Gibson GA-20 hum

[Diverted]

Hi all,
I was just given a very clean Gibson GA20 (12AX7 version) to look at. Owner bought it on Reverb last week and it’s got a pretty good hum. I haven’t been able to knock it out and am wondering if there is something I missed.

The hum is present with both volume controls at 0, with V1, V2 pulled. No sound out of amp with V3 pulled.
Running the amp at 115VAC, all voltages right on spec per schematic. The amp had been recapped recently: New Illinois 22uf filter caps and new Sprague 6V6 cathode bypass cap, new Mallory 150 coupling caps to 6V6s.

It looks like someone else tried to address the hum by elevating the 6.3v center tap to 6V6 cathodes. In any case elevated center tap isn’t on the schematic.

Here is what I did to try to trace out possible issues:

Checked filter caps, ran new F&T caps in parallel to them temporarily. No change.

Checked power stage coupling caps for leakage; ran new .01 caps in parallel to stop any DC leakage that might be there. No change.

Temporarily disconnected the 6.3v heater tap and put in an artificial center tap just in case that center tap wasn’t well balanced. No change in hum.

Temporarily moved center tap to ground. Hum increased marginally, so I put it back on the cathodes.

Checked all resistors for drift. All are within 10 percent of spec.

Chopsticked quite a bit, moved wires etc. Very little change.
Cleaned jacks, checked all grounds, solder joints etc. All appear to be fine.

What else could I check here? I’m wondering if the sloppy heater wiring is introducing hum. The pairs aren’t twisted and are flying all over the place at the bottom of the chassis.

[shooter]

ran new .01 caps in parallel to stop any DC leakage

need to run  in SERIES to block DC

have you identified 120hz hum, 60hz hum, or some random hum?

[Diverted]

Ran caps in series, no change in hum.
I believe its 60 cycle, but I am not sure. I can post a sound clip in a little bit here. Thank you,

Ted

[shooter]

Microsoft won't let me listen 

you want to know because it helps get one the right track

[bmccowan]

I have repaired a few of these and have an earlier all octal version on the bench right now. Reducing hum in these takes patience. Elevated heaters is the common setup in these amps - what schematic are you referring to? The schematics I have for 9 pin and octal versions have the elevated heaters. The grounding scheme is kind of strange - you may want to do some experimenting there, as the ones I've worked on tied all grounds together and runs them to the input jacks. I have grounded the power tube cathode resistor/bypass cap near the PT with a separate terminal and stuck with the rest of the Gibson scheme. I have seen OT failure in the ones I've worked on, but that was total failure, not a hum issue. The one I have on the bench now is 6SJ7x2, 6SL7, 6V6x2 and 5y3. It has the non-twisted heater wiring you describe, but lots of amps do. Is the heater wiring properly referenced to ground? Have seen that issue in many amps and lots of hum results.
Keep at it - the GA-20s are great sounding amps, and a real alternative to Fender (or as Shooter says, car parts amp) clean.

[Latole]

The only schematic of G20 with 12XX7 tube and heater with center tap I see use  12AY7 and 12AX7

https://el34world.com/charts/Schematics/files/Gibson/Gibson_ga20_12ay7_preamp.pdf

Can you show us the right schematic ?
And pictures ?
_______________
You wrote :
"  I’m wondering if the sloppy heater wiring is introducing hum. The pairs aren’t twisted and are flying all over the place at the bottom of the chassis. "

Yes it may be your issue. A well done heater wiring with center tap is a must. Start by fixing that.

Amp may have many issues, you may need to fix all, not only one.
Noisy tube (s)
Unmatched 6V6's

Also tubes's plate resistors come noisy with time, replace them all. They hiss more than hum

[Diverted]

I ended up re-wiring the heater winding for him, as he wanted it done. It lessened the hum level almost to the level I wouldn’t be surprised to hear. Still trying to root it out further. Here is the schematic as well. Thank you.
Sorry for the photo. It’s the only one I have at the moment; I’m at work. I will post some circuit pix when I get home. I appreciate the help.

[Latole]

3 prong power cable with good ground is a must. Did amp have one ?

Heater wires look good.
Schematic show 6.3 volts transformer secondary have no center tap ! A must. Built one with two 100 ohms resistors.

[bmccowan]

Schematic show 6.3 volts transformer secondary have no center tap ! A must. Built one with two 100 ohms resistors.

Do not trust the schematic. Gibson constantly changed circuits and did not stick to their own schematics. All who have worked on early Gibsons have experienced this. Diverted said that the heaters were elevated at the PT cathodes - that indicates a center tap.
Glad the heater wiring helped. Other things I've done to help old Gibsons:
Adding a ground point near the PT and grounding the first two filter caps and the power tube cathode resister/bypass cap there.
Use only one ground point for the preamp grounds - Gibson tended to have multiple ground points.
Make sure the input jacks are in good shape and grounding when not used.
Replace plate resistors with metal film resistors.
Clean all the tube sockets and clean/lube the pots.
That amp looks to be in good shape. The one on my bench was beat! All resistors way out of spec. Bad OT. Bad speaker cone. Exploded filter caps, etc.

[Diverted]

Agreed. The schematic isn’t right on several counts. The transformer has a 6.3v center tap. I actually removed it and tried an artificial center tap in its place in case there was a flaw in the winding and the center tap wasn’t centered along the coil. No change so I put the CT back in circuit. Another difference from the schematic is the center tap running to the 6V6 cathodes; that’s not shown. Not sure if it is original or if a previous owner had tried to elevate it a bit to cut down on the hum.

[Diverted]

Also, yes the three prong had ground, but person who installed hooked it up to the HV center tap/cap/etc node. I disconnected it and gave it its own grounding point.

[Diverted]

I may pull all the filter caps and lay them out in a better arrangement, and shuffle the grounds a little bit. You are right; there is a terminal strip with several grounded lugs that are being used. I’ll probably just pull the screws that connect the strip to chassis and run nylon screws in their place, then run those points to the preamp ground buss which connects from the first input jack across to the pots.

[shooter]

The schematic isn’t right on several counts

 
done a handful of G amps, our job besides fixing the amp is to make the next techs life easier - send it on with corrected docs 

[Diverted]

 I’m not sure what the original grounding scheme was, but I shuffled the caps around a little bit. The first filter is tied off to the power transformer lug, which also serves as the ground for the 6V6 cathodes. The second two filters were tied off at a midway point on a grounded lug of the terminal strip. I moved them to the preamp ground bus which runs through the jacks and into the pots.

Rooting around, grounding V3B (PI) grid eliminates the hum and all volume. Grounding V3A grid, and/or pulling V1 and V2 have no effect on the hum. Issue is with the phase inverter? Or further down stream? 6V6 coupling caps have been replaced with new Mallorys.

[bmccowan]

Good idea Shooter.
So here are a few shots of the earlier GA-20 that's currently on my bench. It was so filthy that I was tempted to remove the PT and power wash the chassis! But I cleaned it up and found all the resistors way off spec and a bad OT. So I documented the layout and started a gut rehab. I relocated (this is starting to sound like a housing project, eh?) the filter caps and grounds and followed the original layout for the rest. Sounds really good except that the speaker needs a recone and can't handle low notes. Cabinet's a mess too, with 90% of the covering (thin book cloth, not Tolex) torn off. I'll recover or wait until reliced amps become the rage like guitars (stupidly) have become.

[shooter]

grounding V3B

try clipping in a 47uf across the 10uf feeding the PI

do you know if it's 120 or 60hz?

If that didn't do anything, hang the same cap across the cathode cap for the 6V6's

[Diverted]

Stumbled upon something interesting, and while I don’t know if it’s a solution, perhaps it can help get there.

THe 6.8K resistor on the same node as the V3B grid and 220K to the 6V6 ... I grounded that at that junction, grounding both the 220K and the V3B grid.
This COMPLETELY eliminated the hum. The amp sounds fantastic, quiet as a mouse at idle, zero hum. My question is this, how can I use this info to address the amp? Obviously the 6.8K is in the circuit for a reason (provide some imbalance to the power tubes/proper biasing on V3?) and I don’t think it would be wise just to cut it out and ground that point.
So ultimately how can I use this information to get rid of the hum?

Thanks!

[shooter]

So ultimately how can I use this information to get rid of the hum

I'm not a PI guy so;
the 6.8k is needed, V3a gets "tapped" at the plate to "feed" V3b.
since you're narrowed down to the PI, if it's yours I'd rebuild the PI starting with the socket.  If it's a customer.....maybe a talk without guarantees 
have you lifted the cap the jumps the plates?  that can be eliminated, but might stop some noise

[bmccowan]

Like Shooter I am not a PI expert by any means, but I imagine that PRR, Sluckey or another will join in. But I agree that you have likely found the location of the problem. The voltage is divided at that location, sending some of the signal to the other half of that PI tube (I think). But I only really want to point out that I don't think you mentioned trying different tubes. If you did not that is the easiest place to start - replace that PI tube with a known good tube before you get the soldering iron warmed up.

[sluckey]

It's a paraphase PI. Read all about it on the net. When you short out that 6.8K you remove the grid signal to the bottom half of the PI tube, which removes the signal from V5. So now your Push/pull amp is only operating on V4! Don't do that.

[Diverted]

Like Shooter I am not a PI expert by any means, but I imagine that PRR, Sluckey or another will join in. But I agree that you have likely found the location of the problem. The voltage is divided at that location, sending some of the signal to the other half of that PI tube (I think). But I only really want to point out that I don't think you mentioned trying different tubes. If you did not that is the easiest place to start - replace that PI tube with a known good tube before you get the soldering iron warmed up.

I tried quite a few tubes in all positions, all known good. There was little to no change from one to the next.

[Diverted]

It's a paraphase PI. Read all about it on the net. When you short out that 6.8K you remove the grid signal to the bottom half of the PI tube, which removes the signal from V5. So now your Push/pull amp is only operating on V4! Don't do that.

Got it. I’m not doing that. I guess then that it was interesting to see it eliminate the hum, but now I’m pretty much back at square one. I don’t want to just throw parts at the problem. I would really like to try to figure out the root of the issue and address it correctly. Just don’t know what else to try. Thanks for the words.

Ted

[sluckey]

Measure resistance of that 6.8K to chassis. (Don't just put the meter leads across the resistor.) Then measure the resistance from pin 5 of V4 and V5 to chassis. What have you?

[Latole]

Test  6V6's:  how they are matched or not.
With 1 ohms resistor at plate; read current

[bmccowan]

I mean no offence to Latole, Shooter, or myself. But follow along with Sluckey. From what I've seen he will walk you right through it. And of course do read up on phase inverters, but be prepared to be confused, I always am.

[Diverted]

Thanks.

6.8K to chassis: 6.62K

V4 5 to chassis: 221K
V5 5 to chassis: 224K

[66Strat]

Excerpt from RC19 regarding paraphase PI.....

[PRR]

> THe 6.8K resistor .... .... ....This COMPLETELY eliminated the hum.

Very strange. This kills half the push-pull. Power output will be down. More to the point: proper push-pull action can *reduce* some forms of hum/buzz (particularly supply crap). Here it does the opposite.

Are there any AC or dirty DC lines in the middle of that inverter? Is the ground under the 6.8k good?

Lighter schematic with all official typos preserved.

[bmccowan]

True to Gibson heritage, the amp Diverted is working on has some elements of that schematic, and some from this one.

[Diverted]

The heater wiring was messy throughout. I moved it out and tucked it up against the edge of the chassis. That cut down the hum some, but it’s still there. The heater wires are now a pretty decent distance away from sensitive leads, I think, and chop sticking throughout has no effect.
I honestly can’t find anything else wrong ... have traced out circuit many times and all is in place  ... all resistors are on spec. Coupling caps to output stage are brand new, filter caps brand new. Voltages check out vs. schematic.


I don’t know what the original grounding scheme was but I have the first (new F&T) filter on the power ground; the second and third (they are now F&Ts, the pics show the Illinois caps that came with the amp and are now gone) are on the preamp ground which runs from the first input jack, through the three others, into the pots, and via a wire down into the 12AX7 socket, then the next and the next, which have a grounding center. from there, other stuff is tied into them. Before settling on this I tried different configurations: All three on the power ground at transformer lug, first two on the power ground with third on the preamp, etc etc etc. None eliminated the hum and were all very close, hum-wise.

I’m going to reestablish the main power ground by removing, putting a new grounding lug in, cleaning chassis around/underneath it and re-bolting. The little solder lug tied to the transformer screw is very blobby with solder. I re-flowed it a few times with no effect but will just re-do it cleanly and see what happens.

Note: The pix show the power cable ground wire alligator clipped. Previous person had tied it in to the power ground; I removed it, clipped temporarily but have since put it on its own ground on a different transformer lug.
Here is a sound clip of the hum, plus some photos:

&feature=youtu.be

[66Strat]

Sounds like 120 hz line noise to me. The first two filter caps after the rectifier feed the 6V6 plate supply and screen grids. You might try grounding these filter caps along with the 6V6 cathodes to a common ground near the power transformer. The last filter cap feeds the PI and preamp plates. Ground this filter cap near the input.

Thanks.

6.8K to chassis: 6.62K

V4 5 to chassis: 221K
V5 5 to chassis: 224K

These look kind of wonky to me. V4 pin 5 to ground has a 220k resistor in series with the 6.8k resistor. I would expect the V4 reading to be higher than V5.

[pdf64]

Sorry if I missed mention of it, but has a known good tube been tried, especially in that V3 last gain stage / paraphase socket?
By ‘known good’, I mean one where neither section introduces hum to the signal path, when it’s  operating as an unbypassed common cathode gain stage.
eg due to excessive leakage between heater and cathode.

[pdf64]

There seemed to be a little ‘chirp’ of oscillation for a moment after power down?
Are you sure all the HT decoupling caps are good?
Does bypassing the shared V3 cathode resistor affect it?

[Diverted]

Good catch. Yes, a chirp at shutdown. As for the coupling caps, I swapped out the two final caps feeding the 6V6s. I assumed (wrongly?) that ones further upstream in the circuit were not at issue as pulling V1, V2 and grounding v3A grid had no effect on the hum.

I will bypass V3 in a moment.

For now here’s an update. Per 66Strat’s comment, I rechecked.
V4 220 to ground is 221.2 (that’s comprised of the 220K resistor which actually measures 214.8 and the 6.8K resistor which actually measures 6.4K). So that looks OK.
V5 220 to ground is actually 214.6. So there’s approx 6.6K difference.

As for grounding: I undid the power supply ground and put a new ground lug on that transformer bolt, cleaned the area up. It had a negligible effect but glad I did it anyway as it was looking pretty dodgy/oxidized.

As for grounding the various filter caps. I tried every arrangement I could on the second two, leaving the first filter cap in place at the transformer lug. This includes:

1. Second cap to power ground, third cap to preamp ground. No change.
2. Third cap to power ground, second cap to preamp ground. No change.
3. Second and third cap to power ground. No change.
4. Second and third cap to terminal strip lug where they were tied off when I received the amp. No change.
5. Second cap to power ground, third cap to preamp ground buss. No change.
6. Third cap to power ground, second cap to preamp ground buss. No change.
7. Second cap to terminal strip lug, third cap to power ground. No change.
8. Second cap to terminal strip lug, third cap to preamp ground. No change.
9. Third cap to terminal strip lug, second cap to power ground. No change.
10. Third cap to terminal strip lug, second cap to preamp ground. No change.

Tubes: Responding to PDF64, yes, at this point I have tried more than a dozen known good 12AX7s in V3. No reduction in hum.

Thanks for the patience!

[Diverted]

Update:

I bypassed V3 shared cathodes with a 22uf. Minor reduction in hum ... I left it in place for the time being.

Also, I tried paralleling a 47uf cap across the second and third filters. This had no effect on the hum.

[shooter]

I tried paralleling a 47uf cap

did you throw it at the cathode of the 6V6's to test?

how does the amp sound guitar wise?
where on the dials does the guitar volume "mask" the hum

[Diverted]

I didn’t try it on the 6V6 cathodes. Forgot to check that one. Not sure that’s it as I replaced that cathode cap with a Sprague 25uf. I will try it now.

The hum is not really noticeable past about 2-3. This amp gets loud QUICK. Other than the hum, the amp sounds fantastic.

[Diverted]

Tried upping value of 6V6 cathodes cap. No change in the hum. Eliminating many things here! It’s got to be somewhere.

[bmccowan]

It’s got to be somewhere.

Are you set up for a probe and listening amp? I have found a problematic hum by going through the signal chain with a probe.

[shooter]

It’s got to be somewhere.

just like the car keys after a bad night 

did you remove and or replace the plate to plate cap on the PI?

[Diverted]

What is that, a snubber cap? No, I have not touched that cap.

[66Strat]

The cap across the plates is there to address/prevent potential parasitic oscillation.

[Diverted]

Thank you. Disconnected one lead, tried it without the cap, and also put another in its place. No change.

[shooter]

IF you're convinced the tubes and Ecaps are good you're down to ~11parts and 3 sockets to rebuild the PI &PA.
otherwise I would rebuild the wiring that spaghetti's around the 6V6's.  re-tension and really clean the sockets, do  the 9 pinners have a "gasket" around them, can't really tell in pic. 

[Diverted]

Yes, they have those little rubber shock grommets.
I may be at that point. I’m sure that everything in the PI/output stages is good. Filter caps I’m sure are good, coupling caps, new resistors etc. Grounds are good and solid. I’ve cleaned all tube sockets, reflowed leads in suspect areas, chopsticked like mad, and neatened up the wiring where I could. .... the gremlin persists. I hate the thought of giving up and not being able to address the problem, but I don’t know if he’ll go for a full rebuild of the suspect stages. We’ll see.
Anyway, many thanks for the help.

[66Strat]

The heater wiring was messy throughout. I moved it out and tucked it up against the edge of the chassis. That cut down the hum some, but it’s still there. The heater wires are now a pretty decent distance away from sensitive leads, I think, and chop sticking throughout has no effect.
I honestly can’t find anything else wrong ... have traced out circuit many times and all is in place  ... all resistors are on spec. Coupling caps to output stage are brand new, filter caps brand new. Voltages check out vs. schematic.


I don’t know what the original grounding scheme was but I have the first (new F&T) filter on the power ground; the second and third (they are now F&Ts, the pics show the Illinois caps that came with the amp and are now gone) are on the preamp ground which runs from the first input jack, through the three others, into the pots, and via a wire down into the 12AX7 socket, then the next and the next, which have a grounding center. from there, other stuff is tied into them. Before settling on this I tried different configurations: All three on the power ground at transformer lug, first two on the power ground with third on the preamp, etc etc etc. None eliminated the hum and were all very close, hum-wise.

I’m going to reestablish the main power ground by removing, putting a new grounding lug in, cleaning chassis around/underneath it and re-bolting. The little solder lug tied to the transformer screw is very blobby with solder. I re-flowed it a few times with no effect but will just re-do it cleanly and see what happens.

Note: The pix show the power cable ground wire alligator clipped. Previous person had tied it in to the power ground; I removed it, clipped temporarily but have since put it on its own ground on a different transformer lug.
Here is a sound clip of the hum, plus some photos:

&feature=youtu.be

It's hard to tell from your description, but the input ground should be at a lower ground potential than the filter cap. If the ground is near the power transformer, the resistance of the wire could be enough to put the preamp at a higher ground potential than the filter cap. Ideally, the preamp should be grounded to a single ground point near the input jack. The preamp filter cap should be grounded at this point.

[Diverted]

This amp’s preamp ground consists of a solid 16g wire running from the first jack all the way on the other side of the chassis, then through the three other jacks, and then on into the three pots, each of which are grounded. So the preamp ground comes pretty close to the power transformer. That ground buss also has ground wires coming off of it that feed the center post grounds on V1-V3, and components (cathode resistors, other parts etc) are grounded to them.

[bmccowan]

Before you start a major rebuild, I'll suggest a couple of things:
Suggest again probing the signal path with a listening amp. Easy to make the probe and it allows you to narrow down the problem area. With that info you could end up replacing a couple of caps/resistors or a pot rather than all.
Try a 12AY7 in V1. It is not likely to cure the hum you describe when the amp volume pots are all the way off. But it may help as you turn up the volume a bit. Note that the other schematic for your amp uses AY7s for V1&2. You may also like the sound with that tube.
Just for the heck of it, try a 12AY7, 5751, or 12AT7 in the PI position.
And - I have no idea why, but one of my amps does not like one of my guitar cords - a cord that works fine with my other amps
Although I don't think its the source of hum - you have the hot 120VAC line pulled pretty tight over to the fuse holder.

[bmccowan]

And I forgot - PI tube socket - as Shooter suggested clean it real well, but I'd probably replace it - it is old!
Speaking of old, see this old thread https://el34world.com/Forum/index.php?topic=18259.50

[Diverted]

Thanks for the suggestions. I have stepped away from the work for the moment due to a family emergency, but will get back to it.
I think the pilot lamp/AC cord is not an issue here. Originally it was routed differently, with the green ground wire grounded at the main power lug. I removed it, put it on its own transformer lug and re-routed the ac line at that time. Neither configuration made any difference with the hum.
I don’t know that he would go for replacing the socket entirely. As it was, I cleaned and re-tensioned all the socket pins, cleaned/de-oxed/lightly scrubbed them to improve contact. No difference.

I don’t have a signal probe to inject a signal along the way. But I don’t know that that would help (?) as it is coming from the PI or downstream. I cleaned up the wiring around the PI and output sockets, reflowed solder joints, tried to separate wires where I could (and if I couldn’t separate them enough I reworked them so I could) etc. It’s just a phantom and I’m wondering what else I could do. All caps and resistors in the PI circuit are new ... the 220Ks, the coupling caps to the output stage, the 6V6 bypass cap, the 6.8K to ground, etc. It’s a frustrating one!