Tracking down noise

[idontknowyou]

Got a real weird noise problem that I am having trouble tracking down.

I built an amp about 18 months ago. Its effectively a JCM800 2204 with a few tweaks to the preamp and it runs KT88s instead of EL34s. I used the mojotone turret board. So far it's been perfect, I took it on tour and it worked great until the last couple days and now it makes a loud buzzing noise intermittently.
The buzz is affected by the gain and master volume, but is still present even with v1 and v2 removed. I swapped out all 3 12AX7s and have tried 2 sets of good KT88s. I have cleaned and tightened all tube sockets and reflowed almost every solder joint.
Here's the weirdest part:
When I initially power on the amp, the buzz is present, but I found that when I even gently tapped on the power tubes, it would go away for a bit. Then if I wait long enough or play loud through it, the buzz returns. I thought it was clearly the tubes but when I put in a new pair of svetlanas, the buzz is present and DOESN'T go away when tapping on the new tubes. If I replace the originals (rubys), the tapping works to stop the noise.
I feel like im going a little crazy.

I haven't yet gone through the amp to measure every single component, but I have double checked the heater balance resistors and they are fine, the grid stoppers and screen resistors on the power tubes, all good. All ground connections seem solid, but it really seems like something is losing grounding briefly. But everything still conducts and the amp functions, the buzz is just really loud and comes in and out.
Anyone had anything like this before? Any tips on trying to find the source?

[sluckey]

Disconnect the NFB wire. Any joy?

[Merlin]

Could be a lot of things.I'll place my first bet on: a connection you forgot to solder when you built it.

[idontknowyou]

Here's a video to hear the noise

[idontknowyou]

Disconnect the NFB wire. Any joy?

This build has presence and an added resonance control. Should I just lift the connection to the impedance selector switch?

[idontknowyou]

Disconnected the wire from the presence and resonance controls leading to the impedance selector and the noise persisted.

[Esquirefreak]

Kinda sounds like a bad ground connection.

/Max

[idontknowyou]

A small update: I have gone through and reflowed many solder connections, including all the grounds, and checked continuity, all grounds seem to be solid. I have one main ground bus for almost every ground point in the amp, excluding the mains safety ground, the heater balance resistors, and the power tubes (they are grounded at their sockets). I fail to see how this could be noise from a bad ground, especially since the amp has been functioning perfectly for over a year and probably 80+ gigs.
The noise going away by tapping one set of power tubes but not the other has me incredibly confused. Should I try replacing the sockets?
Its also not causing a loss of signal, just this huge noise. If the ground was intermittent I would think the tubes would stop conducting, no?
Im also thinking this might be heater noise, but im getting 6.3vac at the heater pins and the balance resistors test ok.

Could a bad transformer sound like this?

[tubeswell]

chopstick it

[idontknowyou]

I have. I have pushed and prodded on every single connection in the amp to no avail. Earlier tonight I also got an alligator clip with one end to ground and verrrry carefully touched it to every point in the amp that should be grounded, and no change.

[tubeswell]

I have. I have pushed and prodded on every single connection in the amp to no avail. Earlier tonight I also got an alligator clip with one end to ground and verrrry carefully touched it to every point in the amp that should be grounded, and no change.

solder a 600V 0.1uF film cap to that wire and do it again

[tubenit]

What do you use to hold the power tubes in place?  Bear claws, springs, gravity ..... ?


With respect, Tubenit

[bmccowan]

If those last 2 suggestions do not lead to any joy, I suggest a listening amp. You won't need a signal generator because the buzz is your signal, you just need to safely intercept the signal path and send the signal to another amp. If unfamiliar, there are discussion threads and instructions on the Forum. Start at the output end and work your way back up the signal chain.

[idontknowyou]

I have. I have pushed and prodded on every single connection in the amp to no avail. Earlier tonight I also got an alligator clip with one end to ground and verrrry carefully touched it to every point in the amp that should be grounded, and no change.

solder a 600V 0.1uF film cap to that wire and do it again

Will do this and report back. Though can I ask the principle behind this? Why is this different/better than a direct ground? Just trying to learn as much as I can.

[idontknowyou]

What do you use to hold the power tubes in place?  Bear claws, springs, gravity ..... ?


With respect, Tubenit

Gravity. Its a head and the chassis sits upright like a Marshall. Also, I built this inside an old AIMS chassis. Gutted everything but the sockets and reused the power transformer. The power tube sockets are pretty close together and they're kt88s so i don't think bear traps would fit.

[idontknowyou]

If those last 2 suggestions do not lead to any joy, I suggest a listening amp. You won't need a signal generator because the buzz is your signal, you just need to safely intercept the signal path and send the signal to another amp. If unfamiliar, there are discussion threads and instructions on the Forum. Start at the output end and work your way back up the signal chain.

Thank you, I will try this next

[shooter]

Gutted everything but the sockets

i'd gut them n get some quality ones with hold-downs




You used a "ground wire" to test grounds
tubeswell is "shorting" AC to ground

[tubeswell]

Yep, it shunts noise to ground so you can work out where the noise is coming from if you work systematically through the signal chain from input to output.


Be careful with the cap. Don’t hold it in your bare hands when you’re prodding through the amp. Use an insulated test clip to attach it etc

[bmccowan]

Agree that that is a good method. The probe is essentially the same as for a listening amp, but the signal is taken to ground rather than to another amp. So as you probe the signal path, the buzz should disappear at some point(s). I find it difficult to work with a bare cap in crowded areas around tube sockets, making a probe that only leaves the tip of the cap lead exposed helps me avoid unwanted surprises. But I am old, arthritic, and a bit shaky. I slip the cap inside an empty ballpoint pen body and trim the tip - but there are many ways to do it.

[idontknowyou]

Ok, finally an update. I just used a 0.1 film cap to probe the signal chain, and as soon as I touched it to the 2nd half of V1, both the plate and the grid (the side that is connected to the low gain input). Nothing happened on the other half of V1. I also probed the b+ node feeding both plate resistors for V1 and it did nothing. I then removed the tube and probed the same spots and it still removed the hum even with the tube out. (The hum is reduced when V1 is out, but its just quieter)

The hum is also eliminated when i probe these spots down the line:

- grid of 1st half v2
- plate of 1st half v2, which is also the grid of the 2nd half
- cathode of 2nd half of v2 (cathode follower)
- grid of 1st half of V3 after the master volume, but not the other grid

So it seems obvious that the noise is following the entire signal path up to the phase inverter. The hum did not go away when probing the grids of the power tubes.

[Dave]

Fluorescent light nearby?


Dave

[idontknowyou]

Sadly this hum has been present in multiple locations with no florescent lights around.

I think my plan will be to try replacing a couple resistors (ive dealt with many noisy carbon comps but this seems ridiculous) and checking for DC leakage.

[shooter]

The hum is reduced when V1 is out, but its just quieter

The hum is also eliminated when i probe these spots down the line:

so it gets quiet "through V1" but gone completely

grid of 1st half v2

IF SO;
you most likely have "cross-talk" between input jack n V2 grid....Sensitive signal close to higher power signals or power wires.
other possibilities, bad sold connection, sloppy tube pins, sketchy grounds, ground loop

[idontknowyou]

The hum is reduced when V1 is out, but its just quieter

The hum is also eliminated when i probe these spots down the line:

so it gets quiet "through V1" but gone completely

grid of 1st half v2

IF SO;
you most likely have "cross-talk" between input jack n V2 grid....Sensitive signal close to higher power signals or power wires.
other possibilities, bad sold connection, sloppy tube pins, sketchy grounds, ground loop

Oh sorry, my wording must have been confusing. I meant that the hum goes away completely when I probe any of those spots. And when I remove the V1 tube, the hum volume goes down overall but doesn't go away

[tubeswell]

If the hum goes away when you A/c ground the grid of V2, the hum source is prior to V2. What happens if you ground the connection immediately before the V2 grid? (Or the connection before that?)

[Carlsoti]

I may be off the mark here, but the two AIMS amps I have use a steel chassis and thick aluminum faceplate. Is it possible that galvanic corrosion has created an unnecessary impedance between the two? Have you tightened up all the jacks and pots to ensure solid ground connections?

[idontknowyou]

I may be off the mark here, but the two AIMS amps I have use a steel chassis and thick aluminum faceplate. Is it possible that galvanic corrosion has created an unnecessary impedance between the two? Have you tightened up all the jacks and pots to ensure solid ground connections?

I removed the face plate and drilled new holes for the pots/jacks directly into the steel chassis. Im using cliff style jacks but the ground connections have a direct wire connection to the ground bus, as do all of the pots

[idontknowyou]

If the hum goes away when you A/c ground the grid of V2, the hum source is prior to V2. What happens if you ground the connection immediately before the V2 grid? (Or the connection before that?)

Well, when i probe the 2nd half of V1, the hum goes away as well. This is basically a 2204. My version has some small deviations from this schematic but for simplicity's sake let's refer to the attachment. The earliest part of the signal chain where the hum goes away when i use the probe is the wiper of the preamp volume pot, which is electrically the same as grid 2 of V1 (pin 2 on the attached schematic screenshot). The hum does not go away when i probe on the other side of the resistor/cap connected to the low gain input.
Bad component there? Im using a silver mica for the 470pf. Its a CDM so of decent quality but I've had some bad luck with silver micas before. There's no bright cap on the preamp volume pot. Bad pot maybe? The pot still functions properly so its ground connection must be good and the wiper still is making contact.

[shooter]

cut out both 470p and test

[idontknowyou]

Still present with both 470pf caps removed

[bmccowan]

Bypass the pot with a fixed value voltage divider and test?
Any chance of a cracked resistor in a path to ground?

[shooter]

what type input jacks did you use?
are all the preamp grounds connected together and grounded near the input jacks ONLY?

[idontknowyou]

what type input jacks did you use?
are all the preamp grounds connected together and grounded near the input jacks ONLY?

Shorting cliff jacks, standard marshall style.

I am using a single bus for almost every ground in the amp, with the exception of the power tubes which are grounded at their sockets, and the safety earth ground. The termination of the ground bus is by the power transformer.

Im more happy to re-route grounds and experiment with it, but I would be very curious to know why/how the amp would work perfectly, and amazingly quietly, for over a year and then manifest suddenly like this if it were an inherent ground problem.

[shooter]

work perfectly, and amazingly quietly, for over a year and then manifest suddenly

if i had a dollar.....


ever have a car just die, blender.......the only answer lies in the solution

a single bus for almost every ground in the amp,

take the time, look n read the grounding posts in archived posts, favorite posts

i would not rule out the cliff-jacks, they got used X times a week for over a year

[dude]

Usually there are two grounds in an amp, star ground for power section SEPARATE from the preamp grounds, preamp grounds down by the input jack. Preamp grounds should be on their own ground close to the input jack with a wire from jack bolted to the chassis at input area.
If you have a buss bar don’t ground the bar in two places, make sure the power section ground is as far away from preamp ground bolt as possible. The buss barr should be grounded in one place and just hanging not connected to the power supply ground.  A ground loop can cause buzzes like that video. I think Shooter is on the right track.
Buss bar should pick up the grounds from the pots down to input and then that bolt. Power supply caps, PT, CT’s, bias, OT down by power section. Preamp E-cap ground to input ground but l’ve seen many preamp cans all grounded to power supply ground and no issues but if able ground to input bolt.

[idontknowyou]

Haha fair enough! I definitely get stuck on "this used to work, what changed" mentality. Sounds like i somehow got away with an unideal ground situation for awhile.

Since building this amp i have gotten merlin blencowes books and have been slowly wrapping my head around them. Will spend a bunch of time on the grounding chapter, and will try to separate the preamp grounds from the power amp stuff.

In general, should the ouput tube grounds be with the power section? What about the presence/resonance controls?

[tubeswell]

I can't see anything in your posts from the start of this thread about whether you subbed a known good tube into the V1 slot already? Did you?

[bmccowan]

In general, should the ouput tube grounds be with the power section? What about the presence/resonance controls?

This page includes a JCM800 6v6 layout with a good grounding scheme - it might be helpful. https://robrobinette.com/RR2104_Master_Volume_Micro.htm

[idontknowyou]

I can't see anything in your posts from the start of this thread about whether you subbed a known good tube into the V1 slot already? Did you?

Yeah, tried known good tubes in every position to no avail

[BrianS]

You have to start looking at non-obvious stuff.  You may have a wonky V1 tube socket.  I can't say specifically what might make it wonky, but maybe poking at it with the chopstick will reveal something.  Maybe sharpen it on one end so you can poke down into the pin connectors.
I'm surprised no-one has mentioned your metal workbench.  Seems like a bad idea to me. 

Good luck.

[tubeswell]

What happens if you put the A/c grounding cap on the plate of V1a?

[idontknowyou]

You have to start looking at non-obvious stuff.  You may have a wonky V1 tube socket.  I can't say specifically what might make it wonky, but maybe poking at it with the chopstick will reveal something.  Maybe sharpen it on one end so you can poke down into the pin connectors.
I'm surprised no-one has mentioned your metal workbench.  Seems like a bad idea to me. 

Good luck.

Thanks. Yeah my next plan is to replace the volume pot (worth a try?) and play with grounds, and maybe replace tube sockets.

Most of the time i work on amps horizontally lifted on boxes or wood blocks so it's isolated from the table, but you're right, its a pretty bad idea

[idontknowyou]

What happens if you put the A/c grounding cap on the plate of V1a?

Nothing

[tubeswell]

What happens if you put the A/c grounding cap on the plate of V1a?

Nothing

So by 'nothing', you mean it still hums?


What about the plate of V1b?

[idontknowyou]

What happens if you put the A/c grounding cap on the plate of V1a?

Nothing

So by 'nothing', you mean it still hums?


What about the plate of V1b?

Yes, no change in the hum when I probe the plate of V1a.
Hum is gone completely when I probe the plate of V1b

[tubeswell]

What happens if you put the A/c grounding cap on the plate of V1a?

Nothing

So by 'nothing', you mean it still hums?


What about the plate of V1b?

Yes, no change in the hum when I probe the plate of V1a.
Hum is gone completely when I probe the plate of V1b

So it's around V1B then.

It could be a bad plate resistor (or coupling cap) on V1b - try prodding V1b plate resistor and/or coupling cap with a chopstick (and be careful not to electrocute yourself)

[idontknowyou]

So it's around V1B then.

It could be a bad plate resistor (or coupling cap) on V1b - try prodding V1b plate resistor and/or coupling cap with a chopstick (and be careful not to electrocute yourself)

I tried replacing the plate resistor of V1b the other day and it made no change

[BrianS]

Also check your input jack wiring and look for sprung contacts on those Cliff jacks.  The amp makes the buzzing sound when you have a guitar plugged in though, right?

[idontknowyou]

Welp, it was the ground scheme.

After temping in a new preamp volume pot and getting no results, I had eliminated every possible source of noise around V1b except the tube socket. I figured playing with the grounds would be slightly less labor intensive than replacing the socket, so I snipped the ground bus in half, installed a machine screw near the input jacks (one nice thing about re-purposing a chassis for a simpler circuit is having lots of extra holes! Thats a good thing, right...?) with a solder lug, and ran a wire from the preamp end to this new ground point. I also ran a new wire between both power tube cathodes to the power amp ground (after separating from the existing ground tab at the sockets), aaaaaaand voila! No hum!

Thank you all for your help. As a novice who is eager to learn more, this forum has been a huge help and wealth of information.

I'm left with lots of questions. Why did it work fine for over a year? Why did the hum go away when tapping on one set of power tubes but not the other? Why did the hum seem to not reach V1a? But all of these questions are a reminder that there are so many extra unknown variables when putting theory into practice, and (most of the time) that's the fun part.

[BrianS]

It's unlikely it was the ground scheme.  You yourself claim that the amp worked fine for over a year.  You traced the hum to a specific point, and then chose to ignore that data and try something else based on some comments from a few people.

The only thing you really learned is that you changed something, and that seems to have solved your problem.  You did too many things at one time to truly know what the actual problem was.  That is fine...unless the problem crops up again. 

You say you are a novice, so, if you want to actually learn stuff,  I believe it's important for you to not draw the wrong conclusions from your process.  With more patience, you could have found the specific thing that was causing the problem.

I disagree with the following quote. This technology is around 100 years old.  There aren't really any "unknown variables".

But all of these questions are a reminder that there are so many extra unknown variables when putting theory into practice.