Jim kelly single channel layout ?

[Colossal]

Very cool Smokebreak!

[renshen1957]

He claims to be getting 60 watts RMS from a quartet of 6V6's. I wonder how he's doing that.

Hi,

The how of 60 Watts from quartet of 6V6GTs (Sylvania in the schematic and Groove Tubes also mentioned in later literature) rated at 12 W was accomplished with an OT transformer of 4300 ohms, and voltages 485V on the plates at idle which under load would have 423V on the plates.

Kevin O'Connor featured the two channel amp in TUT vol 5, but with an updated (Super Stock) Power Supply (different from Kelley's original).

After Suhrs involvement and reissue, the website deleted the two channel amp schematics (with voltages); one as originally built and a later version with LED Bias and modifications to the Gain Channel

As originally built, one channel had a 12AT7 tube and the other had a 12AX7 with the signal diodes. The later version had 12AX7's in both channels, the gain channel had removed the signal diodes.

Kelley's concept was to keep the amp clean and let the power tubes supply the distortion, which was contrary to what MESA and Marshall were doing at the time.


Best regards,

Steve

[renshen1957]

According to Jim's schematic, the OT primary and the screens are supplied from B+1 (100uF filtering). There is a 2k 10W dropping resistor which drops the screen voltage by a mere 3V and is "shared" across all four power tubes. This is similar to Vox's approach where the OT and screens share the same supply node.

Any thoughts on what is effectively a single, shared screen resistor versus individual ones for each power tube? I think I recall Tubeswell commenting about this having a positive effect on sensitivity (don't quote me and I'm not quoting you Tubeswell)...just trying to remember.

EDIT: (or maybe it was Gingertube/Ian who commented?)

Hi,

The single shared screen resistor is either a) in Vox typical English frugalness, don't pay for 4 resistors one you can bet by with one; Traynor in Canada also used single shared resistor) or b) the shared resistor was supposed to make the tubes interact more uniformly.

I have never bothered to check out the latter claim, as I always use a separate screen resistor for each tube.

The 10 Watt resistor is over spec'd for the current, maybe the 2K ohm 10 Watt resistor was available for cheaper than lower wattage. You might write to Jim Kelley and ask why the 10 Watt instead of 2 W or 3 W?


Best regards,

Steve

PS Gar Gillies of Garnet (Canada), pointed out that some 2 power tube HiWatt amps used 100 ohm tube from the screens; the other leads went to a single 500 ohm resistor.

[renshen1957]

They help to reduce blocking distortion, kind of in lieu of a grid stopper.
Normally you could use a diode and a zener back to back set so they come into play at around the tubes cut off input voltage say around 5v ish.

In this case jim used it at around 1.2V to add a little diode clipping at higher settings of the volume pot.

Shane

Hi,

And I thought the diodes were to keep the preamp tubes clean, as Jim Kelley said in various interviews that he designed amp so the distortion came from the power tubes, he wasn't into preamp distortion.

bEst rEgards,

Steve

[Colossal]

The single shared screen resistor is either a) in Vox typical English frugalness, don't pay for 4 resistors one you can bet by with one; Traynor in Canada also used single shared resistor) or b) the shared resistor was supposed to make the tubes interact more uniformly.

Steve,

Thanks for the reply. My suspicion is b), that this was a design choice, not a cost issue.

[hbamp]

John Suhr changed the shared screen resistor in 2 X 3K9@5W. There must be a good reason for that.

[martin manning]

A shared screen resistor results in less screen voltage sag, and so presumably less compression. This is because when one side of the p-p pair goes into cut-off, the other side has the screen resistor all to itself, and sees less voltage drop as it draws more current.

[oj]

This schem is accurate.

Well, V1b are missing 2,2k/20uF on the katod.

[Synchu]

Alrighty.
I am in the process of finalizing a build that is ressembling Jim Kelley amp.
It is the pre section that I am trying to mate with a SE, Class A 6V6/5881 PA. No reverb as well. Don't ask why it turned like this

Using both schematics posted (and the Ampeg SB12/7868 one).

Sounds good so far and I even did the PI, using only the plate leg to feed the PA, no issues whatsoever and sounds good, albeit it is leaning very much on the dirt side and up to 9 o clock there's sort of cleanish sounding crunch which when Gain is maxed turns into very respectable lead sound.

There might be few things that I am missing from the schematic and most probably load from the reverb section hasn't been well simulated with a single 470k grid leak just after the recovery stage, but seems to be good so far.
I will test how it sounds with the PI skipped alltogether (as well as bleeding the cathode non-inverted output to ground), will play some more with the NFB, put the thing on the scope and will post the results.
Certainly it wouldn't get as clean as PP 4x6V6 pushed to the limit, but worths trying clean it up a bit:)
Overdrive sounds great.

Niki

[dorrisant]

Very cool Smokebreak!

Yes... any clips?

[dorrisant]

Pm me for the layout!

Not to bug you too much, but I pm'd you... and just for the record - I'm not Ceriatone...

[Synchu]

I pm'ed too some time ago, as I am tweaking my creature of madness and want to verify that I caught beside obvious errors in the posted schematics.
As I am feeding the pre into SE, Class A, PA, I did some tweaking so far, but some reassurance would be right on time.

Niki

P.S. Not Ceriatone too

[bal704]

I've tried to contact him as well. If anybody has an updated schematic, or layout, let me know.

[Synchu]

The results from my build are rather promissing, despite the quite different architecture of the power amp - SE 5K output, Class A, works fine for 6V6, 5881, 6L6WGB, to an extent 6L6GC.

Power transformer is putting 334V on the plate and I keep screen few volts down (high ratio voltage divider of the screen resistor and 220k to ground).

I played with the preamp voltages to what sounds sort of best to me as this PT cannot supply 375V to the 3rd/PI stages, as marked on one of the schematics.
I did the cathodyne, using only the anode leg to feed the power amp, not really any particular need for that, except sort of following the gain structure and tone nuances that come with it.

Class A power stage takes much less input voltage before it gets overdriven, so I removed the bypass cap of the second stage and increased global NFB to approx. 10:1 ratio.
Overall sound, at first, came a bit bass heavy so put a 150pf bright cap on the Volume.

Few other tweaks are under way before call it complete, but even now I am quite happy with the result.

Would like a bit more clean headroom and may do the split anode resistor thing on the first stage (as in the Ampeg SB-12 early versions, from which this amp borrows the overall architecture and some values) and put in on a switch, as the K-60 is done.
Will not definitely get to the clean headroom of the original, given the very different power amp and voltages I use, but worths trying get something that sounds good The original is tighter sounding due to the push pull, Class AB design while my thing is abundant of even order harmonics (2nd esp. blend almost like a chorus on few notes) and more tweedy.

Will post pictures and sound in a separate thread, once completed, just wanted to share.
Niki

[hbamp]

My virusdetector is blocking me from TAG so I have to startup with no protection at all! Please contact Collosal for the layout. It's OK for me to share! Please don't publish openly!