Marshall JTM 60 problem with gain channel with FX loop

[torcamaniac]

Hi guys,
I'm new to this forum and I hope some of you could help me with an issue on my modded JTM 60.
First I've done all the mods posted by Gunnar on Metroamp forum, such as Diodes clipping out, changing some resistors and capacitors. Anyway the amp sounds great on both clean and gain channel.
But the actual problem I'm having is with serial or parallel fx loop both, when using the gain channel.
It doesn't matter what stompbox I use in the loop (delay, booster etc.), the gain channel gets much quieter!!!
And I really love JTMs gain channel and would like to be able to boost it and delay it through the loop for leads.
On the clean channel all the pedals in loop work great and it doesn't get quieter like the gain channel.
I also replaced the cheap send/return Marshall jacks with new ones from Neutrik. Doesn't seem to be impendance issue 'cause I also tried swapping the original 470 ohm and 100K resistors with 47 ohm and 10K and anyway if that was the problem I guess it wouldn't work for the clean channel either.
I suspect that the problem is somewhere in the preamp stages maybe???

I hope somebody has some ideas and could help???

[Jazz P Bass]

Schematic?

[pdf64]

I'm sure many are willing to review what you've got and hypothesise what the issue may be.
However, the post lacks the basic info of a link to the stock schematic and list of changes.
Chasing that down may take a while, and then the stock schematic has to be analysed, amendments transferred to it and their affect assessed.
It would be helpful if you could provide an amended schematic, eg print it out, write on the changes, scan and upload; or use MS Paint to amend the schematic and upload.

[torcamaniac]

Thanks for answering.
I'll post the original schematic (it's the same as the ones from JCM600. The original for JTM 60 have much worse resolution and it's harder to read) and will try to post the one with changes I made.

[torcamaniac]

Here are the shematics. Modded stuff is in red (note that I didn't change R119 and R132).
jcm6-61-02.pdf

And here is the shematic of the FX loop which doesn't have any mods:

[torcamaniac]

Also on the modded image I erased the diodes and C105. In reality I just removed C105 to get the clipping diodes out of the circuit.

[J M Fahey]

Sorry but it's still unclear.
To begin with, the Marshall so called "schematic" is horrible, it only shows what's inside the PCB but not pots, jacks, switches, relays or interconnection with other PCBs or chassis/panel mounted stuff.
In fact it's not a schematic but the .SCH file used to draw the PCB.

And we can't study what your changes/mods could have done, because we don't know the original values.

Please repost the original schematic, all original values and connections visible, and cross out or add whatever's needed to show the mod; red colour is fine.
And leave the diodes, etc. visible, if crossed out with a red line.

That said, and just as an advance, and purely from memory:

1) I assume at least part of the clipping was diode generated (standard diode/red Led/Zener?) which means audio is clipped to a peak value between 0,7V and a couple volts, so the Loop design is based on that, and loop expects a certain value, compatible with what floor pedals can handle (notice that I'm assuming a lot here, based on memory, let's see the real Mods side by side).

Typically when you pull diodes off, audio voltages rise a lot , Op Amps will clip but now with almost 15V peaks and tubes may show up to 150V peaks ... neither of these values can be handled by any regular floor pedal so they will clip anything above , say, 4V peak.
So if you have, say, a 1 or 2V signal *tops* going through the loop and you insert a pedal, nothing changes (hint: clean channel signal) but if modded dirty channel signal is now, say, 10/20/40V it will go trough the loop unmolested BUT as soon as you insert a pedal , it will be brutally clamped down to 2 or 3 Volts ... instant "volume loss".
Anyway, will wait for your "dual" schematic to confirm or deny this.

2) Re the V102a stage, bypassing R123 with a cathode cap again rises audio levels way beyond what original designers expected , including Loop levels.

So in a way we could say that the cause of your problems are the Mods themselves.
But again, let's see the Mods; mods posted by Gunnar on Metroamp forum is not enough.

[torcamaniac]

Sorry but it's still unclear.
To begin with, the Marshall so called "schematic" is horrible, it only shows what's inside the PCB but not pots, jacks, switches, relays or interconnection with other PCBs or chassis/panel mounted stuff.
In fact it's not a schematic but the .SCH file used to draw the PCB.

And we can't study what your changes/mods could have done, because we don't know the original values.

Please repost the original schematic, all original values and connections visible, and cross out or add whatever's needed to show the mod; red colour is fine.
And leave the diodes, etc. visible, if crossed out with a red line.

That said, and just as an advance, and purely from memory:

1) I assume at least part of the clipping was diode generated (standard diode/red Led/Zener?) which means audio is clipped to a peak value between 0,7V and a couple volts, so the Loop design is based on that, and loop expects a certain value, compatible with what floor pedals can handle (notice that I'm assuming a lot here, based on memory, let's see the real Mods side by side).

Typically when you pull diodes off, audio voltages rise a lot , Op Amps will clip but now with almost 15V peaks and tubes may show up to 150V peaks ... neither of these values can be handled by any regular floor pedal so they will clip anything above , say, 4V peak.
So if you have, say, a 1 or 2V signal *tops* going through the loop and you insert a pedal, nothing changes (hint: clean channel signal) but if modded dirty channel signal is now, say, 10/20/40V it will go trough the loop unmolested BUT as soon as you insert a pedal , it will be brutally clamped down to 2 or 3 Volts ... instant "volume loss".
Anyway, will wait for your "dual" schematic to confirm or deny this.

2) Re the V102a stage, bypassing R123 with a cathode cap again rises audio levels way beyond what original designers expected , including Loop levels.

So in a way we could say that the cause of your problems are the Mods themselves.
But again, let's see the Mods; mods posted by Gunnar on Metroamp forum is not enough.

Thanks.

Yes, when removing the diodes the signal will get too loud and would saturate the next stage. This is why the VR6 pot would work very sensitive and saturate extremely and why I've put two 100K voltage dividers before the pot as before CN5 going into CN108 where the PCB with potis is located and right after V102B.
Like I wrote before, the JTM 60 original shematics are a mess. That's why I've attached the JCM600 shematics (they are basically the same amp with same circuit) in PDF because it's much easier to read. The shematics are in 3 pieces as 3 PCB's inside these amps (Power diagram, Preamp diagram and mains/output diagram).
I am posting these 3 pdfs once more and attach a 4th file with the exact mods I did.

jcm6-61-02-1 (1).pdfjcm6-61-02-1 (1).pdfjcm6-62-02-1.pdf


Anyway just tried the amp with a booster in fx loop again. If I put the gain knob down to around 2 from 20 the amp does get a big boost same as on the clean channel. While going up it's less noticable and with more than 3,5 it's starts lowering the dry signal volume. Weird???
Also, the master volume was low, around 3, gain channel volume middle flat.
And than I tried the clean channel with same master setting but channel volume maxed and I didn't get any boost. Mind you, after the mods the clean channel breaks up earlier and when maxed it sounds like a really nice old-school crunchy overdrive.
So I guess it all has something to do with gain saturation and headroom???

[torcamaniac]

Here some pics of the inside. You can also see the two blue mini fans I've built in to deal with the infamous overheating issue these JTM's are having. They are running really slow and quiet. There's also a third bigger fan on the outside.

[J M Fahey]

Just stopped 5 minutes for a cup of coffee .... these d*mn Musicians who INSIST on picking up their stuff on Friday afternoon or Saturday .... what's wrong with waiting until Monday 10 AM?
But on a quick overview about:

Anyway just tried the amp with a booster in fx loop again. If I put the gain knob down to around 2 from 20 the amp does get a big boost same as on the clean channel. While going up it's less noticable and with more than 3,5 it's starts lowering the dry signal volume. Weird???

think of it this way:
* suppose (just for analysis) that the booster has 5X gain, *and* its output clips at 3V RMS (maximum possible from a 9V supply and in practice closer to 2V)
Also suppose it's true bypass.
a) you hit it with 100mV (volume @2 on a 0-20 scale) ; when activated signal rises 5X to 500mV , that's a BIG volume boost

b) now you raise volume or switch to a higher gain channel, but set for relatively low gain/volume:
you hit it with 1V ; when activated signal tries to rise to 5V but is clamped to dirty 3V ... there is a volume boost but less than expected (the added dirt does not bother you and even goes unnoticed because you are overdriving other amp stages)

c) now you hit it with the fury of an unclamped (you pulled the diodes) tube stage, say it gets 10V ... output can not rise above 3V RMS no matter what, and you LOSE volume.

Ouch, got carried away, coffee is lukewarm ... oh well, back to the salt mine.

[torcamaniac]

So you think the only way to be able to get the fx loop function with the gain channel would be if I put the diodes back?
Hope there's another way 'cause the gain channel sounds much better without those

[J M Fahey]

No, not necessarily, and as Enzo repeats now and then, "there's more than one way to skin a cat".

Just thinking aloud and in no particular order:
* if you like the tube distortion, fine

* at the point the Loop sits today, signal level increased *a lot* , so it's advisable not to use it.

* I suppose there is at least one extra stage after the Loop related one which provides the final distortion , should check the schematic.

* signal there will be very high, way more than what's needed to drive the power amp.

* so it's probably attenuated a lot, plus a master level volume must be fitted to control dirty level.

* if signal level there is, say, 2 to 5 V RMS (whatever's needed to drive the PI) , it's reasonable to attenuate that to some 200mV (which any 9V pedal can handle easily), maybe 500mV but not higher, "Send" to loop Jack; "Return" from the other and then add a simple, say, 4X to 10X stage (may be a single transistor or FET or, of course, an Op Amp) to rise signal back to former unattenuated level.

In fact I remember seeing a small PCB "add on loop" which does that, and if I'm not wrong, needs +/-15V ... which you already have.

Do not suggest modding the current loop for 2 reasons:

1) it's *very* simple, just a single transistor buffer, no extra gain possible, and return is fully passive

2) in any case, *now* it's in the wrong place.

I'd start at the PI input grid and work backwards, or plain insert the active loop between said grid and whatever is driving it.

You might also copy the Valvestate 80/100 loop which again does something similar.

No need for a real PCB (which implies design, burning, etching and drilling, sometimes a deal killer for a total production count of just 1) , it's within what is still "comfortable" on perfboard or veroboard.

[torcamaniac]

No, not necessarily, and as Enzo repeats now and then, "there's more than one way to skin a cat".

Just thinking aloud and in no particular order:
* if you like the tube distortion, fine

* at the point the Loop sits today, signal level increased *a lot* , so it's advisable not to use it.

* I suppose there is at least one extra stage after the Loop related one which provides the final distortion , should check the schematic.

* signal there will be very high, way more than what's needed to drive the power amp.

* so it's probably attenuated a lot, plus a master level volume must be fitted to control dirty level.

* if signal level there is, say, 2 to 5 V RMS (whatever's needed to drive the PI) , it's reasonable to attenuate that to some 200mV (which any 9V pedal can handle easily), maybe 500mV but not higher, "Send" to loop Jack; "Return" from the other and then add a simple, say, 4X to 10X stage (may be a single transistor or FET or, of course, an Op Amp) to rise signal back to former unattenuated level.

In fact I remember seeing a small PCB "add on loop" which does that, and if I'm not wrong, needs +/-15V ... which you already have.

Do not suggest modding the current loop for 2 reasons:

1) it's *very* simple, just a single transistor buffer, no extra gain possible, and return is fully passive

2) in any case, *now* it's in the wrong place.

I'd start at the PI input grid and work backwards, or plain insert the active loop between said grid and whatever is driving it.

You might also copy the Valvestate 80/100 loop which again does something similar.

No need for a real PCB (which implies design, burning, etching and drilling, sometimes a deal killer for a total production count of just 1) , it's within what is still "comfortable" on perfboard or veroboard.

Thank you Juan,
that makes sense. Yes, I love the gain without the diodes. Before it there was too much fizz.
As far as I know V1 and V2 are the first preamp tubes, V3 is the tone stack and V4 is PI.
What I noticed is also that V1 and V3 glow bright and V2 looks a bit dimmed in comparison and PI V4 glows even stronger than V1 and V3.
I could make an active loop board, been searching and reading about loops a lot lately, but can't think of a way where to place this. There's just not enough space in the chasis.
I guess something like Morley's effect loop corrector woudn't help here either?
Also, you've said before that the mod where I bypassed the R123 with a Cap raised the signal a lot. Think it would help a bit if I removed the Cap again?

Thanks Juan

[pdf64]

Have any others that have tried these mods commented on your issue, eg on the metroamp forum?
As Juan notes, the Marshall drawings aren't a full circuit schematic, ie it's not apparent where the connectors go to, how the fx loops are arranged etc, so it seems next to impossible to work out what's happening from that info alone.

[torcamaniac]

Have any others that have tried these mods commented on your issue, eg on the metroamp forum?
As Juan notes, the Marshall drawings aren't a full circuit schematic, ie it's not apparent where the connectors go to, how the fx loops are arranged etc, so it seems next to impossible to work out what's happening from that info alone.

Unfortunatrly,
Metroamp forum seems defunct. I posted there in the original thread where Gunnar posted the mods(also wrote directly to his inbox) and also opened a new thread about the FX Loop. But it seems there's no activity there.
You can see where the leads are from PCB to PCB. There are 3. They are connected through CNxxx two wire plugs to each other.
For example CN5 on tubes PCB(Power circuit diagram) after V102B leads to CN108 to Volume pot VR6 on the front PCB (preamp circuit diagram).