77 Marshall 2203 Check-up

[BlackCrowe604]

Hello,

Me again with another 70s Marshall. A friend of mine asked me if I could take a look at his friends 77 2203. I found some interesting things... it looks like this amp has been on the bench a few times.

Looks like the OPT has been replaced for a current-production Dagnall C2668. The sticker looks very new as compared to the old ones, and it's condition is immaculate compared to the PT & choke. Can anyone confirm if Dagnall is making transformers today?

It currently has Svetlana EL34s installed. I checked the PV & Bias current using a bias probe.

Here's what I got: 527-529V B+ with all the controls set to zero. Bias current was 20.1ma and I could only increase it to 22ma with the bias pot maxed. This seems awfully low, and the voltage awfully high. At this voltage, shouldn't I aim towards 30-33ma?

I have found conflicting info on schematics. Where the white wire from the power transformer connects to te board, I see one 2203 schem with a 10k resistor on the cathode side of the diode, and another schem with 27k & (22k) in parenthesis. The actual circuit uses 2x 15k resistors (see below). I am not sure why.

Would changing the 15k resistor in series with the WHT PT  to 27k help my case at all?

Thanks in advance!!

[silverfox]

I would check the capacitor to determine if it is dried up and dropping the voltage. Beyond that, something is dropping the voltage and I don't think it's the tubes, as a guess.

silverfox.

[DummyLoad]

running a SIM on that bias ckt. with those values suggests that amp may have been equipped with 6550s. is that a possibility? SIM shows -60V with 22k pot fully CW and -50V with 22k pot fully CCW. if you change the 15k that connects to the white secondary winding to a 27k, SIM shows that the range changes from -47V with 22k pot fully CW to -37V with 22k pot fully CCW. that might be what you're looking for. OTOH, your tubes may be near end of life. have you tried a know good set? volt chart (attached) shows -50V bias with 530V B+. have you measured voltage at pin5 of EL34s?


--pete

[BlackCrowe604]

running a SIM on that bias ckt. with those values suggests that amp may have been equipped with 6550s. is that a possibility? SIM shows -60V with 22k pot fully CW and -50V with 22k pot fully CCW. if you change the 15k that connects to the white secondary winding to a 27k, SIM shows that the range changes from -47V with 22k pot fully CW to -37V with 22k pot fully CCW. that might be what you're looking for. OTOH, your tubes may be near end of life. have you tried a know good set? volt chart (attached) shows -50V bias with 530V B+. have you measured voltage at pin5 of EL34s?


--pete

THANKS DUMMYLOAD!!! After following your suggestion, you're right. Looks like the pair of grid leaks are 150k and bias range is 15k. That would have been appropriate for 6550's yes!! (I should have known.... I recently did a conversion from 6550 to EL34, but I blindly followed the recommended substitute values on the shcem).

I'll try the 27k bias range & change grid leaks to 220k. I'm looking at a 2203 800 schem and it has a 56k in series with the 22k bias pot for EL34 config. It has (47K US) in parenthesis I suppose for 6550. Can I get away with leaving this one?

[sluckey]

I'm looking at a 2203 800 schem and it has a 56k in series with the 22k bias pot for EL34 config. It has (47K US) in parenthesis I suppose for 6550. Can I get away with leaving this one?

I would not change this one just yet. Check your bias range after changing that 15K to 27K. If the desired bias current can be achieved with the bias pot about midway I'd leave the 47K in. But if the bias pot has to be set near the end of rotation, I'd change to 56K.

You want to decrease the negative voltage on pin 5 to cause the bias current to increase. Changing the 15K to 27K will do that. But changing the 47K to 56K does just the opposite (increases the negative voltage on pin 5). So, it 'may' take changing both resistors to get the desired bias current ***AND*** bias pot near center of rotation. But do the 27K first and check. Then do the 47K only if need to.

Maybe put a sticker on the chassis that indicates use EL34s ONLY. Bias mod required to use 6550s.

[BlackCrowe604]

I'm looking at a 2203 800 schem and it has a 56k in series with the 22k bias pot for EL34 config. It has (47K US) in parenthesis I suppose for 6550. Can I get away with leaving this one?

I would not change this one just yet. Check your bias range after changing that 15K to 27K. If the desired bias current can be achieved with the bias pot about midway I'd leave the 47K in. But if the bias pot has to be set near the end of rotation, I'd change to 56K.

You want to decrease the negative voltage on pin 5 to cause the bias current to increase. Changing the 15K to 27K will do that. But changing the 47K to 56K does just the opposite (increases the negative voltage on pin 5). So, it 'may' take changing both resistors to get the desired bias current ***AND*** bias pot near center of rotation. But do the 27K first and check. Then do the 47K only if need to.

Maybe put a sticker on the chassis that indicates use EL34s ONLY. Bias mod required to use 6550s.

Makes total sense, thanks Sluckey! I wasn't sure about the affect the 46/56k would bring if changed. Will try the 27k first!

I like the sticker idea too. Will give it a try!

Cheers!!!

[92Volts]

It probably wouldn't hurt to leave the 100k grid-leak resistors. New-production EL34s are known for being fragile, partly because these amps push them pretty hard. Unless you don't like something the 100k does to the sound (likely small if any change), it should improve tube reliability compared to 220k. The lower value stops grid current from messing with bias and causing runaway.