Identify a Mysterious Wire...

Si_

Just took some internal pictures of my late 70s JMP 2203 and noticed a strange white wire.. Can anyone identify what this is actually doing?

It's the wire connected after the resister just about the "m" of Marshall. It seems to come off one of the pre-amp tubes.

martinw

It feeds the grid of V2A. It looks unusual because it's a screened cable. Normally there would be a green wire going from the top of the PCB near where the cable tie is fitted.

Obviously the intent is to reduce noise or prevent oscillation, but I've never had  a problem with the standard amp in this regard, so I can't see why someone would go to that trouble.

Misterg

Looks like someone has replaced the wire to the grid of V2A with screened wire...

I can't tell from the photo, but is the resistor 470k ? (Yellow violet yellow ?)

If so, the same pin of V2 would normally be connected to the top of that resistor (via the PCB). The bottom of that resistor is connected to ground, so I guess that someone has used it as a convenient place to ground the screen of the wire.

Compare to this pic:

http://i433.photobucket.com/albums/qq60/charveldan/100_0299.jpg

Possible reasons would either be an attempt to reduce noise, or to avoid instability / oscillations.

If it ain't broke...

Misterg

D'oh - there were no replies when I started typing! Now it looks like I'm copying you @martinw !

martinw

Great minds and all that

Si_

Cool, Thanks

ICBM

martinw said:

It feeds the grid of V2A. It looks unusual because it's a screened cable. Normally there would be a green wire going from the top of the PCB near where the cable tie is fitted.

Obviously the intent is to reduce noise or prevent oscillation, but I've never had  a problem with the standard amp in this regard, so I can't see why someone would go to that trouble.

They do quite often self-oscillate there. The Marshall factory fix is to stick the wire to the chassis with a spot of white silicone-type glue - usually that's enough to suppress it. In misterg's pic you can see where the glue used to be, although like most amps that have been serviced requiring the board lifted, the glue is broken. (And less usually, it looks like the V1B plate wire has been stuck there as well.)

It looks like Si's amp never had it in the first place, or it came cleanly off the chassis - either way, it's been fixed better by using shielded wire.

Some later ones have an extra bypass cap on V1 as well - despite the relatively low gain, these amps are fairly close to self-oscillation normally. The layout is not exactly ideal!

martinw

ICBM said:

They do quite often self-oscillate there....

....... The layout is not exactly ideal!

I've never had one that needed attention, but that said, if I'm doing any mods on these, I usually go round and shorten wires and can end up with enough wire to build another amp! It can make quite a difference.

On my own amps I shield grid wires as a matter of course.

ICBM

martinw said:

I've never had one that needed attention, but that said, if I'm doing any mods on these, I usually go round and shorten wires and can end up with enough wire to build another amp! It can make quite a difference.

I think this is largely why these amps are so variable. They're not only built very haphazardly, the layout is inherently poor because it's derived from the non-MV version by reconnecting the existing PCB layout arse-backwards! There are other variations and factory tweaks as well as the glue - eg grid resistors at the valve socket rather than on the PCB, like in Si's one.

It can be quite frustrating getting them to sound really good occasionally - some just don't sound quite right no matter what you seem to do. That said, when you do get one that's right, it really is right!