3, v5-opv1, v6-opv4. ok so it blew a fuse on v4 and v7.....i found a loose screen resiator and extreme low bias on v5. AND v7 or v6 had a bad tube in it....most likely it was v7. the tube in v7 was a little low on emissions but other wise fine, just different than all the rest.

with v7andv6 tubes in place the dimbulb grows in intensity and started to strobe after a minute. with fresh tubes in those positions the and maintains a healthy_looking power draw.

im not sure why v5s screen resistor would just come loose like it was never soldered.

[g1]

You can easily find out which is OPV 2&3 by installing only a pair of power tubes, and removing 1 OPV fuse.
Or follow the OPV fuse to the cathodes.
I doubt they would mislabel the fuses, more likely the schematic has wrong names on tubes.

[georage]

alright guys. I found the bad tube and soldered the screen resistors back on and put a fresh quad of tubes in and now i am trying to bias this amp and im running into trouble with it. I would really appreciate some advice. I set up some under-tube bias probes on v4 and v5 (opv1 and 2). they measure amps not volts. Im also measuring the voltage at the bias test point on the schematic. I should be reading -32 to -48 volts at the test point and 38ma at the probes. what im getting is -40v at the test point and 22ma on the probes. in order to get to 38ma, I have to turn the bias adjuster pot all the way to maximum. that makes the test point read over -50v. thats not right. i dont get it. then there is an odd problem that may be caused by my two separate meters reading the individual probes, but, lets say at 22ma both meters read 22ma, but when the pot is turned all the way up, tube 1 reads 38 and tube 2 reads 27ma. i need to run the procedure again with only one meter, but i think there is a problem or im not doing it right. any ideas??

[georage]

I just thought of something! I was running the amp with the dimbulb in series...maybe that is causing the low bias. I should run the whole bias procedure again with one meter and no dimbulb.

[g1]

The more negative the bias voltage test point is, the lower the tube current should be. What you have described sounds backwards (-40V gives 22mA, -50V gives 38mA).

[georage]

yeah youre right. i dont know...im trying to set up a better experiment. i set up a new under tube probe with a 10 ohm resistor in series with the leads and read mv with one meter, with no dimbulb....and got 38, 36, 29, 40 ma. on opv 1 through 4 and a negative bias voltage of -58. is that normal? it seems off...i need a better method of reading the bias, these under tube probes are no fun at all.

[Greg_L]

yeah youre right. i dont know...im trying to set up a better experiment. i set up a new under tube probe with a 10 ohm resistor in series with the leads and read mv with one meter, with no dimbulb....and got 38, 36, 29, 40 ma. on opv 1 through 4 and a negative bias voltage of -58. is that normal? it seems off...i need a better method of reading the bias, these under tube probes are no fun at all.

Use the OT resistance/voltage drop method.

[g1]

Tell us more about your set-up with the bias probes. You said they measure current rather than millivolts? They have worked fine in other amps and still do?
Are you using 2 meters, one set for current and one set for volts? Otherwise it would be a bit of work switching the probes back and forth to measure the bias voltage and then the tube current.

Aside from that, the range of your bias trim pot sounds more like the 4100's set up for 5881's rather than EL34's. Do all the bias circuit components match the EL34 schematic?

[georage]

thisbamp takes 5881s for starters. i will explain my set up, but im thinking that the bias probes may have been the problem. i had 2 probes under opv1 and 2 with an ab switcher to a amp meter. then i put a 10 ohm resistor aross the bias test leads and tested the voltage drop with a voltmeter. that was the upgraded set up...the previous was no abbawitch and too different meters reading current. so all that was terrible and the problem followed the tube...or maybe the probe!

[georage]

so i got out a new push pull amplifier and and tested all the tubes it pairs and some more known good pairs of tubes and recorded the bias. and all my quad came in between 44 and 50ma in my setup. so i put the quad in the jcm900 and biased using the ot method where i measure 15.4ohm or so on each side of the ot. then dialed in the voltage drop across the ot to about 1.07 volts and played it through a speaker for 10 minutes. it sounded good...so now idk clean up and run that experiment again and record the numbers for real....and see if it blows up ...i dont trust bias probes at all now.

[Helmholtz]

1.07V across 15.4R means a current of 69.5 mA for 2 tubes. That's a plate current of around 35mA per tube. Should be fine.

[g1]

I should be reading -32 to -48 volts at the test point

thisbamp takes 5881s for starters.

I'm not sure where the '-32 to -48 volts' reference comes from, but the 5881 version of this amp calls for -50V at the test point.