stabilizing an amp

[wormdirt]

Hey guys,
Mostly my first post, after the ampgarage went south I'm in need of a new community.

Anyway I have a friends Nomad HiGain on my bench, and it is presenting oscillation/sputtering with high master volume and gain settings.
No schematic exists for this amp, as far as I know.

Chopstick tests led me to a 1 bad ground connection, which helped, but did not totally cure the problem.
Tubes have all been changed/checked with no help there.
I checked the majority of resistors, and they all seem in spec, from what I can tell.

I rewired up the entire input jack section.

A few things I noticed while I was in there was that there wire from the grid stoppers to V1 isn't grounded on either side, though it is a shielded wire.
The other thing is that the pots do not have a ground bus, which isn't necessarily odd, but maybe something to consider adding.

Anywhere else I should check?

All the testing equipment I have is a solid DMM and a chopstick.

[Leo_Gnardo]

after the ampgarage went south I'm in need of a new community

Went south like hell! TAG aka The Amp Garage still very much alive and well everybody. Reports of its demise are premature. There is a bit of a problem with being identified as a "potential malware site" which is pure BS. You have to be brave enough to pass the warning page, it's like a self-appointed bouncer that nobody hired or asked to be there. Sooner or later this will be ironed out. Just be glad it hasn't happened here. Yet. Now on to your regularly scheduled activities.

[J M Fahey]

Nomad HiGain on my bench, and it is presenting oscillation/sputtering with high master volume and gain settings.
Chopstick tests led me to a 1 bad ground connection, which helped, but did not totally cure the problem.
Tubes have all been changed/checked with no help there.
I checked the majority of resistors, and they all seem in spec, from what I can tell.
I rewired up the entire input jack section.
A few things I noticed while I was in there was that the wire from the grid stoppers to V1 isn't grounded on either side, though it is a shielded wire.
The other thing is that the pots do not have a ground bus, which isn't necessarily odd, but maybe something to consider adding.
All the testing equipment I have is a solid DMM and a chopstick.

1) very high gain amps are tricky to stabilize.
Although it shouldn't be so, sometimes a certain controls setting is unstable/oscillates and that does not mean it's "broken" , at least since it did so straight out of the box and all others do the same
A couple Marshall heads come to mind.
The secret behind some Trainwreck seems to be the peaks and dips introduced by being so close to oscillation that they become very audible

2) Mesa in particular are known for running at the very edge of the cliff, layout is paramount.
I'm quite certain that you do not have "bad parts" per se, definitely resistor and eventual cap measurement and replacement will not help you.

3) Mesa uses some "dirty tricks" for last minute compensation, personally I hate that because it can not be shown or deduced from the schematic , although probably some service notes, read only by Mesa Authorized Techs, mention them and explain how to compensate.

One such trick is to solder at some specific points little solid wire "antennas" , no kidding, we are talking 1 inch or so length wires with only 1 end soldered to a grid or plate, where the free end is pointed "this way" or "that way" , probably with the tip of a chopstick, until oscillation disappears.

The "mystery" shielded wire is also another such trick.
It's not directly grounded to chassis or circuit ground, which is not a strict requirement anyway, shielding works if shield is connected to the lowest impedance point.
So if a wire carries signal to a grid and its shield is connected to the "sending" plate which is a lower impedance point than the grid, it's "grounded" well enough in many cases.

Why do they do it so, instead of grounding the regular way?
For one, it's often used in a local feedback connection, and hot wire to shield capacitance (yes, we are talking a few pF only) is the "invisible" compensation cap.
Don't have your Nomad schematic but I have seen this connection in other Mesa models.

4) so in a nutshell I'm saying you have a wild and tricky amp to repair ... if repairing is the word, you'd actually be "reengineering" it, and to boot you don't have access to some obscure Mesa service note explaining how to deal with that.
Not having a scope does not help either .

5) if I had it on my bench, "just because it's there" I would try a couple things but not lose sleep over it, if not solved after a reasonable time I'd let Mesa themselves deal with it.

a) I would start with all controls set to 0, then one by one set them to 10, working backwards, stage by stage, from output towards input jack.

The idea is to find which is the critical unstable stage, or to be more precise, which is "the straw which broke the camel's back".
Which is the one which if put "out of the way" leaves the rest of the amp stable and usable.
Hiss by itself is not a concern, but oscillating, "chirping" is.

b) once you find a stage which seems to be the bad boy, which clearly causes trouble, then you work around it: you (chopstick) move wires around, you can also make a small shield, say a 3" by 3" piece of metal, maybe a piece of raw copper clad PCB material or a small piece of aluminum , which you ground with, say, an 8" piece of wire and move around, trying to shield the input jacks from the rest of the amp, some input or plate resistors, etc.
Just don't touch a high voltage point with it, for safety I'd hold it with insulated long nose pliers.

You may find a position which solves the problem;if so fashion some custom shield there.

As said above, rather than a "repair" job you will be correcting unfinished Mesa design.

Oh well.

EDIT: read this thread which refers to something similar:
http://music-electronics-forum.com/t11464/

[Chuck H]

I haven't ventured in at TAG yet since the warning page started popping up. But from other posts I've read, Leo has it right. I'm just a lurker there (because of their excellent coverage of the TrainWreck circuits) so it hasn't been a priority worth any risk. But if it happened here I would surely just plow through to get in

As far as stabilizing a high gain Nomad... Without a schem and only vague reporting it's virtually impossible to troubleshoot. I've been inside one Nomad head and IMHO it was a bit of a turd. Considering the construction and wiring I saw in there I would expect instability from a high gain version. We may need some photos so that we can see what sort of spaghetti mess is inside. Hopefully that can be provided, until then...

How do we know that the amp hasn't been modified? A bad ground would hint that maybe it has been.

If the ungrounded, shielded cable is for the input lead to the first grid, go ahead and ground one end of the shield at the input jack ground. Double check to see that neither end is grounded before you do that. Sometimes shielded cables are attached to plate circuits and grounding the other end would be very bad in that case.

If that isn't the case, or if grounding the shielded cable doesn't help, can you tell us more about how the problem manifests? Does it happen with or without an instrument plugged in? What other controls do or don't affect the problem?

It sounds like a classic cross talk oscillation that typically results from novice modding. As questionable as I consider the Nomad amp I saw, I doubt the amp was sold with a problem like that.

It's always best to try and track the problem to it's source in these cases, but that can be frightfully hard to reconcile sometimes. You can always attack the problem with signal bleeders. Cull gain, sometimes at frequency anywhere that it doesn't negatively affect the tone too much.

EDIT: Reading Juan's post I realize I may have the identity of this amp wrong. Is this a Nomad brand amp or a Mesa brand, Nomad model amp?

[J M Fahey]

If the ungrounded, shielded cable is for the input lead to the first grid, go ahead and ground one end of the shield at the input jack ground. Double check to see that neither end is grounded before you do that. Sometimes shielded cables are attached to plate circuits and grounding the other end would be very bad in that case.

I would not straight ground that shield, as the last comment says it may be connected to a plate (~160VDC ) , I'd first check that there's no DC there, specially hundreds of volts.
I think it's there more as a capacitor than actual grounded shielding.
I don't think a Mesa bench builder "just forgot to solder a wire".

[Chuck H]

I would not straight ground that shield, as the last comment says it may be connected to a plate (~160VDC ) , I'd first check that there's no DC there, specially hundreds of volts.
I think it's there more as a capacitor than actual grounded shielding.
I don't think a Mesa bench builder "just forgot to solder a wire".

Well I did mention "before you do that" in the area you didn't highlight in red And though I wouldn't expect Mesa to leave a shield unattached either, we don't yet know if this is a Mesa product or that the amp circuit hasn't been altered. The OP reports that the shield is not connected at either end. That SHOULD be confirmed. A VDC reading on that shield is a very good idea as well in the process.

[wormdirt]

To clear the air, the brand is Nomad, and the model is HiGain.

[wormdirt]

School is going to take up most of my day, but I'll report back with more specifics tomorrow.

[g1]

The Nomad brand hi-gain model are reported to be boogie knock-offs, but from the mid to late 80's, so they won't be like the boogie nomad model.
Exactly how high settings for vol & master are we talking about, and what kind of eq settings?
And have you tried pre-amp tube swaps?

We've had those google generated malware alerts at this forum also. It is usually related to ads which, ironically, are supplied by google ad services.
For this reason, an adblock type plugin is always a good idea.

[J M Fahey]

Oh, sorry, I assumed it was a Mesa Boogie product and that by just mentioning model, parent company was a given.
Sort of calling an amp Twin reverb or Plexi.
That said, layout, wiring routing, grounding, shielding checking still applies, as in any high gain amp.
As of the highlighting , it was a sort of rising a red flag (literally) because OP might start reading that answer and grounding the wire shield before even finishing the post.
And I distinctly remember discussions about that floating screen wire, so we could very well find that same trick here.

OF COURSE grounding that screen would kill oscillation ... because it grounds the plate !!!!
In another thread running now, lowell reported grounding a certain cap negative leg instantly killed noise ... and audio too.
Markus reminded him that both effects were not surprising ... he was also grounding a switching Op Amp output.
Which might bring a third unexpected effect: a dead Op Amp

Not knowing poster credentials, I tend to explain things at the most basic and general purpose level ... later we might make them complex if it fits.

[MarkusBass]

Still, I wouldn't be surprised if the amp is similar to one of older Mesa amps (or Marshall). In such a case I draw schematic of the amp myself. It's due to the fact that often you need to see the whole schematic to find out sensitive areas of the circuit.

Mark

[Chuck H]

The Nomad amp I opened up had a lot of odd wiring inside. Things like a circuit board that was constructed somewhat like the old ProCo pedal boards. With thick copper tape glazed with solder. Components attached to socket pins and pot lugs were usually soldered from there like PTP but run to the board where the rest of the circuit was located. Clearly hand wired and clearly built on a budget and quickly. Cheap caps used throughout with ceramics used in many places where a film cap would usually be. Also often paralleled to achieve a higher value. There were actually a lot of repetitive components used in parallel or series in what I'll guess as an attempt to keep inventory simple. Like parallel 470k plate resistors on most of the preamp tubes soldered directly from the tube to a blob of solder on the board. Really "interesting" build. Considering the odd approximated values in circuit due to parallel components, and the probability that Nomad would probably never publish a schematic that demonstrated such a build practice, schematics may not even exist for some models.

The Nomad I looked into was nothing like any Mesa product. Similar setup to a Marshall with parallel input stages, to another stage, cathode follower, tone stack, phase inverter and then to a quad of 1980's Chinese EL34's with absolutely no printing on them. Of course component values were totally different with 235k preamp loads and such. Sorry I don't remember much more about the amp.

Just random thoughts on the brand. I'll look for a Nomad "high gain" model schematic (just because) but no one hold your breath.