Marshall 2103 Cold Maximum Bias Mystery

[Enzo]

I find that an anode current of around 28-30ma as a starting point sounds good, but I've no objection to going higher or lower.

It sounds good? Then why bother changing it?

[greengriff]

I mean on 100w Marshalls in general I have found this to be a good starting point. On this amp in particular I am unable to bias it because of the problem outlined in this thread.

[greengriff]

So the circuit is different from this: https://drtube.com/schematics/marshall/78331-3.gif ?

That's the right one, and clearer than it is in the Doyle book. Good find, thank you.

[pdf64]

The JCM800 versions get their bias feed from the HT winding rather than a bias winding. So the resistor on the cathode end of the rectifier diode has a bigger resistor. 220K for EL34, 150K for 6550. I think this is what Chuck was referring to.
https://drtube.com/schematics/marshall/jcm800pw.gif

(I think we're all assuming there is a typo in the thread title and this is a 2203, not 2103).

The 70s onward JMP 100W models used a seperate bias winding too.
From https://drtube.com/library/schematic...ll-schemas#JMP
2103 JMP Master Volume Lead, 100W 2x12" combo
This amp was in production from 1976 up to 1981. This is a 2x12" combo version of the 2203 head.
https://drtube.com/schematics/marshall/2103-pic.jpg

[greengriff]

That's the right one, and clearer than it is in the Doyle book. Good find, thank you.

Actually it's not quite right, in that it shows a number of fuses that aren't in the amp, unless they are out of sight under the PCB?

[Enzo]

SOme places require all secondaries to be fused, other places - like the USA - do not. SO in amps destined for USA, the fuses are replaced by wire jumpers on the board. No point in drawing two schematics for that detail. Some schematics will have a note to the effect "not included on USA models" or similar.

[greengriff]

SOme places require all secondaries to be fused, other places - like the USA - do not. SO in amps destined for USA, the fuses are replaced by wire jumpers on the board. No point in drawing two schematics for that detail. Some schematics will have a note to the effect "not included on USA models" or similar.

Thanks for clearing that up.

[Randall]

Isn't it possible that that set of tubes came that way, as in they are on the low current scale and would run cold in most any amp? Have you tried another set?

[greengriff]

Isn't it possible that that set of tubes came that way, as in they are on the low current scale and would run cold in most any amp? Have you tried another set?

I don't have another set of 4. However I could put 2 of these in my 2204 and see what they pull in there.

[Helmholtz]

Is there a formula I can use to work out what your suggested cathode current means in anode current terms?

According to the EL34 datasheet, screen current is around 4mA. So 34mA cathode current means 30mA plate/anode current.
Are you sure, your bias probe measures anode current only?

And please measure (grid) bias and screen voltages (see post#2).

You don't need to replace all 4 tubes. Just exchange 2 and measure idle currents.

(To avoid misunderstanding and to clarify terms: A tube/valve needs a bias (negative grid-to-cathode voltage). This bias voltage controls the anode and screen idle currents (cathode current is the sum of anode and screen currents). So a typical "bias probe" is actually an idle current probe. Most measure cathode idle current but some actually measure only anode idle current. With the latter I recommend pulling the PI tube before measurement to exclude possible errors caused by oscillation.)

[Chuck H]

If your bias supply is taken from the HT winding then the bias feed resistor should NOT be 27k for UK models. It should be 220k. It is R8 in the schematic that Helmholtz posted. Which is the one I looked at for model 2103 and you indicated it was basically correct. Please locate R8 from the schematic in your amplifier and confirm that it's 220k.

[Chuck H]

Also, the piggyback resistor is almost ubiquitous here in the US where tens of thousands of 6550 Marshalls have been converted to use EL34's. Care must be taken not to unsolder the existing resistor by overheating the leads because that does happen. There are three ways to do it.

1) Just solder the add on resistor and hope for the best. Usually even if the board solder gets melted it will just set up again without issue, but not always.

2) Just like above, but take extra care by getting your iron about 14°C hotter before you start. Then pre tin the leads of the resistor to be added and get in and out very quick. This usually doesn't melt the Marshall board solder.

3) If you have a pair of small nosed needle nose pliers you can wrap a rubber band around the handle to to make them want to hold closed. Then you can clip them onto the existing resistor lead very near the board This should sink enough heat to prevent melting the board solder. If you have a small nosed hemostat that's even easier. The only caveat is that there isn't always room for both the heat sink and the new solder joint.

[Helmholtz]

If your bias supply is taken from the HT winding then the bias feed resistor should NOT be 27k for UK models. It should be 220k. It is R8 in the schematic that Helmholtz posted. Which is the one I looked at for model 2103 and you indicated it was basically correct. Please locate R8 from the schematic in your amplifier and confirm that it's 220k.

Confused.
If the amp corresponds to the schematic I suggested (https://drtube.com/schematics/marshall/78331-3.gif), there is a separate bias winding and R8 is shown as 27k.

[Chuck H]

Confused.
If the amp corresponds to the schematic I suggested (https://drtube.com/schematics/marshall/78331-3.gif), there is a separate bias winding and R8 is shown as 27k.

Ah! OK, my bad Helmholtz and greengriff. I was looking at the 50W model when I determined the value for R8 And didn't notice that the 100W has a dedicated bias winding. Please proceed gentlemen. I'll just try to pay closer attention

[greengriff]

Thanks for the detailed input everyone. I'm busy all day today but hopefully I'll get some time tomorrow to run through all the suggestions and report back. I've ordered 240k and 270k resistors anyway (the one I connected with crocodile clips temporarily was a 250k guitar volume pot so not exactly viable as a permanent solution ;-)

@Chuck thanks for the tips re: soldering in the piggyback resistor. It's going to be a tricky job as there's not a lot of the original leads sticking out of the boards, what there is is partially covered with plastic that's bled over from the resistor body, and I don't have particularly steady hands!