Peculiar case of the Gibson GA40 Version 1

[Yeatzee]

Gonna get straight to it. Purchased this at the vintage guitar show here in SoCal, a dream amp of mine. It was worked on by Skip Simmons so I jumped at the amp and purchased it hoping it wouldn't end up another project and I could just enjoy it. Got it home and it had some issues, eventually leading me to discover a cap that wasn't soldered. More details here:

&t

Fast forward to now and the amp seems to work... ok. The sound is cool, but it's low on output. I was expecting it to be a louder amp, say Princeton level, but it definitely does not sound 12ish watts. I thought it might be tired tubes, tired speaker, but when I started working on rectifying both of those issues I decided to try using my scope for the first real time and found some issues I don't understand. Coming with lots of questions, BUT I'm also coming with LOTS of information to hopefully help track down what's going on.

First things first, I'm attaching the schematic that I found here on this forum. God bless the guy that made it, as there does not exist to my knowledge any other V1 GA40 schematic out there. I used it to comb through the amp and check to see if every resistor / cap is present and the right value (Green = good, yellow = off enough to replace, teal = can't get a good reading). To my beginner eyes, nothing is really jumping out to me as majorly wrong to cause low output. The only thing I couldn't really confirm were the filter caps, and all of the wiring junctions like where R20/21/24/27 connect as it's hard to follow in the rats nest.

 I then also checked the voltages at every pin of the tubes and marked those. I don't know what I should expect since there aren't any voltages listed on the schem. See attached for both PDF's with my notations.

Now time for the scope.

First things first, here's the speaker output with a 1k sine wave (everything shown next is with that same tone) and the amp just before clipping.



I've seen a few different ways to calculate output, I'll let you guys tell me the best way, but either way I do it it's real low.

This is really what set me off to go down the rabbit hole, clearly something is up and I'm really hoping it's not the OT because that would be a gut punch. Continuing on, I took some photo's of the scope's results at a few key places:

Plates:




Grids (1/2 volume):



Grids (Max volume):




Some funky wavforms on the grids when cranked. I installed a new, matched set of 6v6's that have been tested in another amp to confirm they're good before doing this test. I also tried swapping the PI with a new JJ and it made no difference.

As for bias, with the amp volume off I was getting 15.74v across the 143.5ohm biasing resistor. Measured with my bias probe I got 283vdc / 48 = 97% just about dead on for both tube sockets.

Based off everything, what do you see if anything that would point to low output?

[mresistor]

So this amp has a field coil speaker?  And is the tremolo working?  Also I find this schematic to be a little more helpful for me.   YMMV   https://el34world.com/charts/Schematics/files/Gibson/Gibson_ga40_2.pdf

[sluckey]

So this amp has a field coil speaker?

No.

[shooter]

1X probe or 10X on the scope?


i use V^2 / R at the speaker

[Yeatzee]

1X probe or 10X on the scope?


i use V^2 / R at the speaker

10x on everything I took a photo of here.

[Yeatzee]

So this amp has a field coil speaker?  And is the tremolo working?  Also I find this schematic to be a little more helpful for me.   YMMV   https://el34world.com/charts/Schematics/files/Gibson/Gibson_ga40_2.pdf

No field coil. I've never seen that schematic before, thank you for sharing that! Trem works, though it needs to be dialed in but that's a later problem I think.

[Platefire]

I'm paying attention to this one because I once owned a 1960 GA40 with Trem. Thankfully mine came fully functional withthe exception of the trem and came with a Celestion Green Back.

The amp was about as loud as my Silvertone 1482 and had very little clean headroom. The amp was fine if you wantedto get into early break up. It could be a lot of things holding the amp back like bad solder joint to a failing component. Probablysomething simple---just finding it hope you find it soon so you can enjoy your amp

[shooter]

set up the scope Ch 1 on the input jack, Ch 2 speaker
put in your 1k signal, set Gain n or volume at ~~ 50%
raise the input sig volume til the speaker trace just starts to clip
 
post a screenshot including volt like you did earlier.


your waveforms all look good til it starts clipping, then it looks like the PI goes into "sagg'n boobs"

[Yeatzee]

set up the scope Ch 1 on the input jack, Ch 2 speaker
put in your 1k signal, set Gain n or volume at ~~ 50%
raise the input sig volume til the speaker trace just starts to clip
 
post a screenshot including volt like you did earlier.


your waveforms all look good til it starts clipping, then it looks like the PI goes into "sagg'n boobs"

Like this?



Smaller, clipping waveform is at the speaker output jack. Larger, non-clipping waveform is at the input jack where the 1k signal is being injected.

[shooter]

yep, 'cept I can't see what Volts per div the speaker is


both probes still 10X?

[Yeatzee]

yep, 'cept I can't see what Volts per div the speaker is


both probes still 10X?

No, 1x here for both. And both using the same volts per division. Looks like it was 50mv

[shooter]

so you have less than unity input to speaker, yet the waveform is fine.  got a spare set of tubes to roll in and re-test?

[Yeatzee]

so you have less than unity input to speaker, yet the waveform is fine.  got a spare set of tubes to roll in and re-test?

Which tubes? It's got a new set of matched JJ's that tested fine in a different amp, and a new JJ phase inverter 6SN7. I didn't notice any difference with the tungsol 6SN7 that was there originally, and I tried switching the 6v6's between tube sockets and the "saggy boobs" waveform stuck with the socket vs moved with the tube. I can steal another pair of 6v6 from a different amp, but it seems like it's probably not the tubes unless I'm missing something!

[shooter]

so, all the tubes but PA have been changed?


is this a "kit amp"?  the schematic you posted sucks, as an abstract wall hanging, I'd be fine with it 


tube are simple to eliminate, once you eliminate everything that's good, you're down to what's bad.

[Yeatzee]

so, all the tubes but PA have been changed?


is this a "kit amp"?  the schematic you posted sucks, as an abstract wall hanging, I'd be fine with it 


tube are simple to eliminate, once you eliminate everything that's good, you're down to what's bad.

See OP, it's a 1953. I didn't make the schematic, it was found on this forum that someone obviously home-made. Power tubes have been swapped, phase inverter has been swapped. Neither seemed to make a difference. I have no spare 5V4 rectifier or 6sj7 preamp tubes, but it seems unlikely any of those are to blame. I did try re-arranging the 3 preamp tubes to see if that would do anything, but no dice. Seems like we're inching towards it being the OT

[shooter]

easy enough to test since you have a scope.


label, pic, unsolder 5 wires
ohm pri CT to each pri PA wire, guessing ~~ couple hundred ohms between CT and PA wire.
ohm secondary + to -, guessing <2 ohms


IF good;


inject a 1Vac sig on speaker wires, scope each primary, relative to CT WARNING: primary side could be > 100VAC
walla

[bmccowan]

the schematic you posted sucks

I have one of these amps. When I first went through it, that schematic, drawn by Jack Hester, was the only one I could find. At first I thought the same as Shooter - high on the suckometer. But zooming in made it readable and I found that it was 100% accurate to my amp. Jack meticulously documented his own amp to create that schematic and a companion layout.  It is a sweet sounding amp at moderate volume. It does not put out the volume of other 2x6V6 amps I have. The trem is tricky to get right. I've never put it on a scope, so I can't help you there. Good luck with it and keep your eyes peeled for the later GA-40 with the 5879 Preamp which is a better daily driver.

[shooter]

I believe it accurate, but all the florescent green mixed with the red lines does a number on the brain, even without herbal help   


I'm working off the hand-drawn Gibson draftsman's special after a noon'er lunch, someone posted.

[sluckey]

Jack meticulously documented his own amp to create that schematic and a companion layout.

I don't think so. Jack copied this schematic in detail (even copied all the quotation marks) to create his neon eye burner. The original is much more pleasant to look at, especially when trying to follow a signal path.

[bmccowan]

This is the thread that I recalled (somewhat inaccurately) when I first worked on my amp. https://el34world.com/Forum/index.php?topic=12452.msg117071#msg117071
It includes Sluckey and PRR helping Jack with the amp and the schematic. The thread helped me get the tremolo working in mine. But as discussed in that thread, the tremolo circuit is a bit dodgy.

[shooter]

 
'bout spit hot coffee all over the computer!


"CCs tend to increase in value as the universe expands."
  "Who wants to draw Nyquist diagrams for a week?"

[tubeswell]

The scope shot of the 6V6 grid pin indicates gain spike distortion from the PI. Start there.  The voltages at the PI don’t look very ideal - not much room for signal, looks like you’ve only got 15V across each load resistor. What value is R22? Is it really connected to ground? (Or is that a schematic error?)


Edit (Just saw Steve's posted schematic R22 is supposed to be 1M - check yours hasn't drifted exorbitantly. Otherwise, C12 could be leaking DC)

[shooter]

agree ^^
 the PI is in it's "simplest form"  a Grid stop R should help the sags

[Yeatzee]

The scope shot of the 6V6 grid pin indicates gain spike distortion from the PI. Start there.  The voltages at the PI don’t look very ideal - not much room for signal, looks like you’ve only got 15V across each load resistor. What value is R22? Is it really connected to ground? (Or is that a schematic error?)


Edit (Just saw Steve's posted schematic R22 is supposed to be 1M - check yours hasn't drifted exorbitantly. Otherwise, C12 could be leaking DC)

Bare with me, I'm still very new to this so gain spike distortion isn't a phrase I'm familiar with. What is "ideal" for this PI? What should I be seeing across each load resistor? Are you talking about the resistor connected to the plate(s)? Happy to check them all and report back, just point me in the right direction.

R22 is 1meg, measured about 1meg on mine. It's connected from Pin 1 of the 6SN7 to Pin 1 of the 6SJ7. I'll check C12 again to see if there is any leaky DC right now and report back.

easy enough to test since you have a scope.


label, pic, unsolder 5 wires
ohm pri CT to each pri PA wire, guessing ~~ couple hundred ohms between CT and PA wire.
ohm secondary + to -, guessing <2 ohms


IF good;


inject a 1Vac sig on speaker wires, scope each primary, relative to CT WARNING: primary side could be > 100VAC
walla

Sorry, trying to follow this but not 100% sure I understand. Also not sure how to inject 1vac signal into the speaker wires..

[Yeatzee]

C12 is not leaking any dc. 43vdc one side, the other side connected to pin 1 of the 6sn7 pi has 0vdc.

[HotBluePlates]

... gain spike distortion isn't a phrase I'm familiar with. What is "ideal" for this PI? What should I be seeing across each load resistor? ...

See Valve Wizard's page for a cathodyne inverter, and read what it says near the bottom about, "Avoiding unpleasant overdrive."

label, pic, unsolder 5 wires
ohm pri CT to each pri PA wire, guessing ~~ couple hundred ohms between CT and PA wire.
ohm secondary + to -, guessing <2 ohms

IF good;

inject a 1Vac sig on speaker wires, scope each primary, relative to CT WARNING: primary side could be > 100VAC
walla

Sorry, trying to follow this but not 100% sure I understand. Also not sure how to inject 1vac signal into the speaker wires..

If you don't have a signal generator, you can do this transformer test a different way:

   - Pull the output tubes from their sockets.
   - Use clip-on jumpers leads to connect 6V6 Pin 2 (one side of the heater) to Pin 3 (plate), and on the other 6V6 connect Pin 7 (other side of heater) to Pin 3 (plate).
   - If you power up the amp now, there should be heater voltage impressed across the output transformer primary.

   - Measure AC Volts from Pin 3 of one 6V6 socket to Pin 3 of the other 6V6 socket (don't assume it will be 6.3vac exactly)
   - Measure AC Volts across the speaker terminals.

   - Turns Ratio = Volts Primary / Volts Secondary
   - Impedance Ratio = Volts Ratio x Volts Ratio
   - Primary Impedance = Speaker Impedance x Impedance Ratio

Example:  You configure the amp as described above, and verify there heater wiring is applying 6.3vac across the primary.  You measure the secondary and find 200mV AC.
   - Turns Ratio = 6.3vac Primary / 200mV AC secondary = 31.5 to 1
   - Impedance Ratio = Turns Ratio x Turns Ratio = 31.5 x 31.5 = 992.25 to 1
   - Primary Impedance = Speaker Impedance x Impedance Ratio = 8Ω x 992.25 = 7938Ω

      - Round up and assume this output transformer is "8kΩ Primary, 8Ω Secondary"
      - Rounding is fine, as there are likely several rounding & measuring errors present anyway.

[Yeatzee]

... gain spike distortion isn't a phrase I'm familiar with. What is "ideal" for this PI? What should I be seeing across each load resistor? ...

See Valve Wizard's page for a cathodyne inverter, and read what it says near the bottom about, "Avoiding unpleasant overdrive."

label, pic, unsolder 5 wires
ohm pri CT to each pri PA wire, guessing ~~ couple hundred ohms between CT and PA wire.
ohm secondary + to -, guessing <2 ohms

IF good;

inject a 1Vac sig on speaker wires, scope each primary, relative to CT WARNING: primary side could be > 100VAC
walla

Sorry, trying to follow this but not 100% sure I understand. Also not sure how to inject 1vac signal into the speaker wires..

If you don't have a signal generator, you can do this transformer test a different way:

   - Pull the output tubes from their sockets.
   - Use clip-on jumpers leads to connect 6V6 Pin 2 (one side of the heater) to Pin 3 (plate), and on the other 6V6 connect Pin 7 (other side of heater) to Pin 3 (plate).
   - If you power up the amp now, there should be heater voltage impressed across the output transformer primary.

   - Measure AC Volts from Pin 3 of one 6V6 socket to Pin 3 of the other 6V6 socket (don't assume it will be 6.3vac exactly)
   - Measure AC Volts across the speaker terminals.

   - Turns Ratio = Volts Primary / Volts Secondary
   - Impedance Ratio = Volts Ratio x Volts Ratio
   - Primary Impedance = Speaker Impedance x Impedance Ratio

Example:  You configure the amp as described above, and verify there heater wiring is applying 6.3vac across the primary.  You measure the secondary and find 200mV AC.
   - Turns Ratio = 6.3vac Primary / 200mV AC secondary = 31.5 to 1
   - Impedance Ratio = Turns Ratio x Turns Ratio = 31.5 x 31.5 = 992.25 to 1
   - Primary Impedance = Speaker Impedance x Impedance Ratio = 8Ω x 992.25 = 7938Ω

      - Round up and assume this output transformer is "8kΩ Primary, 8Ω Secondary"
      - Rounding is fine, as there are likely several rounding & measuring errors present anyway.

Hey HBP,

So try swapping R25 and R26, both 470k, with something even bigger like 1meg? "the spike can be avoided by using unusually large grid stoppers on the power valves, around 100k say." The tone of the amp doesn't bother me, it's got a cool sound, its the low output that's worrying me more than anything. I'm down to experiment and see what this does to the tone though, if I understand correctly.

I do have a signal generator, but I don't know what it means to inject 1vac on speaker wires. I can give this a shot and report back in a sec.

[HotBluePlates]

I do have a signal generator, but I don't know what it means to inject 1vac on speaker wires.

Connect Signal Generator + to Speaker +
Connect Signal Generator - to Speaker -
Turn on Signal Generator
Adjust Signal Generator output until you measure 1vac from Speaker + to Speaker -

Measure AC Volts on the Primary. ----> Danger:  1v x 31.5 = 31.5 volts across the primary

   The point is the Volts at the Primary will be levered-up by the Turns Ratio, giving a voltage that most meters will measure easily & accurately.

[Yeatzee]

I do have a signal generator, but I don't know what it means to inject 1vac on speaker wires.

Connect Signal Generator + to Speaker +
Connect Signal Generator - to Speaker -
Turn on Signal Generator
Adjust Signal Generator output until you measure 1vac from Speaker + to Speaker -

Measure AC Volts on the Primary. ----> Danger:  1v x 31.5 = 31.5 volts across the primary

   The point is the Volts at the Primary will be levered-up by the Turns Ratio, giving a voltage that most meters will measure easily & accurately.

Yeah I'm totally lost, not getting results that make any sense.



White lead is connected to the COM of the meter which is on the ground tab of the speaker jack where the - of the signal generator is. If I connect the + from the signal generator to the + of the meter I get 1vac as you can see. If I connect the + of the signal generator to the + of the speaker jack and measure with the meter across the jack I can't get anywhere near 1vac with the signal generator at it's max volume (about 10vac if I connect it like I show in the photo). What's going on there? Why's that? I tried adding the OX speaker load to the jack and yeah there's no getting 1vac from speaker + to speaker -.

[shooter]

Pull the output tubes from their sockets.

[Yeatzee]

If you don't have a signal generator, you can do this transformer test a different way:

   - Pull the output tubes from their sockets.
   - Use clip-on jumpers leads to connect 6V6 Pin 2 (one side of the heater) to Pin 3 (plate), and on the other 6V6 connect Pin 7 (other side of heater) to Pin 3 (plate).
   - If you power up the amp now, there should be heater voltage impressed across the output transformer primary.

   - Measure AC Volts from Pin 3 of one 6V6 socket to Pin 3 of the other 6V6 socket (don't assume it will be 6.3vac exactly)
   - Measure AC Volts across the speaker terminals.

   - Turns Ratio = Volts Primary / Volts Secondary
   - Impedance Ratio = Volts Ratio x Volts Ratio
   - Primary Impedance = Speaker Impedance x Impedance Ratio

Example:  You configure the amp as described above, and verify there heater wiring is applying 6.3vac across the primary.  You measure the secondary and find 200mV AC.
   - Turns Ratio = 6.3vac Primary / 200mV AC secondary = 31.5 to 1
   - Impedance Ratio = Turns Ratio x Turns Ratio = 31.5 x 31.5 = 992.25 to 1
   - Primary Impedance = Speaker Impedance x Impedance Ratio = 8Ω x 992.25 = 7938Ω

      - Round up and assume this output transformer is "8kΩ Primary, 8Ω Secondary"
      - Rounding is fine, as there are likely several rounding & measuring errors present anyway.

And to clarify how to do this test, because I can interpret it two different ways...

You're saying jumper pin 2 of the socket with NO 6v6 in it to Pin 3 of the same socket, and pin 7 to pin 3 of the other 6v6 socket with again no tube in it?

Heaters measure pretty spot on.

[HotBluePlates]

Yeah I'm totally lost, not getting results that make any sense.

If you're gonna do it Shooter's way (Signal Generator at speaker), then the speaker needs to be attached, and the amp must not be powered on.  Even then, the low resistance of the secondary winding will tend to limit the signal generator's output.

If you're gonna do it my way, the output tubes must be pulled from their sockets.  Heater pins of the output tubes get connected to plate pins, and the amp powered on for the test.  You're looking to find 200mV or a little less at the speaker terminals (speaker may be connected, or not).  Be sure to remove these connections before re-installing the 6V6s.

[Yeatzee]

Yeah I'm totally lost, not getting results that make any sense.

If you're gonna do it Shooter's way (Signal Generator at speaker), then the speaker needs to be attached, and the amp must not be powered on.  Even then, the low resistance of the secondary winding will tend to limit the signal generator's output.

If you're gonna do it my way, the output tubes must be pulled from their sockets.  Heater pins of the output tubes get connected to plate pins, and the amp powered on for the test.  You're looking to find 200mV or a little less at the speaker terminals (speaker may be connected, or not).  Be sure to remove these connections before re-installing the 6V6s.

It appears my signal generator cannot generate a loud enough signal to get to 1vac hooked up I guess.

Regarding your method, can you clarify the connections. You're saying have the 6v6 on the work bench with jumpers, but I'm not following how they're connected to the amp the description leaves me uncertain and I don't want to mess anything up.

[tubeswell]

Since it seems to be a ‘low volume’ issue, and seeing as how C12 isn’t leaking DC, have you test the other coupling caps for DC leakage?

[HotBluePlates]

...
If you're gonna do it my way ...  Heater pins of the output tubes get connected to plate pins, and the amp powered on for the test. ...

Pick a 6V6 socket.  Jumper Pin 7 to Pin 3 (preferably with a clip-on jumper wire).

Look at the other 6V6 socket.  Jumper Pin 2 to Pin 3 (preferably with a clip-on jumper wire).

There should be about 6.3vac measured between Pin 2 and Pin 7 of the 6V6 sockets.
Since you've attached only one of these to the plate pin of the same socket, you have now placed 6.3vac across the output transformer primary.
Do not have the 6V6s in their sockets during this test!




Said another way, you are attaching the heater winding across the output transformer primary, to inject 6.3vac at the primary so that you can measure some small AC voltage at the secondary.

[pdf64]

...
Pick a 6V6 socket.  Jumper Pin 7 to Pin 3 (preferably with a clip-on jumper wire).

Look at the other 6V6 socket.  Jumper Pin 2 to Pin 3 (preferably with a clip-on jumper wire).

There should be about 6.3vac measured between Pin 2 and Pin 7 of the 6V6 sockets.
Since you've attached only one of these to the plate pin of the same socket, you have now placed 6.3vac across the output transformer primary.
Do not have the 6V6s in their sockets during this test! ...

Surely the rectifier valve (5Y3) needs removing too?
Otherwise the OT primary will be up at HT voltage, whilst the 6.3V heater winding is referenced to 0V common, and bad things will happen.

[Yeatzee]

...
Pick a 6V6 socket.  Jumper Pin 7 to Pin 3 (preferably with a clip-on jumper wire).

Look at the other 6V6 socket.  Jumper Pin 2 to Pin 3 (preferably with a clip-on jumper wire).

There should be about 6.3vac measured between Pin 2 and Pin 7 of the 6V6 sockets.
Since you've attached only one of these to the plate pin of the same socket, you have now placed 6.3vac across the output transformer primary.
Do not have the 6V6s in their sockets during this test! ...

Surely the rectifier valve (5Y3) needs removing too?
Otherwise the OT primary will be up at HT voltage, whilst the 6.3V heater winding is referenced to 0V common, and bad things will happen.

5V4 in this amp. I can remove the rectifier and try it if that's the move.. obviously want to avoid "bad things" from happening!