A friend of mine is having amp troubles. Here's his description:
Got a Marshall JCM 800 using the 1987 circuit. User complained about the sound "almost cutting out" on loud notes. He had the low input of the bright channel linked to the high input of the normal, and ran the guitar into the high input of the bright channel. I listened to it in his studio, and the sound does choke up the harder you strum a chord (it's easiest to trigger with a D power chord).
The first time I saw this amp, the output tubes (GT E34L) were significantly imbalanced, with the bias of one tube reading in the 58mV range while the other was 30mV (measured using a bias tool, so it's a direct conversion from mA). I popped the chassis out, and found scorch marks from pin 3 to pin 2 on both sockets. So, the output of the power tubes was shorting to the heaters throughout the amp.
I replaced both sockets, replaced the power tubes with the same GT rating, and tested it at home. Now, I don't have a 4x10 sitting around, so my ability to really pound a 50W amp is limited. However, I did have each channel as loud as I could manage, and the amp sounded fine.
Cue the phone call a few days later. "Dude, it's doing it again."
Here are some measurements I took before digging in:
Tubes are GTE34L's, biased at 49mV (+/- a mV or two because no match is perfect)
Plate voltage is 478 VDC, grid voltage is -31 VDC, both measured to GND.
Incidentally, the screen resistors are 1K, so there shouldn't be any issue swapping E34L's in for EL34's.
If I'm doing my math correctly, the output of those tubes should be 25W apiece. Right at the stated 50W output of the amp.
So, my next step was to start swapping preamp tubes. I swapped in a known-good 12AX7, testing with it in each socket in turn. No change, D power chord still the killer.
After I got the amp back to my bench, I checked pins 2 and 7 of each preamp tube for voltage indicating a coupling capacitor leak.
Pin 2: 0V
Pin 7: 0V (not really surprising at all, since that's the way they should work)
Pin 2: 0V
Pin 7: 218 VDC (the b side of V2 is a cathode follower stage, though, so this is the same as the plate voltage of the a side)
V3-- (phase inverter)
Pin 2: 13 VDC
Pin 7: 13 VDC
And, again, pin 5 on both output tubes was -31VDC. The coupling caps don't seem to be leaky. However, I don't know offhand what the DC grid voltage of a phase inverter should be. Anyone got a bit of math handy to figure that one out?
I then went through and tested every single resistor for resistance in place. Everything checked out (accounting for the occasional parallel load screwing up my measurement). None of them looks damaged, burned, chipped, or feels loose.
The power transformer (when the amp is idle) measures 370-0-370 (VAC), I'm getting 485 VDC after the rectifier diodes, 478 VDC after the choke.
I have yet to pull the PCB out of the chassis, as the lead dress in this amp is very careful (and short) spaghetti. I'm going to get a little dental mirror (plastic handle/enclosure), and try to find any burns or bad solder joints on the underside of the board.
If you've read all of that, what am I missing? The amp seems to be behaving itself, except it clearly isn't. It's not a heat-related issue, it doesn't go away over time, it just sounds like shit.
My current possibilites:
1) Start pulling and testing each cap. (lame.)
2) See if the coupling cap from V2b/the tone controls to V3 really is shot.
3) Replace all preamp tubes with the assumption that the original pin 3 to 2 shorting screwed up every tube in the thing.
4)Something else, which you much older and wiser techs will educate me on, and I will feel silly for missing. Especially if it's one of the basic laws of physics.
What can I say, this one is humbling.