Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Marshall JCM 600 oscillating
187 views
Skip to first unread message
Gareth Magennis
Dec 8, 2015, 1:55:08 PM12/8/15
to
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.
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
Dec 8, 2015, 3:15:47 PM12/8/15
to
Gareth Magennis
Dec 8, 2015, 4:11:49 PM12/8/15
to
"N_Cook" wrote in message news:n47dkr$t7g$1...@dont-email.me...
(Typical of these amps is to have the chassis ground connected to circuit
ground via 100 ohm resistor in parallel with a cap)
I've checked the grounding around the input socket and how the signal is
then transported to the board with the valves on it etc.
Everything appears to be OK. Nobody has fitted a non insulated jack socket
or anything like that.
But yes, I suspect something like the feedback is coming via the Ground, or
lack of, somewhere.
This amp HAS been messed with, however. The soldering on the footswitch
jack was appalling, resulting in disabling all channel change functions,
which I have now fixed.
There was other work that had failed that was equally poor quality.
Thanks,
Gareth.
N_Cook
Dec 9, 2015, 7:45:12 AM12/9/15
to
Put a dummy load with scope on the amp and run via a variac.
100% mains until oscillating, then turn down to 60% or whatever , that
still maintains oscillation, then probe a 600V 1nF (if ultrasonic,higher
C if lower f) cap around to see where there is a change in oscillation
frequency.
Gareth Magennis
Dec 11, 2015, 5:38:12 PM12/11/15
to
"Gareth Magennis" wrote in message news:aIF9y.963250$FM6.2...@fx42.am4...
Gareth.
Phil Allison
Dec 11, 2015, 8:22:07 PM12/11/15
to
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.
>
> Come on Phil, give us a clue, I'm stuck here.
>
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
N_Cook
Dec 12, 2015, 7:31:04 AM12/12/15
to
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
Dec 12, 2015, 7:35:19 AM12/12/15
to
"Phil Allison" wrote in message
news:22272735-e7a9-4080...@googlegroups.com...
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
Dec 12, 2015, 9:11:19 AM12/12/15
to
"N_Cook" wrote in message news:n4h3tc$6i9$1...@dont-email.me...
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.
Black Iccy
Dec 12, 2015, 8:11:31 PM12/12/15
to
On Sat, 12 Dec 2015 14:11:13 -0000, "Gareth Magennis"
<sound....@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
<sound....@btconnect.com> wrote:
>> Any quick hints or tips from anyone? (It's not the Prescence feedback
>> circuit, or the valves)
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)?
Phil Allison
Dec 12, 2015, 8:51:37 PM12/12/15
to
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.
>
>
>
> 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)?
>
>
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
Gareth Magennis
Dec 14, 2015, 2:43:20 PM12/14/15
to
"Phil Allison" wrote in message
news:22272735-e7a9-4080...@googlegroups.com...
interconnects was way too close. (Some cable ties had not been replaced)
All is good now, thanks.
Gareth.
Phil Allison
Dec 14, 2015, 8:30:25 PM12/14/15
to
Gareth Magennis wrote:
> "Phil Allison"
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
> "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.
>
>
>
> > 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:
> interconnects was way too close. (Some cable ties had not been replaced)
>
>
>
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
Dec 15, 2015, 4:33:03 AM12/15/15
to
"Phil Allison" wrote in message
signal wire lying beside an earth wire.
I don't know what level of protection that actually offers.
Cheers,
Gareth.
Phil Allison
Dec 15, 2015, 7:00:52 PM12/15/15
to
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.
>
>
>
>
>
>
>
> ** 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.
> signal wire lying beside an earth wire.
> I don't know what level of protection that actually offers.
>
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
Dec 15, 2015, 7:42:01 PM12/15/15
to
"Phil Allison" wrote in message
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
Dec 23, 2015, 7:52:18 AM12/23/15
to
Chuck
Dec 23, 2015, 2:04:54 PM12/23/15
to
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
Dec 23, 2015, 8:14:17 PM12/23/15
to
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.
> >
> >
>
>
>
> >
> > 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.
> 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.
>
.... Phil
Phil Allison
Dec 23, 2015, 8:37:16 PM12/23/15
to
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.
>
>
>
> 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.
>
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
N_Cook
Dec 24, 2015, 7:24:25 AM12/24/15
to
to a main agent for repair sometime in the next 3 months(even if
manufacturer pays the return cost), then finds out a good chance of said
item being repaired locally for much the same cost , in half a week,
they always tend to one option, as they have a gig next week.
John Heath
Dec 25, 2015, 11:43:24 AM12/25/15
to
Related to wire dressing I have noticed in some preamps the power transformer is mounted in an odd orientation. Often at 45 degrees to the chassis frame. I imagine an engineer when prototyping move the power transformer here and there finding the sweet null spot where stray 60 CPS would not couple into the inputs. I have a bench frequency counter that is the same with the power transformer mounted at 45 degrees relative to the chassis frame.
Phil Allison
Dec 25, 2015, 10:06:39 PM12/25/15
to
John Heath wrote:
AC supply transformer orientation can be critical in sensitive devices, graphic equalisers and tape machines are good examples - plus CRT based scopes.
In tube amps, AC and output transformers are normally mounted close together with the cores at right angles to prevent direct injection of 50/60 Hz hum.
In case of the OP's JCM600, Marshall elected to mount the two at opposite ends of the chassis, maybe to balance the weight, but doing so can easily lead to instability.
.... Phil
>
> Phil Allison 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 Allison 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.
> >
> >
>
>
> I can see how wire dressing would cause this with high gain of a guitar
> amp. However this is Monday night quarterbacking. One does not know it
> is wire dressing ,
>
>
** Moving the input signal and output plate wires BACK to where they were originally and having all oscillation stop is enough proof for anyone familiar with tube guitar amps - but apparently not you.
> amp. However this is Monday night quarterbacking. One does not know it
> is wire dressing ,
>
>
AC supply transformer orientation can be critical in sensitive devices, graphic equalisers and tape machines are good examples - plus CRT based scopes.
In tube amps, AC and output transformers are normally mounted close together with the cores at right angles to prevent direct injection of 50/60 Hz hum.
In case of the OP's JCM600, Marshall elected to mount the two at opposite ends of the chassis, maybe to balance the weight, but doing so can easily lead to instability.
.... Phil
0 new messages