Gibson GA20 HUM!!!!!!!!! UPDATE!

[plexi50]

I have a very bad and loud hum in a Gibson GA20 amp. I have replaced the power supply capacitors and PI coupling capacitors. You can pull V1 & V2 12AY7 preamp tubes and the hum is still there. This means the problem is in V3 the PI. But, after making sure the pins are clean and the wires are re-soldered it persists

The hum does not effect the Microphone channel input. It is clear and nice. No hum.
The hum increases in loudness in the instruments channel volume pot rotation.
I changed the coupling cap for the instrument channel V2 with no change in hum.
All resistor values in the amp are within there values. What in haids is causing this?
It's crazy. Im missing something. The tubes are excellent.

[shooter]

The hum does not effect the Microphone channel input. It is clear and nice

I'm going by the schematic in your other thread, if this is correct, swap parts til you find it.  they look close enough to identical.  Otherwise bust out a scope and compare.

[plexi50]

The hum does not effect the Microphone channel input. It is clear and nice

I'm going by the schematic in your other thread, if this is correct, swap parts til you find it.  they look close enough to identical.  Otherwise bust out a scope and compare.

Their are only (5) original coupling caps left and im sure i will wind up changing them as well because of it's sheer age and leakage.
The Hum sounds like it is being injected straight into the grids of the power tubes. But the PI has new coupling caps now.

Yes this is the same schematic as in the other post of the Gibson GA20 Schematic.
 

[shooter]

unless I missed something, all the parts I Didn't circle work just fine when you plug in the mic?

so that just leaves the things I circled?

[sluckey]

Is this a true statement?

You can pull V1 & V2 12AY7 preamp tubes and the hum is still there.

If so, then the hum ain't coming from the preamp.

I suspect this statement is a red herring? At least it needs clarifying because it conflicts with the above statement.

The hum does not effect the Microphone channel input. It is clear and nice. No hum.

[plexi50]

Yes if i pull V1 & V2 the loud hum still persists in the instrument channel.

The Microphone channel is not effected, (super weird). That's why i believe the PI is the issue.

I found an old post from 2015 here that looks like the very same issue i am having.

http://el34world.com/Forum/index.php?topic=18259.0

The problem turned out to be the PI tube socket itself. Though i would have like to seen more about the tube socket itself and what the actual issue with it was.

[plexi50]

unless I missed something, all the parts I Didn't circle work just fine when you plug in the mic?

so that just leaves the things I circled?

Yes that would be correct.
I changed V1A & V2A coupling cap's earlier going to the microphone & instrument volume pot wipers.
There has to be something going on with the PI socket area.

[plexi50]

Pic's of new caps so far. Mica cap was removed and replaced. Hum not effected by removal of Mica cap. (snubber)
I also changed the bias r from 200r to 250r to see if the bias was to hot and creating a hum. No change.

[pdf64]

...The hum does not effect the Microphone channel input. It is clear and nice. No hum.
The hum increases in loudness in the instruments channel volume pot rotation...

So does turning the instrument vol control to minimum kill the hum?

[sluckey]

Connect the hot side of a volume control to a good ground point. Does the hum go away?

[shooter]

what happens if you disconnect C8/R9

[plexi50]

...The hum does not effect the Microphone channel input. It is clear and nice. No hum.
The hum increases in loudness in the instruments channel volume pot rotation...

So does turning the instrument vol control to minimum kill the hum?

Increasing the instrument volume pot increases the volume to an insane level.
Returning the pot to CCW position is minumal but present. Still very loud.

[plexi50]

Connect the hot side of a volume control to a good ground point. Does the hum go away?

No it does nothing to mute the hum.

[plexi50]

what happens if you disconnect C8/R9

Disconnected C8 tone cap. No change.

[plexi50]

If i ground pin 7 of the PI it kill's 95% of the hum. Grounding pin 2 of the PI has no effect.

[terminalgs]

If i ground pin 7 of the PI it kill's 95% of the hum. Grounding pin 2 of the PI has no effect.

If pin 7 comes from preamp and grounding it kill the hum, the hum isn't in the PI.  It's between pin7 and the preamp.

[plexi50]

If i ground pin 7 of the PI it kill's 95% of the hum. Grounding pin 2 of the PI has no effect.

If pin 7 comes from preamp and grounding it kill the hum, the hum isn't in the PI.  It's between pin7 and the preamp.

Pin 7 of PI feeds into 220k / 4.7k.  Pin 2 feeds microphone & instrument volume pot's. From instrument pot to tone cap to tone pot

[sluckey]

Measure resistance directly between the PI pin 7 and ground. What have you?

Look at that solder blob on the terminal strip where the yellow wire from pin 7 meets the 4.7K. Could be bad. Clean it up. Also look at the other end of that 4.7K. It is connected to ground. Does it look good?  Recheck resistance directly between PI pin 7 and ground.

[plexi50]

Measure resistance directly between the PI pin 7 and ground. What have you?

Look at that solder blob on the terminal strip where the yellow wire from pin 7 meets the 4.7K. Could be bad. Clean it up. Also look at the other end of that 4.7K. It is connected to ground. Does it look good?  Recheck resistance directly between PI pin 7 and ground.

Resistance from pin 7 to ground is 4.9k.  Looking at the blob now.

[plexi50]

Resistance from pin 7 to blob is 4.9k. From blob (4.7k / 220k junction) resistors read 4.9k / 215k across their leads.

[shooter]

Resistance from pin 7

which side of Pin 7, tube or bottom?

[plexi50]

Resistance from pin 7

which side of Pin 7, tube or bottom?

Good question!
From the bottom. Checking the inside of the socket now.

[plexi50]

Measured from inside V3 PI tube socket pins all show connection to their prospective point's though resistors and wire leads.
It's crazy. All resistors and paths (leads) are making good connection. I'm going to look at the heater filament wiring again. It's pretty straight forward though.

[plexi50]

Here is a short video of my issue. I just posted it.

&feature=youtu.be

[sluckey]

1. Pull the PI tube. Does the hum go away?

2. Do you have a meter that can measure frequency? Such as Fluke 87V? If so connect the meter across the speaker leads and measure the hum frequency. What have you?

[plexi50]

1. Pull the PI tube. Does the hum go away?

2. Do you have a meter that can measure frequency? Such as Fluke 87V? If so connect the meter across the speaker leads and measure the hum frequency. What have you?
[/quot

Hum is gone after pulling the PI tube. No frequency meter but maybe some crazy app on my phone will measure the field.

[plexi50]

Audio Frequency Counter with phone app shows 240.0085 HZ.
Phone put near amp o.oooo before turning on amp.
The frequency settles at 240.0085 HZ. after speaker starts humming.

[shooter]

for fun, can you take out c13?  100Pf? bridging the plates

[plexi50]

Shooter Do you mean C17 the one across the plates of the PI?  Can barely see if that is C17 on the smudged schematic.
The Mica snubber cap? Did that a while ago. No change.
THIS IS FUN ISN'T IT!

[shooter]

Yup, IT IS, if it's NOT, take up yoga
I've spent the day on my sand project
I'd be glad to swap troubleshooting

if this is a keeper, or a profit maker, keep swappin, eventually the gremlin can no longer hide

[plexi50]

Do you all hear 120 cycle hum here?

[DummyLoad]

Do you all hear 120 cycle hum here?

yes is 120Hz buzz.

where is R17? i see C15 plate filter and C14 screen filter with yel/red wire back to screens for 6V6 but DO NOT see R17 (10K) that should connect from C14+ to C13+ which is the 10uF filter for PI and preamp tubes. C13,C14,C15 used to be the can cap. R18 is visible, R17 is not - at least not visible in the pics provided.


--pete

[pdf64]

Thanks for the video; yes, to me that's hum from the HT line, not the heaters.
Are you sure all the HT caps are good, and have a good connection to the chassis 0V?.
Mine had a chromed chassis; your's is painted, so electrical connections to the chassis might be a bit more problematic.

[plexi50]

Do you all hear 120 cycle hum here?

yes is 120Hz buzz.

where is R17? i see C15 plate filter and C14 screen filter with yel/red wire back to screens for 6V6 but DO NOT see R17 (10K) that should connect from C14+ to C13+ which is the 10uF filter for PI and preamp tubes. C13,C14,C15 used to be the can cap. R18 is visible, R17 is not - at least not visible in the pics provided.


--pete

R17 is well hidden behind the back of the terminal strip behind the lower coupling cap. Even with the lower coupling cap removed you still can not easily see it without a mirror and flash light. And then use DVM to confirm it is actually there in place. It took me 3 days to actually find it. It is that well hidden.

[plexi50]

Thanks for the video; yes, to me that's hum from the HT line, not the heaters.
Are you sure all the HT caps are good, and have a good connection to the chassis 0V?.
Mine had a chromed chassis; your's is painted, so electrical connections to the chassis might be a bit more problematic.

The PS caps are new.  I also paralleled  a new 100uf / 350VDC cap across the new PS caps to test each new cap.
Upon first contact the hum is deadened 50% and then immediately returns. This is nut's. Could the transformer itself be bad?

[sluckey]

Find where the HT center tap wire connects to chassis. Now use a gator clip lead to connect the negative lead of a filter cap to the same point the HT center tap is connected to. Repeat for all filter caps you have replaced. Schematic shows three filter caps so you need three gator clip leads. Make sure you have a good connection with all six gator clips. Fire it up. Hum any better?
 

[plexi50]

Find where the HT center tap wire connects to chassis. Now use a gator clip lead to connect the negative lead of a filter cap to the same point the HT center tap is connected to. Repeat for all filter caps you have replaced. Schematic shows three filter caps so you need three gator clip leads. Make sure you have a good connection with all six gator clips. Fire it up. Hum any better?

The high voltage CT is soldered to the Pin #1 of the first 6V6G tube along with a small white wire which is ground for the entire preamp including the bias 220k voltage r feeding pin 5 grid. I'm taking off the CT from the first 6V6 and moving it to the chassis with a alligator clip wire to see what happens.

[plexi50]

Victory!!!!    There is no way that HV CT was ever soldered at the factory to pin#1 passing through the 220K grid resistor to pin #5.
I just noticed the OT plate leads look soldered funny on the 6V6 tubes. Not clean work or original.

[sluckey]

There is no way that HV CT was ever soldered at the factory to pin#1 passing through the 220K grid resistor to pin #5.

Sure it was. Pin 1 was often connected to ground back in those days. Heck, the old metal 6V6s had the metal shell tied to ground. Look at this pic...



My point is that pin 1 was a ground tie point. It was convenient to connect the center tap to that ground point. It was also convenient to connect the 220K grid resistor to that same ground point. So quit thinking about the HV CT was passing through the 220K grid resistor to pin #5. That's ass backwards!

Gibson did not rely on that painted chassis to provide a good ground. That's why they strung a ground wire around the circuits. The real problem occurred when you replaced the caps but began connecting them directly to chassis (which ain't a good ground in this amp). That's why I wanted you to use gator clip leads to connect the negative end of each filter cap directly to the HV CT.

There are several ways to permanently fix this problem. Personally, I'd connect the HV CT back to pin 1. Then I'd run a wire from the negative lead of each filter cap you replaced back to the CT, or that bare wire on the pots. However you decide to do it, when all is done make sure that 220K resistor between pins 5 and 1 still measures zero ohms to the CT. That CT is like the negative terminal of a car battery. It is the ultimate ground point and everything on the schematic that should be connected to ground better have a good connection back to the CT.

Edit and fixed... Don't know why I kept saying pin 6 when I meant pin 1.

[plexi50]

There is no way that HV CT was ever soldered at the factory to pin#1 passing through the 220K grid resistor to pin #5.

Sure it was. Pin 6 was often connected to ground back in those days. Heck, the old metal 6V6s had the metal shell tied to ground. Look at this pic...



My point is that pin 6 was a ground tie point. It was convenient to connect the center tap to that ground point. It was also convenient to connect the 220K grid resistor to that same ground point. So quit thinking about the HV CT was passing through the 220K grid resistor to pin #5. That's ass backwards!

Gibson did not rely on that painted chassis to provide a good ground. That's why they strung a ground wire around the circuits. The real problem occurred when you replaced the caps but began connecting them directly to chassis (which ain't a good ground in this amp). That's why I wanted you to use gator clip leads to connect the negative end of each filter cap directly to the HV CT.

There are several ways to permanently fix this problem. Personally, I'd connect the HV CT back to pin 1. Then I'd run a wire from the negative lead of each filter cap you replaced back to the CT, or that bare wire on the pots. However you decide to do it, when all is done make sure that 220K resistor between pins 5 and 1 still measures zero ohms to the CT. That CT is like the negative terminal of a car battery. It is the ultimate ground point and everything on the schematic that should be connected to ground better have a good connection back to the CT.

Ah! I see what you were trying to get me to do and why.
Yes the grounding to chassis here is painted and basically just depending on the few rivets if that for a ground.
I twisted the new PS caps negative fairly good before i soldered them in place so i don't know if i should take them off and connect them to the CT back to pin 1. But i get it now. Makes me want to try it anyway and see (which i believe you are right) what happens.

[DummyLoad]

cool! it's fixted.   


--pete

[plexi50]

cool! it's fixted.   


--pete

It's fixted and sounds stellar! These amps produce a very articulate tone and the notes run on for days.

[Willabe]

Nice work!   

Plexi, could you give us a list of dc plate voltages while you still have that amp.

(For guys who might want to build a clone.)

[plexi50]

Nice work!   

Plexi, could you give us a list of dc plate voltages while you still have that amp.

(For guys who might want to build a clone.)

The amp is gone and i don't want to give numbers that i am not sure i remember.
But i can get the amp back in a week or two and document all voltages for sure and post them. I am going to build a clone of it. 

[Willabe]

Thanks Plexi.   

[plexi50]

The amp sold for 3,500.00 the day after it left here. It was a fluke sale. The owner was playing his amp when a friend came over. The guy said he wanted it. He laid down $1,000.00 / The owner said no. $2,000.00 and again no. $3,000.00 and a no again. $3,500.00 and he had to take it. Dam! I'll get another one in here soon from the same guy. There one for sale here on CL in rough shape so i should be seeing it soon. What a crazy history.

[Willabe]

WOW!   

Will you be getting the same amp model?

And document the DCV's?

[plexi50]

WOW!   

Will you be getting the same amp model?

And document the DCV's?

Dam right. Same model and year. 9 Pin preamp. Give a year here or there. I'm going to go out on a limb cause i better mention a ball park in voltages before i loose all memory.
This is what i remember of the voltages.

Plate Voltages I Remember:
V1A 55VDC / V1B 75VDC
V2A 40VDC / V2B 42VDC
V3A 100VDC / V3B 100VDC

6V6GT Tubes:
V4 & V5 = 333VDC /(343?)

5Y3 B+ = 350VDC

Bias grid with 250 ohm r was +12.93

Will see how good or bad my memory is in a few weeks after my friend brings another one to me.

[shooter]

It was a fluke sale

It was a Gibson 

I sold my GA5T for more money than the suggested retail price  , I don't feel guilty
was very clean though

[DummyLoad]

https://austin.craigslist.org/msg/d/1950s-gibson-ga20-combo-tube/6541731139.html


1100? meh!


--pete