Marshall 3210 Mosfet head. Can't turn off the boost channel.

[bclemons]

I've got the footswitch in and still can't get the clean channel by itself unless the boost volume is all the way down. The footswitch works in my Marshall 5210 combo. I've pulled the chassis and the long tabs weren't touching the short ones with the switch in. I know I've ruined a jack now, but I even pried up the long tabs and put a jumper from the distal tab to the ground tab. It's still giving me the combined channels unless the boost volume is all the way down. Any ideas?

[g1]

Near as I can tell, the boost is turned off by pins 5 & 9 of IC2.
schematic: http://drtube.com/schematics/marshall/3210-iss7.gif

[dwmorrin]

I would try replacing the CA3046 if you're sure that you've got 0V At pins 9, 2, and 12. I've found that it's common for those to die and I've replaced a few of them in old Marshalls.

[bclemons]

Ok, thanks. Which ones are 9, 2 and 12, and where do I connect the other test lead? Am I looking for AC or DC voltage? Too bad the IC's aren't socketed.

[J M Fahey]

1) to check that the CA3046 is not doing its duty, turn boost hard off by grounding pins 5 & 8 .

2) check that pins 3, 7, 10, 13 are actually grounded.

3) to check that it's actually receiving proper commands, push footswitch on/off repeatedly and check whether you have 0V/0.7V on pins 2, 12 and 9.

Only after these tests you can pronounce it dead and search replacements.

[bclemons]

Alright, I've done #1.With the footswitch, I either get the clean channel or a barely audible distorted signal. If the clean volume is all the way down, The barely audible distorted signal is still there. I assume this is normal, as any clean volume would probably drown it out. Off to pull the board and check the soldering on 3, 7, 10, and 13.

[bclemons]

The soldering looked good on the the board. On pin 2 and 12, I was getting either 0 or .77v depending on the footswitch. On pin 9 it was .77 or the volts were slowly dropping.
But now I'm barely getting any output from the amp with the volume maxed. Did I screw something else up now? No weird smells or pops, I don't think I shorted anything out.

[bclemons]

NM about the output. in the process of moving the board, the yellow wire on the effects return jack came loose. I did manage to pull the shaft out of the reverb pot while removing the knob

[J M Fahey]

On pin 9 it was .77 or the volts were slowly dropping.

OK, this is the first abnormal thing I see.

Pinout is:


Either Q4 (inside CA3046, let's call it CA-Q4) BE is open or switching feeding it is defective.

To test its function, :
a) pull all jumpers I asked before
b) check that now we are on square 1: dirt channel works and can't be turned off
c) measure voltage at pins CA-11/4/6
d) now short CA3046 pin 9 to ground, what voltage now at pins CA-11/4/6 ?
e) does dirt channel now switch off?

Please test results and voltages asked.

As you see, we first test voltages and function, parts replacement comes at the end, if necessary.

No please.

Not effective, not even if you use one of these:

[bclemons]

Pins 11, 4, and 6 read 0v with or without the jumper on pin 9. The dirty channel was still active.

[J M Fahey]

Ok, just to leave no stone unturned, check that R10 still has +15V at the other end, although by now I guess it will, and then yes, the only option is CA3046 dead.

Hope you can find one.

If not, you can use actual transistors with legs bent and pushed through proper IC pads .

[Jazz P Bass]

Now we are getting somewhere.

Pin 4, 6 & 11 are tied to the +15Vdc supply through R10/ 22K resistor.
So find R10 & check it for an open (or bad solder/ connection).
You should also check that the +15Vdc supply reaches it.
3210 Ft Sw circuit.pdf

[bclemons]

Well, I get 15.88 volts prior to the resistor. The resistor reads 22.1k The solid link that feeds pin 6 from the resistor has no voltage on it, and the solder joint after the resistor looks a little suspect. I'm going to reflow it and cross my fingers.

[bclemons]

Now I am getting 0.3v on pins 4,6,and 11.

[Jazz P Bass]

Ground the footswitch input.