Marshall 9000 amp Channel B Way too hot

[mikeskory]

Well it's my 1st post. Hopefully I can edit it if not understandable. I am working on a Marshall 9000 stereo amp (only). We started out with nothing. The customer acquired it from a flea market and said it just hummed loud then popped now dead. I looked it over and easily saw 2 power caps were blown open. There are 4 the same. Dual section caps with 50mf each section. I soldered some scrub caps and heard both channels play with no fuses popping. So I ordered 4 nice caps from Studio Sound Electronics. they fit fine and look awesome! But after leaving it of for a while the channel B runs very hot. Way louder then channel A and also the EL34s on the B channel were emanating lots of heat. Tried tubes....same. I checked & checked and finally decided it was the output trans. The guys at Classic Tone helped me pick one out. Of course after I installed it the problem was exactly the same. I have the schematic but could use a directional plan to trouble shoot this amp. 4:27 Update. I'm starting from scratch. I removed all 4 EL 34s and checked DC voltage on the socket. Pin 1 0v, pin 2 450v, pin 3 450, 5 450 vdc, 6 0vdc, 7 0vd, 80vdc. Pin 4 has a discrepency. Channel A its -47vdc on both tubes at pin 4, Channel B it's -.03 Vdc on both tubes at pin 4

[Enzo]

Hi Mike...
Transformers don't usually make things run hot. My first reaction was where is the bias set? I am assuming that one has bias adjust for each channel? Monitor mains current and see if dialing the bias down some decreases mains draw. Or use a bias probe on that channel output tubes.

Or go back a step and pull those power tubes, and verify you have good B+ at both pins 3 and 4 of each socket. then make sure you have reasonable bias voltage at pins 5. I would guess about -445v, but if the other channel is running OK, use it as a rough comparison.

I'd expect bias because the signal level won't really affect heat. But in general, if we apply the same signal to both channels, like with a Y cord, we can trace it through, Is the signal at the phase inverter grids stronger on the loud channel? Or not?

[mikeskory]

Remember I owe you lunch. Yes I updated my original post. I pulled all 4 power tubes. Here's what I have. Update. I'm starting from scratch. I removed all 4 EL 34s and checked DC voltages on all 4 sockets. Pin 1 0v, pin 2 450v, pin 3 450, 5 450 vdc, #6 0vdc, #7 0vd, #8 0vdc. Pin 4 has a discrepancy. Channel A its -47vdc on both tubes at pin 4, Channel B it's -.03 Vdc on both tubes at pin 4.

There is a PR1 SET BALANCE and PR2 Set BIAS for each channel. On the good channel they are -40 to -50vdc range. On the hot channel is -036 VDC at all 4 spots. 2 for each PR pot

[Jazz P Bass]

Here is the schematic: http://drtube.com/schematics/marshall/2001pwr3.gif

Track down why you do not have any bias voltage on Ch B.

[The Dude]

A its -47vdc on both tubes at pin 4, Channel B it's -.03 Vdc on both tubes at pin 4

You are missing bias for channel B. You need to figure out what's wrong with the bias supply. This might help (if it's the right one).

http://drtube.com/schematics/marshall/9005.gif

Edit: <simulpost> Jazz beat me to it!

Edit (again): Remove the tubes from channel B until you get the bias fixed on the channel.

[Enzo]

Make sure to count socket pins right, pin 5 should have the -47v. The pins count the opposite direction when viewed from the top.


Lunch one day would be great. I am glad you got to meet RJ, we all have to stick together to keep Lansing playing.

[mikeskory]

Thank you guys! I found it! KNB1530 .047m polypropylene rectangular capacitor. I soldered in a blue .047 regular poly and it plays normal. I'll add the KNB1530 to my current Mouser order, install it, bias the tubes....and that should be that. 2:26 pm Last thought...Im thinking the output trans was still bad. Prior to installing it, my light bulb tester would light up like a beacon if I engaged the bad channel via the stand by switch with the power tubes OUT. The only way I could calm down the light bulb was to unhook the trans from the circuit. If took about 15-20 seconds.
After the new one was installed, the tester acted normal with the trans wired in. Thank again

[Enzo]

Hey, it is Thursday... I thought you only worked three days a week.