Circuit Analysis Jim Kelley Amp

[Mr. Lime]

Hi guys,

I'm currently having some 6V6 amp projects laying around which I'm not completely happy with. Then I came across the Jim Kelley amp circuit which seem to sound really good - both clean and dirty.
Basically the Kelley amps started life as Ampeg SB12 Portaflex but Jim seemed to make some minor changes.

See the schematics for reference:
http://www.prowessamplifiers.com/schema ... y_facs.pdf

https://www.thetubestore.com/lib/thetub ... ematic.pdf

The first stage has a grid leak R decreased from 5M6 to 1M (lower input impedance) and the plate R is summarized to a single 150k. Optional 470p bright switch on the volume pot followed by clipping diodes.

Second stage cathode is fully bypassed and the tone stack got some more traditional values suited for guitar followed by the reverb circuit. The 2M2 grid stopper bypasses the clean signal but also attenuates the signal quite a lot - I guess that's why the 3rd stage was added with a relative low gain level of only 11.62. Again with clipping diodes in front of it.

Presence switch was added to the PI and the 7868 output tubes were replaced by 6V6 with very high plate voltage of around 500V.



My question is, how important is the 3rd stage as cold clipper? No reverb circuit in my amps and I understand that two cold clipper (with 4k7 cathodes) result in symmetric clipping. Wouldn't I get similar results with one gain stage after the tone stack with Rk 1.5 fully bypassed and leaving the 2M2 out?

[teemuk]

Yes. The design tries to maintain symmetric clipping even in the preamp stage. One should note that Kelley ultimately prefers push-pull power stage clipping (i.e. FACS and the later "Line Amp" power amp emulation), which by its nature is very symmetric.

Reputedly basis of Jim Kelley amp design was indeed a rebuilt SB12 fitted with 6L6 output tubes and Fender output transformer. The distortion was all power amp -based and Kelley preferred how the James -style tonestack operated in comparison to common "Fender/Marshall/Vox" -type circuit.

[ChopSauce]

I don't know how the cold clipper stage is important to this circuit but I'm pretty sure it's worthy.

At least this is what I could experience on a simple circuit, with and without a 3rd cold clipper stage. I was very sorry I had to suppress the cold clipper, which added too much noise, its triode being V1b.

[two tone]

The Jim Kelley ˋFortune‘ Amp is without Reverb

[Mr. Lime]

Yes. The design tries to maintain symmetric clipping even in the preamp stage. One should note that Kelley ultimately prefers push-pull power stage clipping (i.e. FACS and the later "Line Amp" power amp emulation), which by its nature is very symmetric.

Reputedly basis of Jim Kelley amp design was indeed a rebuilt SB12 fitted with 6L6 output tubes and Fender output transformer. The distortion was all power amp -based and Kelley preferred how the James -style tonestack operated in comparison to common "Fender/Marshall/Vox" -type circuit.

Funny that some people compare the Kelley with kind of Dumble distortion which is basically an overdriven preamp.
I would rather compare it to an 5e3 maybe?

I don't know how the cold clipper stage is important to this circuit but I'm pretty sure it's worthy.

At least this is what I could experience on a simple circuit, with and without a 3rd cold clipper stage. I was very sorry I had to suppress the cold clipper, which added too much noise, its triode being V1b.

KOC has his clone circuit in TUT5 leaving the reverb out as well. The 3rd cold clipping stage and the 2M2 are still there and probably introduce hum as well.

The Jim Kelley ˋFortune‘ Amp is without Reverb

Is there a schematic somewhere?

[dorrisant]

Is there a schematic somewhere?

Here: https://ampgarage.com/forum/viewtopic.php?f=6&t=29580


Edit: Wrong schematic... Doh!

[Mr. Lime]

Those diodes have a Vf of 1V each, so they clip one halfwave at 2V after the first gain stage.
I guess the signal doesn't even clip if the volume pot isn't maxed..

Ditching the 3rd stage and replacing the 1N914 diodes with back to back zeners might do the trick as well.
Someone might consider modifying the global feedback loop to a real presence control and fully bypass the 4th stage's cathode to recover the volume that gets lost at the diode clipping circuit and the missing gain stage.

I have to get into circuit simulations..

[Mr. Lime]

If I would like to add a FX loop, where do I insert it best?

Using all 4 gain stages + split load PI, one unused triode is left over.
I guess placing a cathode follower behind the tone stack and using the first cold clipper as recovery stage (with additonal cathode bypass cap) would be a good start?

The reverb circuit is was placed quite early in the preamp..

Any other suggestions?

[musant]

Hi guys,

I'm currently having some 6V6 amp projects laying around which I'm not completely happy with. Then I came across the Jim Kelley amp circuit which seem to sound really good - both clean and dirty.
Basically the Kelley amps started life as Ampeg SB12 Portaflex but Jim seemed to make some minor changes.

See the schematics for reference:
http://www.prowessamplifiers.com/schema ... y_facs.pdf

https://www.thetubestore.com/lib/thetub ... ematic.pdf

The first stage has a grid leak R decreased from 5M6 to 1M (lower input impedance) and the plate R is summarized to a single 150k. Optional 470p bright switch on the volume pot followed by clipping diodes.

Second stage cathode is fully bypassed and the tone stack got some more traditional values suited for guitar followed by the reverb circuit. The 2M2 grid stopper bypasses the clean signal but also attenuates the signal quite a lot - I guess that's why the 3rd stage was added with a relative low gain level of only 11.62. Again with clipping diodes in front of it.

Presence switch was added to the PI and the 7868 output tubes were replaced by 6V6 with very high plate voltage of around 500V.



My question is, how important is the 3rd stage as cold clipper? No reverb circuit in my amps and I understand that two cold clipper (with 4k7 cathodes) result in symmetric clipping. Wouldn't I get similar results with one gain stage after the tone stack with Rk 1.5 fully bypassed and leaving the 2M2 out?

It's really a great analysis, man. Well done. I would like to clarify a couple points, if I may. Regarding the diodes, they're really not clippers as much as they're voltage clamps - to keep the grids from going too far negative. That causes a shift in duty cycle at lower frequencies (asymmetrical), gives a flubby, compression-like effect. It's more of a DC thing than an AC thing. A grid stopper is the latter-day method of dealing with the issue.

The 2.2 meg resistor is used to provide isolation between reverb send and reverb return. The smaller the value, the less reverb you end up with. The following stage is a mixer and reverb recovery amp.

[Mr. Lime]

It's really a great analysis, man. Well done. I would like to clarify a couple points, if I may. Regarding the diodes, they're really not clippers as much as they're voltage clamps - to keep the grids from going too far negative. That causes a shift in duty cycle at lower frequencies (asymmetrical), gives a flubby, compression-like effect. It's more of a DC thing than an AC thing. A grid stopper is the latter-day method of dealing with the issue.

Thanks for the kind words and the input!

I reallly dig this circuit but can't get around an elegant solution for an effects loop.
In my amp I plan to convert some day I have one triode left over (no reverb circuit) and I'm afraid another triode as recovery stage for the loop could influence the overall sound because of phase shifting.
My first attempt would be to set a cathode follower after the tone stack and fully bypass the third gain stage to recover from fx-level. It's the same problem like the Trainwrecks have as the distortion comes from the power amp what makes the fx loop somewhat useless..

I would be glad to have some suggestions to think about.

[musant]

It's really a great analysis, man. Well done. I would like to clarify a couple points, if I may. Regarding the diodes, they're really not clippers as much as they're voltage clamps - to keep the grids from going too far negative. That causes a shift in duty cycle at lower frequencies (asymmetrical), gives a flubby, compression-like effect. It's more of a DC thing than an AC thing. A grid stopper is the latter-day method of dealing with the issue.

Thanks for the kind words and the input!

I reallly dig this circuit but can't get around an elegant solution for an effects loop.
In my amp I plan to convert some day I have one triode left over (no reverb circuit) and I'm afraid another triode as recovery stage for the loop could influence the overall sound because of phase shifting.
My first attempt would be to set a cathode follower after the tone stack and fully bypass the third gain stage to recover from fx-level. It's the same problem like the Trainwrecks have as the distortion comes from the power amp what makes the fx loop somewhat useless..

I would be glad to have some suggestions to think about.

I use a pair of audio transformers. Requires a high impedance ratio. Resulting Voltage ratio gives instrument level send and return.

[Mr. Lime]

I think one main reason for using such a high anode voltage is to maximize the cathodyne PI voltage swing and prevent PI clipping.
Like Trainwreck amps this amp creates the distortion by clipping the power tubes so I guess a FX loop might not work as expected.
I would rather try an attenuator Line out and use a transparent power amp in parallel..

[Mr. Lime]

Another idea what to do with a spare triode;

Parallel cathodyne PI for even more headroom and less clipping so we might get away with less PSU voltage..

[Suckseeder]

Looking at the 30/60 schematic,I noticed the 6V6 screens are tied to B+ via a 2k/10w resistor.
Why is it done like this,and not supplied from a lower voltage stage?
450v plus is really hard on 6V6 screens

[novosibir]

Looking at the 30/60 schematic,I noticed the 6V6 screens are tied to B+ via a 2k/10w resistor.
Why is it done like this,and not supplied from a lower voltage stage?
450v plus is really hard on 6V6 screens

I think, Jim has used a given PT w/o a 2nd lower HT secondary, presumably a PT designed for use w/ 6L6's

But the higher the screen voltage, the quicker the response of the output tubes

And yes these high plate & screen voltages are extraordinary stress for 6V6's, even for i.e. 50' RCA 6V6GTA Blackplate's

You can de-stress the output tubes by swapping in a 3...5W/33...50V zener diode in series w/ a 2W 470 ohms resistor in front of each screen...

... but you'd loose dynamics & livelyness of the amp's output stage

Or you try to still find a quad of Red Bank Bendix 5992 output tubes, which easily can handle these high plate & screen voltages:

https://tubedepot.com/products/bendix-5992-6v6gt

https://www.rfparts.com/5992.html

Btw. I'm one of the very few lucky guys on this planet, who still does have a quad of these rarities in his NOS tresor. But mine aren't for sale

Larry