Rebuilt Gibson GA-8T Loud Hum at 120hz.

[One10]

If all circuit points showing the ground symbol measure ~zero Ohm to chassis, they are effectively grounded .
There needs to be only one connection to chassis and that seems to be at the input.
Actually a single chassis ground is best as this avoids ground currents in the chassis, so it doesn't matter where the chassis is grounded.
Any additional chassis ground will cause a ground loop and is likely to make things worse.

I don't trust acoustic hum level measurements with a phone.

Balance the power tubes and see/hear if this silences the amp.

Hello Helmholtz, that is a relief to hear. I was beginning to think maybe I should separate the reservoir cap from the filter caps and ground it separately, but as I believe you already said, that shouldn't have been causing the 120hz. hum. Yes, I would have to agree with you, the phone app was just a way of informally checking out what frequencies I thought I was hearing come from the amp. Unfortunately, I no longer have my oscilloscope, or my other equipment to use. Just an LCR meter and a Fluke Multimeter.

The power tubes are balanced / matched. Both are same brand, same readings on a Sencore TC109. both are 106.5 & 107 Em. I use it only for matching, grid leakage & shorts.

I did not notice the (extra?) 'thin' nuts & lockwashers inbetween the chassis and powertransformer and they were loose, but the chassis nuts were tight. So I tightened them.and soon after the 120hz & 240 hz. appeared. THEN, moments later while testing the amp, alll the notes suddenly had distortion on top of them (clear note & a separate fuzz sound at the same time). Then, the low 'E' string buzzed, followed quickly with the low 'A' string warbling!

I checked every single part on the amp again. I was going bug you guys about this new problem, but then I thought, I had never taken the speaker out of the amp and actually looked at it & physically pushed on the cone to see its condition. The speaker must have went bad moments earlier. The voice coil scrapes and catches horribly. The cone cannot move back & forth smoothly. I just ordered a new replacement speaker.

I am hoping somehow, the speaker may have caused some of the hum, or no?


Sorry about the .. posts above, I was trying to revise them when I kept encountering more problems with the amp. I didn't know I couldn't put them back (the posts) later.

[g1]

Yes pin 8 is 45v (48v on schem); while pin 1 is 25.5v, not 48v as written in schematic, as it has a couple of resistors before it. I can see where the voltage drops and it is because of the resistors. I did actual measurements before a certain point and after. 25.5v is on pin 2.

The grid pin will not always give an accurate reading with respect to ground (due to meter loading effect). The cathode will. And the grid with respect to cathode will give you the difference (red probe to grid, black probe to cathode). A triode like this will only have a couple volts between grid and cathode, it would not run with 20 some volts between them. If you measure from grid to cathode, and subtract that from the cathode voltage, you should get something near what is shown on the schematic.

Sorry, I must have must the pic of the cardboard insulator for the cap can. Or is it a paper cylindrical type? That would be a multi-cap but not usually be called a 'can'. I just saw you had mentioned multi-cap can was replaced with separate caps and I assumed the can was the metal type. Here's a pic of the usual can type. All the tabs around the outside are negative terminal and connected to the shell. It's fairly common to see them with a ground wire to one of those tabs, and the other tabs soldered to the chassis. This holds the cap can in place and also connects ground to chassis. Or sometimes they have a metal clamp or strap around them that is bolted or riveted to the chassis, again completing the ground to chassis connection.
This one is a triple-cap can:

[The Dude]

I believe the original cap was a paper multi-section "firecracker" type cap. See the picture in post #41.

[Enzo]

Agree with g1. Cannot measure voltage to ground at phase inverter grid, dur=e to meter loading. Measure cathode voltage, then measure cathode to grid directly. You will find grid voltage is only a bolt or two different.

[g1]

Now realizing the original filter multi-cap was insulated, and seeing the mounting of the Kemet's, is it possible one of those Kemet's was scratched or worn on the outer plastic and making contact with the chassis? I think the Sprague's have a thicker outer plastic and that's just a wild guess as to why the Kemet's had hum where the Sprague's did not.

[SoulFetish]

So…. are we back to thinking it’s grounding again? When did that happen? I thought the added capacitance eliminated the hum?

[Helmholtz]

I was beginning to think maybe I should separate the reservoir cap from the filter caps and ground it separately, but as I believe you already said, that shouldn't have been causing the 120hz. hum.

No, that's not what I said.
I said, if adding 10µ to filter caps stops the hum, a ground problem is unlikely.
This assumed that the additional caps were connected to the same (ground) points as the existing ones.

Without a clear picture of the complete grounding scheme of the amp, I can't advise on improvements.

A choke in place of the 470R dropper wouild result in much lower screen and preamp ripple.

The power tubes are balanced / matched. Both are same brand, same readings on a Sencore TC109. both are 106.5 & 107 Em. I use it only for matching, grid leakage & shorts.

Did you check idle currents in the amp for balance?

I am hoping somehow, the speaker may have caused some of the hum, or no?

Very unlikely.

[Helmholtz]

Another remark:

Doubling filter capacitance can be expected to drop the ripple voltage at this node by 6dB (assuming similarly low ESRs of both caps).

So changes in hum level by 40dB (corresponding to a factor of 100) or more by increased capacitance are unrealistic.

A phone just isn't a sound level meter and even with professional equipment such acoustic measurements at low levels are very tricky.

[Enzo]

My land line phone has the typical response, something like 300-3000Hz. I was working on some amp and had 100Hz screaming through it for test. The person on the other end of the phone line couldn't hear it at all.

[One10]

The grid pin will not always give an accurate reading with respect to ground (due to meter loading effect). The cathode will. And the grid with respect to cathode will give you the difference (red probe to grid, black probe to cathode). A triode like this will only have a couple volts between grid and cathode, it would not run with 20 some volts between them. If you measure from grid to cathode, and subtract that from the cathode voltage, you should get something near what is shown on the schematic.

Agree with g1. Cannot measure voltage to ground at phase inverter grid, dur=e to meter loading. Measure cathode voltage, then measure cathode to grid directly. You will find grid voltage is only a bolt or two different.

Ok.

6BM8, V2A measurements:

1.) Pin 8 = 40.480 vdc.
2.) Pin 1 to pin 8 = -1.195 vdc.

A choke in place of the 470R dropper wouild result in much lower screen and preamp ripple.

Did you check idle currents in the amp for balance?

Ok, thanks. I will most likely end up using a choke after I figure out where the hum is coming from.
Yes, the idle currents were fine.

I ended up purchasing one more time, newer "Kemet" capcitors and replacing them all. These had remarkably close, to almost exact readings as the first two sets. Amp still hums a good bit from the start, at '0' and gets quite loud when turned up pass '5' on the volume pot. So again, I removed these and I put back in Sprague Atoms from 1990's (N.O.S.) and even 2 that were 'bubbling', or 'blistering' on the end and these were still 10db to 18 db quieter than the brand new Kemets. Quite puzzling why this is, It shouldn't be. As I have no degree and only very simple knowledge of guitar amps, it will have to remain a mystery to me.

I really wanted to use the Kemets, extremely long life and quite small, (because they all fit inside the amp easily).

(I would like to see how the Sprague Atoms & Kemets measure at under actual working voltages & conditions, but unfortunately don't have the $$$ or knowledge for that).

I am just going to purchase new Spraque or F&T capacitors in a week or two.

Sorry, that I couldn't respond right away.
Thanks for everyone's helping out and great knowledge!

[Helmholtz]

I ended up purchasing one more time, newer "Kemet" capcitors and replacing them all. These had remarkably close, to almost exact readings as the first two sets. Amp still hums a good bit from the start, at '0' and gets quite loud when turned up pass '5' on the volume pot. So again, I removed these and I put back in Sprague Atoms from 1990's (N.O.S.) and even 2 that were 'bubbling', or 'blistering' on the end and these were still 10db to 18 db quieter than the brand new Kemets. Quite puzzling why this is, It shouldn't be. As I have no degree and only very simple knowledge of guitar amps, it will have to remain a mystery to me.

Did you compare ripple voltages with different caps?

[g1]

Also, I'd be interested in the amps performance with a low frequency tone (60 to 100Hz range) with the old and new caps.
As the hum seems to be when you open up the volume control, maybe with the new caps you are hearing better bass response. Other than that, it's a real head-scratcher.

[One10]

Did you compare ripple voltages with different caps?

Hello Helmholtz. Yes, I tried with a multimeter, I do not have an oscilloscope right now and had to use a Fluke 89IV multimeter. The only 'good' capacitors I have at the moment are a little too large in capacity, but I connected them just for a test. Multimeter automatically displayed the "Hz".

The readings seem suspect.

Kemet Peg124 Capacitor ACV Ripple

1st (reservoir): 43.4µf 450v = 8.16V @ 120hz. Then replaced that with (1) Sprague Atom (circa 1993): 21µf 500v = 3.34V @ 120hz.

2nd: 44.37µf 450v = 510mV @ 120hz.

3rd: 34.27µf 450v = 3mV @ 0hz. (V1B pin 1)
4th: 33µf 450v = 670mV @ 60hz. (V1A pin 6)

Also, I'd be interested in the amps performance with a low frequency tone (60 to 100Hz range) with the old and new caps.
As the hum seems to be when you open up the volume control, maybe with the new caps you are hearing better bass response. Other than that, it's a real head-scratcher.

I will end up doing that later also, when I can access the rest of my test equipment.



I will have to set aside this amp aside for now.
Thanks for your help!

[One10]

As the hum seems to be when you open up the volume control, maybe with the new caps you are hearing better bass response. Other than that, it's a real head-scratcher.

Hello g1, I do get a much better bass repsonse, it isn't muddy, or loose any more and the tone knob can now be used from '0' all the way to '10' as I put in a 500pf cap in place of the .005µf on the tone pot. The bass has an almost piano-like sound, but with some very slight distortion, or should I say harmonics starting at '2'. I don't particularly like that part, but it does sound pleasing to my Dad.

I have recently liked a rather clean sound with a little sparkle or chime to it. The notes really flow and sustain for quite a long time, it is a very touch sensitive amp.



I hope to recover some equipment, including an oscilloscope in December and then I can maybe find the source of the hum.

Thanks for the reply!