Gibson GA50 Clone Noise Issue

[Duck Dodgers]

Previous noise issues were pretty well taken care of by re-doing the grounding scheme per forum input and, in particular suggestions and reference to an article by Blencowe. from “didit”. Thanks guys (and gals?).

Very, very slight 60Hz hum, barley discernible and not affected by volume setting. OK with this.

Also have a “buzz” that is annoying and definitely not OK. Not affected by volume setting. Sounds 120Hz “ish”.

I think the noise is the spikes that are imposed on the 60Hz signal at the output jack. Traced (scope) from output jack back through to plates of V2, inverter. Seems to be on inverter grids as well but my ancient scope can't trigger on the low amplitude.

What I've done:
Re-routed, twisted and whatever filaments. Also ran on 6.3vdc....No change
Swapped 6sn7 between a new JJ and a new Tungsol,,,,No change.
Moved everything into the house; different 120 leg and away from the solar inverter....No change.
Endless poking around....Waste of time.

Some other component changes and voltages are noted on the schematic.

The good news is that the tone is very mellow 50's-60's jazzy with a first year production ES-165 and a mediocre speaker.

[Stevem]

Try sitting the amp on a sheet of Aluminum foil that makes good contact with the chassis .

Are your output tube heaters wired pin 2 to 2 and pin 7 to 7?

[sluckey]

Seems to be on inverter grids as well but my ancient scope can't trigger on the low amplitude.

Set the scope trigger source to Line instead of internal. Then display will be stable regardless of vertical input amplitude.

[Phil_S]

I'm guessing here. I think you may have some B+ mixing with signal. The lead dress might benefit from some revision. Many of the leads are much longer than necessary. It is hard to know without being there. It could be that my suggestions will yield no improvement. Here goes...
-The blue wire on the ground side of the the pair of filter caps mid chassis. Route it directly towards the tubes, folding down onto the chassis floor at the edge of the board, and then tucked into the bottom chassis fold. Run along the fold and then 90 degree bend to final connection point.
-Two reds and one yellow coming from the 6SL7. Take out as much slack as you can. Do not make taught, just tidy and shorter.
-Blue and black between the two power tubes. Take out as much slack as you can. Not taught, just tidy, shorter.
-What is the blue wire from the first power tube, hops across, connects on the buss near the red/black? Depending on what it is, I might be tempted to push this under the board instead of flying like that.
-Three flying white wires going to the pots. It looks like they are near the B+ (120Hz) supply. See if you can run under the board to get them away from the power supply rail.
-Move the two 100Ω faux CT for the heaters (on the pilot lamp) to another ground that isn't the circuit ground. If that was my build, I'd drill another hole in the chassis for that ground because I don't see another convenient place for it.
I think you can tell, I don't care for any of those flying leads. Don't do everything at once. Do this one wire at a time or one group of wires. Quit when it improves to your satisfaction.

Maybe you'll narrow down the source of the problem. I wouldn't be to quick to do any of what I suggested.

[Duck Dodgers]

Try sitting the amp on a sheet of Aluminum foil that makes good contact with the chassis .
Are your output tube heaters wired pin 2 to 2 and pin 7 to 7?

Forgot to mention that I tried shielding the bottom. No effect
Heaters are wired per the schematic. I used a blue and white wire to keep track. It's a little hard to see in the photo.

Phil_S.
It didn't occur to me that there is 120Hz on the B+. Makes sense that the rectification isn't 100%.
Since there is a small signal on the grid(s) of the inverter I'm wondering if the pot leads are picking up 120Hz where they cross the B+.
Seems like a reasonable place to start.

[Stevem]

How close is your choke to the output transformer?

You might try spinning it 90 degrees from the way it faces now.

[Duck Dodgers]

Been trying to dress some of the wiring. Some of the pot leads and the "Blue Wire" ground. No silver bullet . At least not yet.
Also, the blue wire running to V3 & V4 P1 comes from C2. I'm using the empty P1 for the screen resistor connections to P4.

With the scope set to line (thanks, sluckey) I was able to get waveforms from the inverter grids. Scale is 1mv/division. Interesting that P4 (the "inverting" half) is 2+ times the amplitude of P1. Can't find the 120Hz spikes anywhere before V2.

Included a shot of the top component layout. Turning the choke (per stevem) will be a chore. Things are pretty tight. I did try a steel shield between the choke and output trans. No change.

[Stevem]

Can you unhook L2 and jump it for a quick test?

[Phil_S]

Can you unhook L2 and jump it for a quick test?

+1 on that suggestion and my hat is off to Steve for asking about this one!

Looking at placement of the iron, if nothing else is working, I'd be inclined to rotate L2 and the output transformer 90 degrees. If there is enough slack in the wires, you can unscrew them and temporarily tack them down with hot glue or silicone, or something similar. It will be a very brief test and doesn't need to hold for very long. You might be able to use one screw hole to hold them in place, which would be better than using some sort of goo you need to scrape off. If you do this, do it one at a time or you won't know cause and effect.

[didit]

I have little to add to help as you've got the basics down given analysis presented and there has been various others giving sound advice already. It is fair to assume the 120Hz is coming from the B+ chain with the rectifier, power transformer and L1, along with wires connecting it all will be your likely source(s). One thing not mentioned yet, so wonder if you are confident your 5V4 is up to spec?

Hindsight can be painful. There are a few things strongly recommended changing if I had paid attention to the locating of all that iron. The location of L1 would best be on the other side of the power transformer. That output transformer could be turned 90 degrees, or even better positioned using headphone or scope after the power supply is live, to minimize that source of coupling noise.

Good hunting regardless. Sounds like you're getting close to figuring it out.

Good nudge as I really should make finally finishing the build here a priority soon.

Best .. Ian

[bepone]

power transformer need to be rotated 90 degrees, without that, there will be no improvement.

[Duck Dodgers]

Disconnected L2 and jumped the connection. 60Hz hum went up. No change to the 120Hz.

Borrowed a GZ34 from another amp to replace the 5V4. No change.

Replaced the JJ 6sn7 with a new Tungsol. Can't conclusively say it helped but, the grid waveform amplitudes were just about equal as opposed to the 2:1 difference in the JJ.

Turned L1 and L2 90deg (pic below). 120Hz still present but a little better. Not sure if this is due to the re-positioned chokes the Tungsol tube or my ear getting used to it.

I will try to move/turn the OT. over the weekend. Pretty tight for space. What I'm gathering is that it would have been better to have the OT on the other side of the Power Trans. Also, checking some other builds (didit's) looks like the lamination's should be at right angles. Turning the PT might be easier the turning the OT. Live and learn.

I included a drawing of the chassis. If anyone has the inclination to scribble on it feel free.

Thanks!

[Duck Dodgers]

Quick one...

To turn the PT I will have to move the OT as well. So, turning the OT is the way to go if it will help the noise issues. I would leave the PT alone.

[Phil_S]

There is no ideal solution with transformers/inductors. The general thinking is that orientation of the lams at 90 degrees from each other helps to control hum. Since the PT is more work to turn, I'd go the path of least resistance. See if you can turn the OT. Layout is PT OT L1 L2. Orientation to try for is -- | -- |. I see you have temporarily turned L1 and L2 in approximately the same orientation. L1 can stay where you originally put it. In the end, you may need to follow the earlier suggestion to turn the PT, so the orientation will be opposite of what I suggested. This kind of problem can be frustrating. Keep at it. There is a solution that you haven't found yet.

[Stevem]

Keep in mind that only the output trans and chokes have 120 hz in them, so 1) I can’t see how 120 hz could get impressed into the Pt core.

2) he has no issue with 60 hz from the PT getting into the Ot, or at least not to any amount of it being worth messing with spinning it.