Small mischievious troll hiding in my amp

[Chuck H]

This one has me really bent. My personal amp started making crackle and static noises that happen with, but don't interrupt normal operation. These noises only happen from C to C# and any cent interval in between with 50 cents right between being the worst. The first thing I tried was a tube swap with no effect. This only happens with the chassis in the cabinet. Once I remove the chassis I have no problem to diagnose. Vibration? Nope. I plugged into an external cab and I get the same results. Chassis in amp bad, chassis out fine. Also, the amp doesn't need to be loud nor do the strings need to be plucked hard for the problem to present. But to difinitively rule out vibration I poked every component and lead in the amp with no effect. With the chassis in the cab I induced the problem and damped each tube with no change. I even took the chassis and placed it directly in front of the external cab with no problem. Put the chassis back in the cab and the problem is back no matter what speaker I plug into. So, thinking the shield in the cab may be part of the equation I place a sheet of the same shielding that is in the cab on the open chassis with no change. Smacked it around, no change. To further test what I could with the amp in the chassis I scoped the output at the effects loop send and the speaker output. A little voltage spike peppering the waveform is visible at both. I haven't been able to get my signal generator to induce the problem. It only happens with a guitar. I tried the guitar with another amp of the same schematic with no problem. Prior to this the amp has always worked fine since I built it four years ago.

Sorry for the long explaination but I thought it might save on redundant Q&A.

Realizing I was at standstill I actually smiled. A little happy in a sick, wierd way to have such an unusual problem.

[big_teee]

Could it be a heat issue in the cab?
T

[soundguruman]

The little spike peppering the waveform is often the sign of a bad resistor. A lot of times it's in series with a grid.
In hi fi amps we see this kind of thing upsetting the distortion rating.
The resistor breaks down over time and does that spike thing. It may read perfectly good when tested, and show no sign of damage.
After replacing it, bingo, the distortion is gone.
but, it can be anything in any component really. At one certain threshold, it will act up. But usually, it's a resistor...the thing breaks down right at one critical weak point, voltage, current, frequency, etc...
You will see this with carbon resistors because the leads are glued in. Soldering, tiny internal cracks, will eventually give in.
Sometimes freezing them one at a time, then pausing between tests, while the amp is ON will reveal the defective ones. When you freeze it, the crackle may start happening. As it returns to normal temperature, the crackle goes away.

[J M Fahey]

Well, it *is* vibration, of course.
Ugh !!
The problem is finding it (duh !!)
Reproduce the noise, then without touching anything else move the guitar plug to a generator, and sweep the suspect frequency range.
And the anti test: move said plug to another guitar and play the exactsame notes, at same level, etc.
So you differentiate whether said gremlin lives in your amp, or your guitar.
Good luck.

[Chuck H]

Right. And I did try the guitar with another very similar amp with no problems. The only difinitive symptoms so far are that it only happens when the amp is in the cabinet (and I did actually swap it in and out with the amp live. So it's not a heat related issue) and it only happens when I play any tone between C and C# on a guitar (I did duplicate the frequency and sweep with the sig gen, I also duplucated and increased/decreased the input voltage and could not induce the problem with the sig gen.)

SGM. The first thing I need to answer is "Why does it only happen while the chassis is in the cabinet?" Would a bad resistor demonstrate that behavior?. Also, I'm guessing that the frequency specific thing is indicative of a resonant peak or knee. So I'm guessing a cap.

But why only when the chassis is out??? There are actually only three things I can think of to explore. The shielding in the cab isn't perfectly smooth, so that would be a difference? The speaker magnet or EMF from the speaker coil? The speaker location is pretty typical. And I have yet to plug another guitar into it. Mostly because I only play one guitar and my others need care to one degree or another. But should certainly work for testing. I may try some sort of ferrous metal sield between the speaker and the chassis if I can find something that will work.

[Steve Conner]

I once had a similar problem with my bass guitar. It made a fizzling distortion on certain notes. Had me stumped, as it is a passive instrument, no electronics to make distortion.

Turned out that the volume pot on the bass was worn out and the wiper was hardly making contact. Some notes would make the body vibrate harder than others, at which point the wiper would jiggle around on the pot track and make intermittent contact.

A new Bourns pot fixed it.

[JoeM]

Combo amp? What if the amp is in the cab. but no speaker.

[1ampman]

You've tried alot and the only thing would be a thin sheet of steel that's grounded to the chassis (shielding) or a different speaker and cord.... So the amp was working fine for yrs and this just started??? I'll be curious what you find???

[R.G.]

chassis flexing stresses an incipient crack in a resistor, cap, lead or solder joint?

[Chuck H]

Still need to try a different guitar. But since I don't have any problem with another nearly identical amp or when the chassis is out of the cab I don't think it's the guitar.

Take the speaker out!?! Good idea. But what if that "fixes" the problem!?! I don't think I'd like a speakerless combo. But I'll try it just because.

I'd considered a copper or steel laminate on the aluminum chassis over the preamp area. But since that usually isn't needed, and the amp has always been fine before, I wouldn't really feel like I fixed the problem. But if it offers a solution I could probably live with that.

I considered chassis flex. That's why I poked everything in the amp with a pointed stick. I even tried NOT snugging the mounting screws. The problem is still there with the amp just hanging in the cabinet. So, magnet proximity? Why now? And how come this isn't a more common problem with other combo's? Or maybe it is but doesn't usually present like this and it's just part of the "combo sound".

[R.G.]

Hmm. C to C#, worst at 50cents between? Wow, that's tricky. Anything that frequency selective has to be some kind of resonance, either electronic or mechanical, or mechanical exciting electronic. The impulse nature of the noise says it has to be some kind of switching action, either relaxation oscillator style, intermittent contact, or some kind of breakover. (I'm just recounting stuff, thinking out loud here.)

If it were me, my next step would be one of two things. If I was in the mood for an intellectual exercise of hunting the problem down and knowing what it was, I would attach wires to each stage and bring them out of the chassis so I could scope the stage where it first appears. Right now you're limited to the effects loop and output. That would tell you where it first appears, maybe.

If I was frustrated, I would remelt every cotton-picking solder joint in the amp.

[g1]

Suggest you try some kind of insulator (cardboard potato chip box etc.) between the chassis and the cab shield.

[R.G.]

By the way, the troll's name is "Maxwell's Demon". 8-)

[grooveunit]

Try changing the freq of the cab- clamp a hunk of 2x4 to it- see if the resonance is shifted.

[GregS]

Try changing the freq of the cab- clamp a hunk of 2x4 to it- see if the resonance is shifted.

That's what I'm thinking. Even though you tried it plugged into a different speaker cabinet that other cabinet is probably still in close enough proximity to resonate the cabinet your amp is in. That in turn is vibrating a bad connection or component. To determine if acoustic resonance is or isn't a factor maybe you could hook your amp (in it's cabinet) up to a resistive load and scope the effects loop and amp output as you did before for that voltage spike. Just an idea.

Greg