Class AB2: $5, .5 hr

[Ray Ivers]

Arthur,

I have to admit, I completely mis-read the Isodyne schematic the first time around - I concentrated on the two left-most triodes, seeing the two right-side ones as push-pull outputs, for some reason - maybe because of the bias control, or more likely just losing it, I guess. This is kinda-sorta a "buffered Schmitt w/local NFB".

The Isodyne could work great for power-grid drive IMO, but it would have to be modified for direct-coupling to the output tube grids. The existing positive bias supply would have to be changed to a low-Z negative one and run to the PI and CF Rk's as well as the bias control, and the 300V PI and 400V CF B+ reduced - the ultimate goal being a correct negative bias voltage at the CF cathodes/output grids, along with low-impedance positive-swing drive provided by suitable (i.e., non-12AX7) relatively hot-biased CF stages. If one was going to all this trouble, the CF's could be eliminated completely and a low-rp PI tube (w/suitable Rp value - ideally less than 5K) used to drive the output grids directly, in a straight-ahead Schmitt or PI/diff amp configuration.

Ray

[Analog Assassin]

Could this be a cure for the loose bottom end on blackface fenders? I find it daunting often to get a tight sound out of my super reverb without thinning down the bass too much.

[Ray Ivers]

Could this be a cure for the loose bottom end on blackface fenders? I find it daunting often to get a tight sound out of my super reverb without thinning down the bass too much.

Yes - I assume you mean the Zener mod (the AB2 circuit would also work, but the Zener thing's so much easier to do). In my testing I used .47uF coupling caps from the PI to the output grids, and I'm sure I could have gone a lot bigger than that. In AB1 these big caps seem like ridiculous overkill - 12Hz F3, who needs to go that low? - but when the grids conduct you're looking at more like 180Hz for F3, inside the NFB loop (where a reduction in bass is a reduction of bass NFB and apparent damping factor, with corresponding reduced control of speaker-cone motion).

Bigger coupling caps + Zener mod = more/tighter bass when the power amp's working hard.

Ray

[Satamax]

Huh, guys! I've tried to follow, but with my limited knowlege, i haven't understood much.

There's one thing i'm thinking tho, isn't an interstage transformer/spliter the easiest solution?

Bye.

Max.

[Arthur B.]

The Isodyne could work great for power-grid drive IMO, but it would have to be modified for direct-coupling to the output tube grids. The existing positive bias supply would have to be changed to a low-Z negative one and run to the PI and CF Rk's as well as the bias control, and the 300V PI and 400V CF B+ reduced - the ultimate goal being a correct negative bias voltage at the CF cathodes/output grids, along with low-impedance positive-swing drive provided by suitable (i.e., non-12AX7) relatively hot-biased CF stages. If one was going to all this trouble, the CF's could be eliminated completely and a low-rp PI tube (w/suitable Rp value - ideally less than 5K) used to drive the output grids directly, in a straight-ahead Schmitt or PI/diff amp configuration.

Ray

Could you give the isodyne it's own floating supply and then use a variable voltage reference to bias it?

Huh, guys! I've tried to follow, but with my limited knowlege, i haven't understood much.

There's one thing i'm thinking tho, isn't an interstage transformer/spliter the easiest solution?

Bye.

Max.

The next question would be "what's the driver?"

[Satamax]

Could you give the isodyne it's on floating supply and then use a variable voltage reference to bias it?



The next question would be what's the driver?

I like the way fender did it on the PS400. 6L6 and interstage transformer/spliter.
There's another weird one, teh SMF tour series from sound city, i've never understood why an EL34 phase spliter with caps afterwards! Weird!

Bye.

Max. Ray, sorry to hijack!

[Ray Ivers]

Arthur,

Could you give the isodyne it's own floating supply and then use a variable voltage reference to bias it?

Yes, absolutely - the floating supply is pretty much what I meant to describe (in my long-winded way), and the VVR should make an excellent bias supply.

The Isodyne's a really nice PI/driver setup - IMO it would also make a great (cathode-)driver for a P-P/P-P xfmr - and if the 4-gain-stage requirement is a problem the CF duties can always be performed by MOSFET's.

Ray

[Ray Ivers]

Max,

Hey, everybody hijacks from time to time...

There's one thing i'm thinking tho, isn't an interstage transformer/spliter the easiest solution?

Yes, it really is - one gain stage, a few associated components, a transformer, and you're done. As Arthur pointed out, though, it's gotta be the right type of gain stage for the job to work well, and there's also the power-supply thing to work out too.

I like the way fender did it on the PS400. 6L6 and interstage transformer/spliter.

...and the 300PS uses a 6V6 and IT - still pretty beefy!

Ray

[Arthur B.]

I like the way fender did it on the PS400. 6L6 and interstage transformer/spliter.
There's another weird one, teh SMF tour series from sound city, i've never understood why an EL34 phase spliter with caps afterwards! Weird!

Bye.

Max. Ray, sorry to hijack!

http://www.schematicheaven.com/barga...y_smf_tour.pdf

Seems like a self split EL34 amplifier for the PI.

[Ray Ivers]

Arthur,

The SMF PI/driver looks to me like like a self-biased LTP using two EL34's as the gain stages. IMO they must have been positively awash in EL34's and octal sockets to do it like this, as I really can't see any benefit the EL34's add; the PI component values and B+ would have worked fine with EL84's, or even a 12AT7! If only they had spent the transformer budget on a good driver unit rather than that balanced out... or that super-weird totem-pole distortion circuit... which might sound OK, for all I know...

Ray

[Arthur B.]

Arthur,

The SMF PI/driver looks to me like like a self-biased LTP using two EL34's as the gain stages. IMO they must have been positively awash in EL34's and octal sockets to do it like this, as I really can't see any benefit the EL34's add; the PI component values and B+ would have worked fine with EL84's, or even a 12AT7!

I think they were trying to pry the grids positive through brute force.

If only they had spent the transformer budget on a good driver unit rather than that balanced out... or that super-weird totem-pole distortion circuit... which might sound OK, for all I know...

Ray

I don't see a totem pole circuit anywhere. The distortion circuit seems to consist of a gain stage casccading into a concertina cascading into two more gain stages which drive the transformer primary.

[BobW]

Ray, I finally got to try the Zener mod at the second stage preamp of my Dumble clone. Looking at the scope wavefrom I determined that a good starting Vz was 6.3 volts, and this stage was the best candidate. I didn't have any 1N4001 - 1N4007s but a fast recovery 1N4936 would do the trick. After installing, the two diodes, the slightly gritty tones at higher volumes disappeared and more of a singing tone resulted. This clamped the higher F transients, although based on the waveforms alone, it could be further improved. Stores were closed on Sun so I'll try a lower Vz's 4.7, 3.9, etc. and report back. Thanks for posting this mod!

[Ray Ivers]

Bob,

Most excellent! I was hoping it would work well on preamp tubes, too.

Ray

[BobW]

Ray, It seems the 6.3 Vz was the best out of smaller 4.7, 3.9, and larger, 9, 12 zeners. This ckt was also tried at the the 3 previous stages w/o any tone improvement, so am satisfied with this mod at OD2 on my dumble clone. This is exactly the ckt I need to help remove the grit tone in the OD channel of my peavey (sorry) classic 30. thanks for your help.

[BobW]

I should have mentioned, you won't hear the zenering effect until you have the amp cranked above normal bedroom volume. Using too low a value of zener in this circuit can cut too much off the top end, so either increase the value or add a series resistor to smooth the zener effect.