Adding Feedback & A Presence Control To A Gibson GA-8T (Cathodyne Phase Inverter model)

[One10]

Good catch. I couldn't really read that bypass cap value on the schem I posted and thought maybe it said "C2". But now I see in the proposed mod image that the actual value of .02 is indicated. And I agree that it chould be full bypass on the v1B cathode

EDIT: I just ran a simulation with a .02 cap for the V1B bypass and Helmholtz is correct that it takes all the punch out of the presence control. If the overall tone becomes too bassy with full bypass then a 1uf bypass cap seems sufficient to get more gratification from the presence control while still trimming some LF gain.

Hmmm...I will probably have to remove the 3-way cathode bypass cap switch completely for now, then.
I just removed the 10k-L pot assembly 'presence' circuit.

Ok, order a 22µf & try it out as Hemholtz & you have explained, if to 'heavy' sounding then maybe a 1 µf (as you suggested), otherwise for now, all I have is a 2.2 µf cap & 5 µf cap on hand.

Thanks for all the help!

[Chuck H]

I just ran some more simulatons with the "corrected" stock 22n cathode bypass for V1B...

To get a similar frequency to the stock amp while incorporating NFB and the presence control you'd want to increase the V1B cathode bypass to 100n. This value allows for some gratification with the presence control. For more midrange keranggg! you can increase the presence cap value to 680n. And I actually like the "look" of this 100n V1B cathode bypass cap/680n presence cap arrangement. When I say I like the look what I mean is that I've teted and simulated many amps and there are certain frequency response markers that seem "right" for what we want out of our guitar amps. If it were my project I would probably roll with the 100n V1B cathode bypass and the 470n/680n presence cap. Staying with the 500 ohm presence pot and the 47k feedback series resistor.

I know you've probably already ordered some parts as I continue to experiment with the circuit. So apologies for any inconvenience but this is where I am right now

[One10]

I just ran some more simulatons with the "corrected" stock 22n cathode bypass for V1B...

To get a similar frequency to the stock amp while incorporating NFB and the presence control you'd want to increase the V1B cathode bypass to 100n. This value allows for some gratification with the presence control. For more midrange keranggg! you can increase the presence cap value to 680n. And I actually like the "look" of this 100n V1B cathode bypass cap/680n presence cap arrangement. When I say I like the look what I mean is that I've teted and simulated many amps and there are certain frequency response markers that seem "right" for what we want out of our guitar amps. If it were my project I would probably roll with the 100n V1B cathode bypass and the 470n/680n presence cap. Staying with the 500 ohm presence pot and the 47k feedback series resistor.

I know you've probably already ordered some parts as I continue to experiment with the circuit. So apologies for any inconvenience but this is where I am right now

it's ok, I haven't ordered parts yet, the shipping (90% of the time) is expensive for Alaska, I am still looking at a few other parts I can add to the order to lessen the shipping $$$ shock.

1.) Chuck H, that's fantastic that you can simulate the amp circuit in a program, it does seem to take the much, or all the guesswork out of the equation.

2.) Does the potentiomer have to absolutely be 500Ω? I have misunderstood as to how the pot interacts, or not with the 47k feedback resistor. I had thought the variable pot. didn't influence the negative feedback greatly from the 'speaker' output, except for negating certain freq. from the negative signal that is feeding back to V1B Cathode.
By this I mean for i.e. mid-high end freq. being 'added' back into the circuit depending on capacitor & resistor value combo you mentioned: 470nF, or 680nF & 500Ω var. pot., etc.
(I am asking because so far It has been easier & cheaper to find online correct size 200Ω, even100Ω var. pots that will fit in my amp, then 500Ω pots. that are instock on EBay.

3.) Hello Hemholtz, I just tried a 22µf @ 450v at V1B + 2.2k bypass and the low end was quite thiick & loud (from vol. 3 and up), but to me it sounds a little indistinct, like a smearing of the low freq. from 82Hz to about 200hz, slightly muddy? even. (I was unable to appreciate the difference). I would prefer to not use it in this case.



(Chuck H and it has never been an inconvience to me for when you, or anybody else here is helping).

I can now add more kerranggg! to see how I might like that. I now have a .68µf to try, since removing it from the 3-way switch taken off the amp.



Thanks again, for the generous help!

[One10]

Here is what actually came with my Gibsonette:

Chuck H, you mentioned the schematic being hard to read. I do have a slightly sharper version.
The reason I didn't post it, was at the time I was using it to document the many changes and soon it became very cluttered.

I also have the original unaltered circa 1960 Gibsonette schematic in my photo album & very sharp for those looking for it. Not sure how to make it available to everyone else.

[Chuck H]

Nice to see the actual voltages. Based on similar Gibson amps I've worked with I correctly guessed the voltage at about HV 300V.

Are these "many changes" your own? Or just the usual Gibson situation where the actual amps never match the schematics

[Chuck H]

I am still looking at a few other parts I can add to the order to lessen the shipping $$$ shock.

Yep. That's how I do it as well. Makes sense.

1.) Chuck H, that's fantastic that you can simulate the amp circuit in a program, it does seem to take the much, or all the guesswork out of the equation.

I'm often surprised how close the simulations are to what I see on the bench with an actual unit under test. I always do final tweaking by ear though. For simple mods like this the simulations are very handy. That said... I have no model for 6bm8 tubes so the CAD I'm using is 12ax7's and el84's. Which isn't going to be entirely accurate but likely close enough.

2.) Does the potentiomer have to absolutely be 500Ω? I have misunderstood as to how the pot interacts, or not with the 47k feedback resistor. I had thought the variable pot. didn't influence the negative feedback greatly from the 'speaker' output, except for negating certain freq. from the negative signal that is feeding back to V1B Cathode.
By this I mean for i.e. mid-high end freq. being 'added' back into the circuit depending on capacitor & resistor value combo you mentioned: 470nF, or 680nF & 500Ω var. pot., etc.
(I am asking because so far It has been easier & cheaper to find online correct size 200Ω, even100Ω var. pots that will fit in my amp, then 500Ω pots. that are instock on EBay.

In the simple circuit I posted above the pot itself IS the shunt resistance for the NFB circuit. Only the capacitor signal is variable on the wiper so the LF NFB ratio remains the same and the HF ratio from the cap is variable. The cap value is relative to circuit impedance.

Since the NFB is being inserted by adding a "tail" resistor to the V1B cathode (in this case that would be the presence pot) the gain characteristics of V1B are altered. Reduced gain and increased headroom at the grid. A smaller value tail would result in less change of the stock circuit characteristics. As long as the NFB series resistor is valued for the same ratio relative to the tail value the actual NFB affect will be about the same (Yes, Helmholtz, I know that it won't be exact because loop gain has changed but that starts to get too mathy for me ). But a lower value tail does mean the capacitor value needs to increase into the UF range because the NFB circuit impedance has been reduced.

Example: If you use a 100ohm pot for the presence circuit/NFB tail resistor then the series resistance would be 9.4k for the same NFB ratio (but I'll sim with 10k since it's a standard value and close enough). Reducing the circuit impedance to about a fifth of what it was means the capacitor value needs to incease about five times what it was for the same frequency alteration with the presence circuit. So the cap value would need to be raised from 680n to 3.4uf (or 3.3uf since it's a standard value and close enough). So...

If you want to use a 100ohm pot then use a 10k NFB series resistance and a 3.3uf presence capacitor.

3.) Hello Hemholtz, I just tried a 22µf @ 450v at V1B + 2.2k bypass and the low end was quite thiick & loud (from vol. 3 and up), but to me it sounds a little indistinct, like a smearing of the low freq. from 82Hz to about 200hz, slightly muddy? even.

See here? I was able to predict this with the simulation It's such a fun tool for me since I love the design process but not so much the actual soldering. So I can get really close to what I want before going to the bench.

[Helmholtz]

My Fender (Rivera) Super Champ uses a similar power amp circuit (NFB but no presence): https://el34world.com/charts/Schemat...hamp_super.pdf
The clean channel sounds great, certainly not muddy or indistinct.
Of course the speaker can make a big difference.

[One10]

Nice to see the actual voltages. Based on similar Gibson amps I've worked with I correctly guessed the voltage at about HV 300V.
Are these "many changes" your own? Or just the usual Gibson situation where the actual amps never match the schematics

Yes, it makes it easier to confirm how the amp is actually working.
Chuck H, Gibsonette mostly matched the schematic, except for them switching V1A & V1B as what was wired to what. Since the amp turned out to need alot of work, I 'gutted' it and put in all new parts, keeping the P.T. & O.T. as they tested well. So I added what I believed to be marked inmprovements not only for longevity & but also for a reduction in buzz, hiss & hum.

Below is 99% correct in how my amplifier originally was configured, this was inside the amp along with an instruction manual. The prior, original owner wrote "ECL82" on the schematic, it had never been worked on.


Thanks for explaining the pot-as-a-shunt relationship and math for a smaller var. pot. So you are saying that the ratio is in relation to the 47k feedback resistor / 500Ω variable pot. making it 94:1?

My Fender (Rivera) Super Champ uses a similar power amp circuit (NFB but no presence): https://el34world.com/charts/Schemat...hamp_super.pdf
The clean channel sounds great, certainly not muddy or indistinct.
Of course the speaker can make a big difference.

Helmholtz, you could be right about how it should sound. Maybe the problem is elsewhere. I have noticed that the Cannabis Rex 10 speaker is rather heavy, almost too heavy I believe, (for what it is mounted to). The 'baffle' board itself is quite thin & flexes from the weight of the speaker. I just attempted to beef it up with some strips of 1 x 1-1/2 solid alder and it stopped the flexing, but there still may be a problem with how the speaker mates to the thin redwood? I did expect the low-end to be thick but clear with the 22µf bypass cap.

[g1]

Below is 99% correct in how my amplifier originally was configured, this was inside the amp along with an instruction manual. The prior, original owner wrote "ECL82" on the schematic, it had never been worked on.

That image won't enlarge like the one in post #19, so can't really see it. ECL82 and 6BM8 are considered interchangeable.

[One10]

That image won't enlarge like the one in post #19, so can't really see it. ECL82 and 6BM8 are considered interchangeable.

I was finally able to fix it. I checked before posting and it showed as a clickable 'thumbnail' = fullsize. I deleted and uploaded the file again and appears to work. I know it's distorted, the scanner I used had problems feeding. Unfortunately I have lost the original copy, or I would have tried to upload a really clean, straight-lined copy.