Unusual Hum/Buzz

[svejkist]

Hi guys,

I started up a new Hoffman style AB763 last week, essentially a single channel DR with higher input stage plate voltage and a mid pot. Like a 23W TR, sorta. I have built this amp before. It sounds great, except there is a very buzzy 60hz hum. Naturally, I first suspected heaters. But it doesn't act like heater hum. In the first place, it only happens when there is signal. If the input is grounded, dead silent. If plug in a guitar, the noise starts, BUT, if I turn the guitar volume down, it disappears. The hum is inaudible at low volume (and unsurprisingly, treble) settings. I tried elevating the heaters, and DC heating the input stage, with no change. Strangely, even though though it is signal dependent, I can pull the input tube, in fact all tubes before the PI, and it is still there. Swapped all tubes and, monkeyed with he grounding scheme looking for loops. I used one star ground for the signal and effects ground busses, and one close to the PT for the power ground busses. I severed the artificial CT from the output section ground rail and ran it straight to the power ground rail, as well as a variety of other fiddly ground changes. I grounded all rails from the end of their associated circuit, not the beginning. Of course, I would expect ground issues outside the heaters to be 120hz, but at this point I am trying everything. I've been tracing the circuit and checking values, but it looks correct so far. All ideas welcome.

EDIT: I also checked the bias supply.

[PRR]

> it only happens when there is signal.

Does that mean: signal connected? Or signal sounding?

What I'm thinking is just poor shielding around the wire from input to first grid.

[Bieworm]

+1 on what PRR said.
Is the wire from input to V1 pin 2 a shielded wire,  with 1 side of the shielding connected to ground and the other side clipped off and sealed with heatshrink? And is that 1M resistor good quality MF?

[svejkist]

> it only happens when there is signal.

Does that mean: signal connected? Or signal sounding?

What I'm thinking is just poor shielding around the wire from input to first grid.

Sounding. Like I said, if I roll the volume back on the guitar it goes away. It needs at least the pickup hum. But it isn't the input shielding. I replaced it twice. And Like I said, it is still there if I pull the input valve. Which is why I can't understand how it can be signal dependent.

[svejkist]

+1 on what PRR said.
Is the wire from input to V1 pin 2 a shielded wire,  with 1 side of the shielding connected to ground and the other side clipped off and sealed with heatshrink? And is that 1M resistor good quality MF?

I have redone that all twice, with different pieces of West Pen 291, different resistors, and different jacks, both Cliff and Switchcraft.

[acheld]

Which is why I can't understand how it can be signal dependent.

Based on your observations, I don't think it is signal dependent.   But, because it occurs when you plug in a cable, we're all thinking it has something to do with how you've wired the jacks, or the grounding of the shielded cable from the jack to V1.   

High res photos of your build will help us.  Focus on the input jacks and your grounding scheme for the preamp section.  (I'm assuming that with your PI tube pulled, the hum disappears.)

[svejkist]

Well, it turns out my observations were outdated. Somewhere along the line, one of the changes I made made it signal dependent, and pulling the input tube does get rid of it now. So that makes a lot more sense. Although I still have no idea where it is coming from. This is the jack and signal rail. The jack, volume/tone, and effects ground rail are all grounded to the signal rail, which is grounded at the end of the signal chain to the chassis. I also put a rail across the back of the pots to level their potentials. I have tried that both grounded and ungrounded, to the signal ground rail's chassis ground point.

Pics: https://imgur.com/a/exsXULl

Sound:

Buzzier than I remember say, an ungrounded heater tap sounding.

[Williamblake]

What leads are clipped in?

[Bieworm]

Almost all wires I can see run parallel.  How can this be a noiseless amp?
That bare B+ string is freaking me out! I think you've got a lot of coupling going on in there.
And.. that 1M resistor on the input.. I hope that's shadow I see. It looks as if it's burned from heat

[svejkist]

The only AC wires running parallel are between the output section heaters, because I replaced them, thinking I had reversed the polarity on them (I hadn't, but it wasn't easy to see, because I hadn't coded the wires there) , and it was too big a pain to twist the replacements. And those are parallel to each other, but perpendicular to the signal wires. It is extremely unlikely to cause a problem, since it is at the end of the signal chain, the problem preexisted that, and that would not be signal dependent.

You are going to have to be more specific than "wires running parallel." What wires do you think are introducing 60Hz hum into the signal chain? All the 60Hz wires (except that tiny run far from the input) are twisted, perpendicular to the signal wires, and chopstcking has no effect at all on the sound. Running an uninsulated power rail isn't unusual, and certainly isn't going to contribute noise. Please tell me what exactly you think should not be parallel. I've built dozens of dead quiet amps this exact way. The fact that turning the guitar volume down eliminates the noise suggests to me there is no problem with internal coupling. The heaters are silent, and there is almost no hiss. I don't see how it can be an LF or HF oscillation, since it is 60Hz at all settings and wire positions. I would think that any parasitic oscillation would respond to chopsticking, and you can move every wire on this with no change. Other than being signal dependent, it sounds and acts like a heater grounding issue, but no changes to that circuit have any effect.

My wire coding is:

Yellow- signal
Red- B+ (twisted ones are AC wall voltage)
Orange- cathodes
Green- ground
Black- oddball stuff like trem signal
Twisted Yellow/Orange- heaters/

I don't see how having signal wires roughly parallel with DC wires is going to cause a problem.

That is a shadow on the resistor pic. 

[svejkist]

What leads are clipped in?

None now. I was just still monkeying with different ground points, looking for loops (even though those would normally be 120hz) when I took the video. Those clips were all grounds.

[mresistor]

wires running parallel


He's talking about your lead dress,  for instance all the wires running parallel that go from the turret board to the tube sockets. 

[svejkist]

wires running parallel


He's talking about your lead dress,  for instance all the wires running parallel that go from the turret board to the tube sockets.

I understand that. If you look at those wires by function/color code, I see nothing that is going to introduce 60Hz hum. Not all wires are created equal. They have different functions. A filtered DC B+ wire or a DC bias wire isn't going to introduce 60Hz AC noise into a signal wire. The wires carrying 60Hz AC are all perpendicular to the signal wires. It isn't clear to me how you are supposed to wire this layout without any parallelism in the B+, signal and cathode wires. They are on the same socket and grouped together on the board. The layout itself shows them wired this way.

[sluckey]

If the input is grounded, dead silent. If plug in a guitar, the noise starts, BUT, if I turn the guitar volume down, it disappears.

This indicates the hum is not in the amp. It's being picked up in the room. Fluorescent lights are a big 60Hz noise generator. So are light dimmers.

Put a metal cover over the open side of the chassis. If no sheet metal available just cut a piece of cardboard to fit. Then cover the cardboard with aluminum foil. Does the amp get quieter?
 

[PRR]

I don't care about bare B+ wires. Or signal running with them. Heaters do not have polarity.

The only place you *should* have 60Hz is from the wall. In your video tour this area is visible for only a few seconds, not clearly.

Is that red/yellow wire your HV CT? Shouldn't it go direct to the first filter cap, instead of dumping buzz into the chassis? Of course that would be 120Hz and you seem to know the difference.

Is there any chance your rectifier is half-broken? I see a tube socket and some 1N4007 which I would have to inspect very closely.

AND Sluckey's #13 post. I too thought it "had to be" external but figured: you are there, I'm not. BTW: 50 foot power cord out in the driveway is a fair 2nd-opinion on hum. Gets you out and away from under the lamps, dimmers, WiFis, icebox, and other hummers. (Ah, but not under the overhead line to the power meter.)

[svejkist]

By heater polarity I just meant accidentally swapping the pins, so they weren't consistent between tubes. My understanding is that would put them out of phase with each other and introduce hum. Although I thought that only mattered between the center tapped heater tubes, I figured it was worth a shot.

Yeah, that is the HV CT. I'll move it to the first cap.

Not the rectifier. I have tried it with a couple of tubes, then put in those SS diodes for test purposes. And replaced them too, to be safe. Also, the B+ is the correct value.

Here is what I am now thinking: Bad ground to the wall in the cable. I meant to test that yesterday when I was there taking the stills, but forgot. I left the original cable from the Peavey donor chassis in, and it has a mechanically attached plug on it. So there is a pretty good chance they miswired it. It isn't obvious to me that this would cause the behavior, but  ¯\_(ツ)_/¯  .

Thanks.

[svejkist]

If the input is grounded, dead silent. If plug in a guitar, the noise starts, BUT, if I turn the guitar volume down, it disappears.

This indicates the hum is not in the amp. It's being picked up in the room. Fluorescent lights are a big 60Hz noise generator. So are light dimmers.

Put a metal cover over the open side of the chassis. If no sheet metal available just cut a piece of cardboard to fit. Then cover the cardboard with aluminum foil. Does the amp get quieter?
 

It really does seem that way. I mean, it is dependent on the pickup signal (hum). I had just never had environmental AC be anywhere near this loud, and I have tried a variety of shielded input configs (isolated Switchcraft, grounded Switchcraft, Cliff) with fresh wire and resistors. I have induced pickup hum by holding the pickups near a source, and it sounds like it rides on top of the existing buzz. But that is pretty subjective, I guess.

Here is what I realized: I think the ground of the AC in (wall, earth, whatever) cable may be broken. Not sure that this would cause this, but it is plausible. I'll be over there in a few days and can check it. Meant to do it yesterday, but forgot.

Even if it is environmental, it means I have done something dumb. I have built dozens of amps, including this same circuit, on this same bench. I even had another AB763 I built open on it in between troubleshooting sessions on this one. The physical location in itself is OK. Ill check that cable, and shield the chassis, and get back. Thanks.

[PRR]

By heater polarity I just meant accidentally swapping the pins, so they weren't consistent between tubes. My understanding is that would put them out of phase with each other and introduce hum....

Have you tried deliberately swopping on an amplifier that is already hum-free?

I know enough of Peavey's production process to really doubt they screwed-up the power cord (more than most builders, Hartley hates to kill customers), but it sure is a thing to check.

[svejkist]

By heater polarity I just meant accidentally swapping the pins, so they weren't consistent between tubes. My understanding is that would put them out of phase with each other and introduce hum....

Have you tried deliberately swopping on an amplifier that is already hum-free?

I know enough of Peavey's production process to really doubt they screwed-up the power cord (more than most builders, Hartley hates to kill customers), but it sure is a thing to check.

I have not tried that. I just read it somewhere, and thought it was why people color code their heaters. So when you twist them, you get all the tapped pins (9 in this case) and on one wire, etc. Why use two colors of wire if that isn't an issue? But it isn't in Merlin's heater section, so I assume it isn't true. Also, you just said it isn't true. Live and learn.

This isn't Peavey's handiwork. Someone replaced the plug with a big honking clamped on affair. Presumably it got cut off at some point. I probably should have rang it right off, but I didn't. I have full faith in Hartley Peavey's desire not to kill me. Less so in Leo's. Leo tried to kill me multiple times.

[svejkist]

By heater polarity I just meant accidentally swapping the pins, so they weren't consistent between tubes. My understanding is that would put them out of phase with each other and introduce hum....

Have you tried deliberately swopping on an amplifier that is already hum-free?

After some googling, it appears this is a myth propagated by Gerald Weber. Once this is done, I'll swap them for empirical confirmation.

[svejkist]

By heater polarity I just meant accidentally swapping the pins, so they weren't consistent between tubes. My understanding is that would put them out of phase with each other and introduce hum....

Have you tried deliberately swopping on an amplifier that is already hum-free?

I know enough of Peavey's production process to really doubt they screwed-up the power cord (more than most builders, Hartley hates to kill customers), but it sure is a thing to check.

Yep, it was an air gap in the power cable earth. Amp is dead silent now. Thanks guys!

https://imgur.com/a/rgzakqo