Marshall JCM900 rebuild

[CJ]

got a JCM900 100w hi gain master vol mk III

complaint was muddy sound, found a .047 across the input jack to ground,

cleared that up, found that there might be room for improvement on this particular model,

 

[CJ]

here is the original circuit, bridge diode used for clipping, then feed it all into a TLO72, ugly clipping when over driven,

master vol 1 and 2 are just two pots coming off the same node,  not real channel switching going on, and one of those A/B switching opamps, not very musical clipping from that either,

 

[CJ]

circuit board was getting knackered as we tried to make this sparrow hawk into an eagle,

moved the tone stack forward where it sounds better, after stage one,

some improvement, but after a week, it was deemed futile to try and get this thing to sound like a real Marshall,

so bite the bullet and scrap the board, commit to a P to P rebuild where we can mod to our hearts content with no colateral  (sp) damage,

 

[CJ]

only have to wire first two tubes, phase inverter is on the power board, getting there, they had 0.1/200 volt caps on the plates of the inverter, do not know how they survived, put some juicy .068/600 film caps (blue) in there,

got rid of the busy effects loop (see previous schematic) nobody uses those anyway, (we hope  )

RMC green top 330 pf,  a Sangamo oil filled 0.1 for the tone sack, gonna put a Eurofoil .047 from home in there tomorrow,

got a real input jack also, gonna use the opamp supply to switch in a relay for true dual channel operation, borrowing ideas from a Kustom 72 Coupe (James Brown)

poor guy bought a MojoTone OPT thinking it would fix this mess,

 

[CJ]

here is what we are going after, preamp of the 72 Coupe, built his circuit before and the results were spectacular, full props going out to JB Ampmaster,  want your Epiphone SG to sound like a 59 Les Paul? this is your amp. 

don't have to put in every tone tweak, gets pretty busy around the vol controls,

 

[CJ]

here is a non destructive test we did on the Mojotone out of curiousuty, (sp)  looks like they tried to duplicate the low inductance/low DCR version that comes stock from Dagnall,

 

[Tubetec]

I think tube amps over the years have had all the guts taken out of them ,once upon a time you had your bassman or bluesbreaker ,gutsy lowdown grunt ,not shrill or screachy like the modern boxes ,nipping out the odd vol bypass cap helps a bit,fx loops with sillicon chips ,gone.
you gotta be carefull not to encroach on your bass players frequencies with such old style amps though ,100w marshall  4x12 ,fat channel with neck pickup ,way way to much mud even on a strat,les paul and your getting down in siesmic low end ,something I always liked about the Orange amps ,FAC (frequency attenuation capacitors) ,you can pull out the wallowy low end nicely before the output stage .Savage treble boost just aint the same as bass cut though
Say Im eq'ing a 58 on vox ,first job cut the bass ,99% of the time you get the bass weight right ,theres very little else needs anything doing ,start from the treble end ,you just end up making a mess .

I once had a home brew 4xel84 style ac 30 dropped into me ,anything above nominal volume and the thing howled with feedback ,like a cat being strangled down a back alley . I set it right so it didnt osscillate,,never had a chance to get the owners opinion on it though ,its just so hard to know how a customer will take a mod ,my guess he was out of sorts till he found his levels again . Took some cobwebs out of a farily well known players amp one time ,the security heavies eventually payed up the hard cash ,after trying to corner me in the back of the venue for their inepptitude and not having a can of iso spray handy .Jmp's have just about the right balance for me ,ones without the master volume always took effects pedals better too .
anyway Cheers Cj ,always an inspiration and an education .

 

[CJ]

good stuff there Tubetec ,  these circuit board pots nowadays, some of them make dialing in tone a hassle, so without the pc board, we can throw some CTS  or Alpha large body pots to get a better taper, notice that JCM schemo uses linear taper for the treble, linear pot for the first vol VR1,  ...

 

[Tubetec]

Will you eventually build the eq's off the pot lugs on that one Cj?
Ive done that on one or two builds and it seems to work ok for me .
From what your saying the dagnall transformer is set at less than optimum winding ratio ,how does this compare to Jmp series transformers .
Thanks again ,

 

[CJ]

this thing is just about ready to fire up,

got relay switching instead of the A/B opamp,

dang thing oscillates, wtf?

check circuit, everything is ok,

start moving wires and the problem vanishes,  lead dress, just a matter of moving a wire 1 inch and we have a stable amp,    ;D

so when you see a weird wire layout on that old blackface Fender, realize there is a reason for that!

bandwidth goes up when you switch from pc board to point to point, so sometimes you have to do stuff,

this is just 1 k hz, can you imagine the poor Ham Radio guys with their 10 meter DIY transmitter, sheesh, 

here is a before and after pic showing the plate wire of the 4th stage being moved away from the grid wire of stage 2,  and the results therein>

 

[JohnRoberts]

If you think about it instability is a gross manifestation of crosstalk between high impedance circuit nodes. Well below becoming a RF transmitter, crosstalk will alter the sonic signature of a high impedance (and high gain) audio path.

I am not a tube guy, but used to chew the fat with guys who designed tube guitar amps for a living and there are lots of subtle interaction involved in realizing a desirable sound signature.

Of course not oscillating is always desirable.  8)

JR

 

[abbey road d enfer]

The quick and dirty fix for oscillations is putting caps on sensitive lines; that's what CBS did to Fender amps, and it killed the tone.

It can also be a dubious artistic choice. The Fender Blues Jr. is an example of how trying to make the amp sound like a blues amp (whatever it means  ??? ) kills its tone.
One of my friends bought one such amp after hearing mine; he came back to me and said it was lacking, seemed dead-ish.
We took the amps to the AP and compared them. There was a slight difference, about 1dB at 5kHz, that I could not explain, until I remembered I had, years before,  lifted one end of a treble bleed capacitor. When I did the same on his amp, the sound became much more open.
One should always keep a log of mods!

 

[peterc]

Would using shielded cable stop the need for such critical lead dressing?

It seems a little odd that in this day and age it would not be used on a wider scale.

Peter

 

[abbey road d enfer]

Would using shielded cable stop the need for such critical lead dressing?

It seems a little odd that in this day and age it would not be used on a wider scale.

Peter

Shielded cable adds capacitance that may significantly alter the response if placed in a critical node.

 

[CJ]

the Shockmeister 5000 is done, sounds like the Springfield  Creamery in Eugene, Oregon.

on my way to the cardio unit for valve replacement, not the amp, me.  ;D

 

[CJ]

pic for ref in case we get another one in for dismemberment,

 

[CJ]

took out the red arrow, was gobbling top end,

may be a typo on the schemo, put in a 47 pf, still had some gremlins when treble and presence on 10 along with master and lead drive, standing near amp with hands off strat, you know that marshall squeal, got rid of that with some coax to overdrive tube stage grid,

 

[alexc]

Now I know what to do with an old pcb based jcm900

hOPE the ticker keeps trucking on strong ...

happy new year  8) 8)

 

[abbey road d enfer]

took out the red arrow, was gobbling top end,

may be a typo on the schemo, put in a 47 pf, still had some gremlins when treble and presence on 10 along with master and lead drive, standing near amp with hands off strat, you know that marshall squeal, got rid of that with some coax to overdrive tube stage grid,

I think 470pF is a steamhammer answer to oscillation; indeed it's a very sensitive noise. Proof that a judiciously chosen short piece of shielded cable is a better fix in that case.

 

[JohnRoberts]

Would using shielded cable stop the need for such critical lead dressing?

It seems a little odd that in this day and age it would not be used on a wider scale.

Peter

As Abbey already shared the C of typical shielded cable could introduce an RC LPF if source impedance is high enough...and all the sensitive nodes inside tube amps are high impedance..  BUT the shield does not necessarily have to be grounded. The shield could be tied to some other low impedance node in the circuit. This would likely introduce very specific crosstalk, that could be OK, or not.

The shift to PCB layout for tube amps instead of p-to-p wiring was a mixed blessing. It made lead dress much more predictable  so the sonic signature was stable and repeatable, for better and worse, often worse. No longer can the design engineer twiddle with how the wires are dressed out for best sound signature.

As I already mentioned I used to watch  tube amp designers at work, and wrestling with PCB layout was a major headache. Especially for budget amps that were the first targets for PCB use to reduce cost. The design engineer was not afforded the luxury to try multiple experimental PCB cuts to get the sound right (I know I used to manage an engineering group and time to market is money, not to mention repeating work costs actual money too). 

In an ideal world PCB based amps will keep getting better as designers learn from previous versions.  P-to-P gear has the potential to sound good but since humans are involved that is not guaranteed.

JR