Explanation of this circuit

[Manic]

This is a single channel Jim Kelly. When I saw this scem I was a bit shocked at the values of the components. Are My rookie eye is seeing cooler gain stages but one to 2 more? Maybe one makes for the loss in the tone stack, or the reverb??? Im just not quit sure what to make of this thing. Its intreiging, I know Jim Kelly amps are supposed to be the BOMB.
I was thinkin about trying some of his "tricks" in My 5e3.

Could someone explain whats going on with this amp.

As always thnks, C Smith

[MWJB]

Which tricks were you considering for the 5E3? These would appear to be substantially different amps. A 5E3 can cover a lot of territory, I'd be concerned about reducing it's versatility.

In short the amp has a volume control following the input triode, a Baxendall tone stack between the 2nd & 3rd triodes, then a cathodyne PI (assuming we are excluding the reverb). You don't have enough triodes to clone this preamp. Converting to fixed bias, or incorporating the Kelly NFB loop would seem quite feasible?

[Manic]

The back to back diodes look interesting on the grid (V1b)...wonder what they do?

The presence switch I might like to try.

I would want to remove the reverb, or not include it.

I have tried a similar plate resistor setup on another amp I have (V1A and V1b), but I dont "get" the cathode values of those stages. I assume it setup for less crunch.

The Baxendall stack has always interested Me.

I have one more spot in My chassis for a triode (9 pin) that is empty.

As far as versatility goes, I feel My 5e3 is a one trick pony. Turned down it sounds pretty bland and barky. Cranked to around 80% it sounds pretty awesome.

Does this amp look like a gain monster? What about the JK looks like it will cut down on the versatility? From what I have read about JK amps, He tried to maximize power amp OD and minimize preamp OD.

Thnx for responding and thnx for the help.

C Smith

[MWJB]

Cathode values at V1a (1.5K) & V1b (2.2K) are pretty middle of the road values. Typical of Gibson, Fender, Danelectro and many, many others.

"As far as versatility goes, I feel My 5e3 is a one trick pony. Turned down it sounds pretty bland and barky. Cranked to around 80% it sounds pretty awesome." Well a 5e3 has 2 different sounding channels and quite a bit of variation depending on where the volumes are set. I'm not saying that it is beyond improvement, to better suit your needs, but there are simple mods that might help it? Check out Mission Amp's mod kits, cathode/fixed bias switch etc?

Kelly - I've never heard one, so I don't know what it sounds like (I'm sure they sound fine). This is the thing, we have the schematic but there's more to an amp's sound that that. How that design is physically executed will affect the sound. If you don't know what the amp sounds like then I don't understand why you would butcher an amp of a totally different design to make it into an unknown (to you) amp, with no tonal reference as to what you are shooting for? If you really don't like the 5E3 maybe sell it and use the proceeds to put towards building the Kelly, exactly as per the schematic?

By the way, I would expect that plate resistor at V1a to be 150K not 150ohms.

I'm not trying to discourage you from experimenting, but I always try and get folks to think about how they want an amp's sound to specifically change and then implement that change, rather than blindly chasing component values & circuits.

[DirtyGeorge]

It looks like the schem is missing a coupling cap between V1a and V1b. I don't think it would work as drawn.

The diodes on the grids of V1b and V3a are probably to clamp the positive signal swing to prevent bias shift from grid current flow.

DG

[Bruce / Mission Amps]

Also... did anyone notice that the cathodyne driver does not have a grid load resistor?
I don't think I've seen one biased quite like that.
I mean with positive plate voltage applied on grid but with out a voltage divider, who's missing resistor would have been the grounded grid load resistor.

[Manic]

here's the link to the page I got the info from........
http://ps.uci.edu/~jwkelley/schem.htm

[d95err]

Could it be DC coupling between the first two stages (rather than a missing coupling cap)? I know the AX84 people use that in their Melissa amp.
EDIT: Nope. Now I see that the other similar schematics have the coupling cap.

The diodes look like they are used for hard clipping. Rather than using back-to-back diodes like in stomboxes or later Marshalls, there are two sets of diodes that clip the signal at different parts of the circuit. The first side of the waveform is clipped by the first set, the signal is then inverted in the following gain stage, and then inverted side is clipped after the 2M2 resistor (at "Y").

Very unusual design. Would be cool to try...

[teemuk]

About those back-to-back diodes: Normally the signal’s “zero point” is the reference defined by the grid resistor. (in most cases it’s 0V due to connection to common). However, when the circuit is driven into clipping the tops of the signal's positive halfwave will be clipped off. The resulting asymmetry of the signal would normally cause the "zero point" to drift into a negative value. However, since the diodes clip everything that is exceeding about –1.2V, their clipping effect becomes increased along the DC level shift. It seems like this is an arrangement to acquire more symmetric clipping and of course more distortion.

[DirtyGeorge]

Teemuk, you are right. I didn't notice the polarity of the diodes when I said they clamped the pos signal swing. I like to think of the diodes in this arrangement as providing a path to discharge the coupling cap during the neg signal peak. As you pointed out this prevents the bias shift ("zero point" to drift into a negative value) that results from the grid current flowing into the cap during the pos signal peak.

DG

[Manic]

It was suggested to me that the diodes might act as some sort of limiter, kinda like a grid leak but gated by the voltage rating of the diodes. Two diodes in series makes the (gate) twice as hard to open.

FWIW, C. Smith