Marshall JCM 600 oscillating

[Gareth Magennis]

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination up
too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a (shorted)
jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you can
hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain dependent)

(Under certain test conditions it will oscillate massively ultrasonically.
It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix the
problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am going
round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere, but
since it's a complete loop broken by muting the input, it's kind of hard to
isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)



Cheers,


Gareth.

 

[N_Cook]

On 08/12/2015 18:55, Gareth Magennis wrote:

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination
up too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a
(shorted) jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you
can hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain dependent)

(Under certain test conditions it will oscillate massively
ultrasonically. It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix
the problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am
going round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere,
but since it's a complete loop broken by muting the input, it's kind of
hard to isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)



Cheers,


Gareth.

Missing or inadequate ground line?

 

[Gareth Magennis]

"N_Cook" wrote in message news:n47dkr$t7g$1@dont-email.me...

On 08/12/2015 18:55, Gareth Magennis wrote:

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination
up too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a
(shorted) jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you
can hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain
dependent)

(Under certain test conditions it will oscillate massively
ultrasonically. It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix
the problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am
going round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere,
but since it's a complete loop broken by muting the input, it's kind of
hard to isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)



Cheers,


Gareth.

Missing or inadequate ground line?



I have investigated that to a degree.

(Typical of these amps is to have the chassis ground connected to circuit
ground via 100 ohm resistor in parallel with a cap)

I've checked the grounding around the input socket and how the signal is
then transported to the board with the valves on it etc.
Everything appears to be OK. Nobody has fitted a non insulated jack socket
or anything like that.

But yes, I suspect something like the feedback is coming via the Ground, or
lack of, somewhere.



This amp HAS been messed with, however. The soldering on the footswitch
jack was appalling, resulting in disabling all channel change functions,
which I have now fixed.
There was other work that had failed that was equally poor quality.



Thanks,


Gareth.

 

[N_Cook]

On 08/12/2015 21:11, Gareth Magennis wrote:

"N_Cook" wrote in message news:n47dkr$t7g$1@dont-email.me...

On 08/12/2015 18:55, Gareth Magennis wrote:
This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination
up too high will break into oscillation (a few kHz) above a certain
gain.
This doesn't happen unless the pre-amp is unmuted by inserting a
(shorted) jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you
can hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain
dependent)

(Under certain test conditions it will oscillate massively
ultrasonically. It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix
the problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am
going round in circles and need some ideas my head doesn't have right
now.

There's some frequency dependent positive feedback going on somewhere,
but since it's a complete loop broken by muting the input, it's kind of
hard to isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)



Cheers,


Gareth.

Missing or inadequate ground line?



I have investigated that to a degree.

(Typical of these amps is to have the chassis ground connected to
circuit ground via 100 ohm resistor in parallel with a cap)

I've checked the grounding around the input socket and how the signal is
then transported to the board with the valves on it etc.
Everything appears to be OK. Nobody has fitted a non insulated jack
socket or anything like that.

But yes, I suspect something like the feedback is coming via the Ground,
or lack of, somewhere.



This amp HAS been messed with, however. The soldering on the
footswitch jack was appalling, resulting in disabling all channel change
functions, which I have now fixed.
There was other work that had failed that was equally poor quality.



Thanks,


Gareth.

This is the next route I'd take in.
Put a dummy load with scope on the amp and run via a variac.
100% mains until oscillating, then turn down to 60% or whatever , that
still maintains oscillation, then probe a 600V 1nF (if ultrasonic,higher
C if lower f) cap around to see where there is a change in oscillation
frequency.

 

[Phil Allison]

Gareth Magennis wrote:

Come on Phil, give us a clue, I'm stuck here.

** Most times I've seen such a problem, it was due to lack of shielding between the output valve plate wiring and input signal wiring. Maybe a shielded signal wire really isn't or the plate wires need to be twisted, tidied and pushed against the chassis.

Everything in the JCM600 is crammed together making unwanted coupling more likely.

Other times it has been due to high resistance grounds on pots or jacks - or because some fool decided to put the OT next to the inputs jacks.


..... Phil

 

[Gareth Magennis]

"Gareth Magennis" wrote in message news:aIF9y.963250$FM6.274839@fx42.am4...

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination up
too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a (shorted)
jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you can
hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain dependent)

(Under certain test conditions it will oscillate massively ultrasonically.
It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix the
problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am going
round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere, but
since it's a complete loop broken by muting the input, it's kind of hard to
isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)








Come on Phil, give us a clue, I'm stuck here.


Gareth.

 

[N_Cook]

On 11/12/2015 22:38, Gareth Magennis wrote:

"Gareth Magennis" wrote in message
news:aIF9y.963250$FM6.274839@fx42.am4...

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination up
too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a (shorted)
jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you can
hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain dependent)

(Under certain test conditions it will oscillate massively ultrasonically.
It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix the
problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am going
round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere, but
since it's a complete loop broken by muting the input, it's kind of hard to
isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)








Come on Phil, give us a clue, I'm stuck here.


Gareth.

I assume you've swapped out valves one by one, in case there is an
internal electrode problem , not picked up in a valve tester.
Another idea, moving a piece of mumetal around in likely areas?
Do aluminium valve shields actually screen from stray, to any great
extent, rather than just for valve retainer function?

 

[Gareth Magennis]

"Phil Allison" wrote in message
news:22272735-e7a9-4080-b4d2-8c88646a60cf@googlegroups.com...

Gareth Magennis wrote:

Come on Phil, give us a clue, I'm stuck here.

** Most times I've seen such a problem, it was due to lack of shielding
between the output valve plate wiring and input signal wiring. Maybe a
shielded signal wire really isn't or the plate wires need to be twisted,
tidied and pushed against the chassis.

Everything in the JCM600 is crammed together making unwanted coupling more
likely.

Other times it has been due to high resistance grounds on pots or jacks -
or because some fool decided to put the OT next to the inputs jacks.


..... Phil





Thanks, Phil, things are indeed particularly tight in this amp, and there
are loads of those pesky jumpers from PCB to PCB.
Makes the whole thing rather unenjoyable.

I'll have a look at the cable dressing etc, maybe something's not been put
back properly by the previous not very skilled repairer.


Cheers,


Gareth.

 

[Gareth Magennis]

"N_Cook" wrote in message news:n4h3tc$6i9$1@dont-email.me...

On 11/12/2015 22:38, Gareth Magennis wrote:

"Gareth Magennis" wrote in message
news:aIF9y.963250$FM6.274839@fx42.am4...

This is doing my head in.

Marhall JCM 600.
http://www.classictubeamps.com/schematics/Marshall/jcm600_60w.pdf


Turning the distortion channel's gain/volume/master volume combination up
too high will break into oscillation (a few kHz) above a certain gain.
This doesn't happen unless the pre-amp is unmuted by inserting a (shorted)
jack in the input socket.

As these 3 series gain controls approach the point of oscillation, you can
hear the inpending frequency that will feed back rise as the gain is
increased. (i.e. before feedback, the boosted frequency is gain
dependent)

(Under certain test conditions it will oscillate massively ultrasonically.
It's probably best not to do that very often)


Another amp repairer has been inside this amp and has attempted to fix the
problem by the looks of it. He's put small caps across some
electrolytics, and there was a resistor piggy backed over the top of R4,
feeding VR5. Not sure why.
I haven't found any other "mods", but that's not to say there aren't any.


Anyway, I've tried to isolate various things to discount them, but am
going
round in circles and need some ideas my head doesn't have right now.

There's some frequency dependent positive feedback going on somewhere, but
since it's a complete loop broken by muting the input, it's kind of hard
to
isolate anything really.
(Actually the input jack mutes both the input and the signal at CN6)


Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)








Come on Phil, give us a clue, I'm stuck here.


Gareth.

I assume you've swapped out valves one by one, in case there is an
internal electrode problem , not picked up in a valve tester.
Another idea, moving a piece of mumetal around in likely areas?
Do aluminium valve shields actually screen from stray, to any great
extent, rather than just for valve retainer function?




I changed all the valves for a "test set" of known good new ones.
No difference to the originals. (One output valve was faulty)

I checked all the pot values (and their soldering), in case someone had put
the wrong one in. They hadn't.


I guess if I can get it to the point where I can hear the obvious frequency
lobe building, I can start poking and moving cables around to see if
anything increases or decreases the lobe.
That will be monday.



Cheers,


Gareth.

 

[Phil Allison]

Black Iccy wrote:

I think I looked at the right schematic. I don't do repairs these days
but going back in time, it was not uncommon for spurious oscillations from
the output stages. Not denying that overall looping may be a problem but
*some* strange ultrasonics etc would get tamed by tiny (small value) chokes
in the plate connections and/or adding resistors into the grid line (so-called
grid stoppers). The circuit shows that there are some there. Maybe
replace/upgrade? Have you checked for RF band oscillations (in addition to
your audio group)?

** The OP's Marshall is unstable, oscillating at a few kHz, when gain settings are high. This strongly suggests that positive feedback is the culprit.

It is surprising how little capacitive coupling from the output valve plate wiring to input grid wiring can cause this to happen - with a voltage gain of over 200,000 between the two, even 1/100th of a pF will do the job.

The input sensitivity to full power of Marshall models fitted with a master volume is typically 60 microvolts at 6kHz, with all controls maxed. Just plugging an open circuit jack into the input results in oscillation.


.... Phil

 

[Black Iccy]

On Sat, 12 Dec 2015 14:11:13 -0000, "Gareth Magennis"
<sound.service@btconnect.com> wrote:

Any quick hints or tips from anyone? (It's not the Prescence feedback
circuit, or the valves)

I think I looked at the right schematic. I don't do repairs these days
but going back in time, it was not uncommon for spurious oscillations from
the output stages. Not denying that overall looping may be a problem but
*some* strange ultrasonics etc would get tamed by tiny (small value) chokes
in the plate connections and/or adding resistors into the grid line (so-called
grid stoppers). The circuit shows that there are some there. Maybe
replace/upgrade? Have you checked for RF band oscillations (in addition to
your audio group)?

 

[Gareth Magennis]

"Phil Allison" wrote in message
news:22272735-e7a9-4080-b4d2-8c88646a60cf@googlegroups.com...

Gareth Magennis wrote:

Come on Phil, give us a clue, I'm stuck here.

** Most times I've seen such a problem, it was due to lack of shielding
between the output valve plate wiring and input signal wiring. Maybe a
shielded signal wire really isn't or the plate wires need to be twisted,
tidied and pushed against the chassis.

Everything in the JCM600 is crammed together making unwanted coupling more
likely.

Other times it has been due to high resistance grounds on pots or jacks -
or because some fool decided to put the OT next to the inputs jacks.


..... Phil





Yep, the plate wiring wasn't tidy or hard on the chassis, and one of the PCB
interconnects was way too close. (Some cable ties had not been replaced)

All is good now, thanks.


Gareth.

 

[Phil Allison]

Gareth Magennis wrote:

"Phil Allison"

Come on Phil, give us a clue, I'm stuck here.


** Most times I've seen such a problem, it was due to lack of shielding
between the output valve plate wiring and input signal wiring. Maybe a
shielded signal wire really isn't or the plate wires need to be twisted,
tidied and pushed against the chassis.

Everything in the JCM600 is crammed together making unwanted coupling more
likely.

Other times it has been due to high resistance grounds on pots or jacks -
or because some fool decided to put the OT next to the inputs jacks.



Yep, the plate wiring wasn't tidy or hard on the chassis, and one of the PCB
interconnects was way too close. (Some cable ties had not been replaced)

** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary wires run next to all the pre-amp valve wiring on their way to the valve sockets on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for oscillation.

Be worthwhile twisting the three primary wires and moving them further back.


..... Phil

 

[Gareth Magennis]

"Phil Allison" wrote in message
news:8a7973ba-0515-40cb-888d-2450e7ada015@googlegroups.com...

Gareth Magennis wrote:

"Phil Allison"

Come on Phil, give us a clue, I'm stuck here.


** Most times I've seen such a problem, it was due to lack of shielding
between the output valve plate wiring and input signal wiring. Maybe a
shielded signal wire really isn't or the plate wires need to be twisted,
tidied and pushed against the chassis.

Everything in the JCM600 is crammed together making unwanted coupling more
likely.

Other times it has been due to high resistance grounds on pots or jacks -
or because some fool decided to put the OT next to the inputs jacks.



Yep, the plate wiring wasn't tidy or hard on the chassis, and one of the
PCB
interconnects was way too close. (Some cable ties had not been replaced)

** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary wires
run next to all the pre-amp valve wiring on their way to the valve sockets
on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for
oscillation.

Be worthwhile twisting the three primary wires and moving them further back.


..... Phil




The multitude of grey interconnects are not actually screened cable, just a
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.



Cheers,


Gareth.

 

[Phil Allison]

Gareth Magennis wrote:

** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary wires
run next to all the pre-amp valve wiring on their way to the valve sockets
on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for
oscillation.

Be worthwhile twisting the three primary wires and moving them further back.





The multitude of grey interconnects are not actually screened cable, just a
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.

** Close to none.

The layout relies on the two plate wires being in close proximity so the out of phase electric fields cancel at a distance.



... Phil

 

[Gareth Magennis]

"Phil Allison" wrote in message
news:6046f3c8-69b7-49ed-bb61-8d0ee0a337e2@googlegroups.com...

Gareth Magennis wrote:

** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary
wires
run next to all the pre-amp valve wiring on their way to the valve sockets
on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for
oscillation.

Be worthwhile twisting the three primary wires and moving them further
back.





The multitude of grey interconnects are not actually screened cable, just
a
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.

** Close to none.

The layout relies on the two plate wires being in close proximity so the out
of phase electric fields cancel at a distance.



.... Phil




Thanks.




By the way, and off topic now, I had a young lad come to me today to ask
about how he might learn how to digitally model valve amps for his
University project.
He was sent by a customer of mine who I've repaired valve amps for, and is a
teacher at his college.

I recommended this book, as this is the one I first read to learn the
basics, and I think it is quite good for that. Certainly taught me a lot.
http://www.amazon.co.uk/Valve-Amplifiers-Morgan-Jones/dp/0750656948

Do you or anyone else have any knowledge/opinion of this publication, or
anything else I could recommend him?


Gareth.

 

[John Heath]

On Tuesday, December 15, 2015 at 7:00:52 PM UTC-5, Phil Allison wrote:

Gareth Magennis wrote:



** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary wires
run next to all the pre-amp valve wiring on their way to the valve sockets
on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for
oscillation.

Be worthwhile twisting the three primary wires and moving them further back.





The multitude of grey interconnects are not actually screened cable, just a
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.


** Close to none.

The layout relies on the two plate wires being in close proximity so the out of phase electric fields cancel at a distance.



... Phil

There is a third option. Stop being a hero. You can not fix everything. Phone the customer and say this is a tricky problem and that they are better off taking it to the manufacturer for service. You will lose money but gain some cookie points for being honest. Better still the manufacturer is now spinning their wheels trying to fix it not you which the way it should be if it is a design problem.

 

[Chuck]

On Wed, 23 Dec 2015 04:52:13 -0800 (PST), John Heath
<heathjohn2@gmail.com> wrote:

On Tuesday, December 15, 2015 at 7:00:52 PM UTC-5, Phil Allison wrote:
Gareth Magennis wrote:



** Found this pic of the insides of one:

http://gitaradiy.pl/uploads/2_974_JCM600_6.jpg

The OT is on the far left ( right behind the input jacks ) and primary wires
run next to all the pre-amp valve wiring on their way to the valve sockets
on the right.

That is a *really* poor layout for a valve guitar amp and sitting duck for
oscillation.

Be worthwhile twisting the three primary wires and moving them further back.





The multitude of grey interconnects are not actually screened cable, just a
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.


** Close to none.

The layout relies on the two plate wires being in close proximity so the out of phase electric fields cancel at a distance.



... Phil

There is a third option. Stop being a hero. You can not fix everything. Phone the customer and say this is a tricky problem and that they are better off taking it to the manufacturer for service. You will lose money but gain some cookie points for being honest. Better still the manufacturer is now spinning their wheels trying to fix it not you which the way it should be if it is a design problem.

Many times in the past it is the warranty stations that figure out
what the design fault is and the modification that needs to be done
and notifies the manufactuer.

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Phil Allison]

John Heath wrote:

The layout relies on the two plate wires being in close
proximity so the out of phase electric fields cancel at a distance.




There is a third option. Stop being a hero.
You can not fix everything. Phone the customer
and say this is a tricky problem and that they
are better off taking it to the manufacturer for
service.

** The OP fixed the oscillation problem, after acting on my advice about cable dress. A previous repairer had been careless.


..... Phil

 

[Phil Allison]

Chuck wrote:

Many times in the past it is the warranty stations that figure out
what the design fault is and the modification that needs to be done
and notifies the manufactuer.

** Lucky if they take any notice of repairers at all - the products are by then all sold and a recall is out of the question.

If the product is made in another country, you got no chance - the reply will likely be: "We do not have that issue here... "

Designers are typically arrogant people and manufacturers very reluctant to admit when there is a serious problem that requires them to change what they are doing.


..... Phil