Rozenblit 6as7 otl questions

[Klimon]

http://www.bonavolta.ch/hobby/en/audio/6as7_2.htm

Few questions:
1) Is the input circuitry sufficient to drive the otl stage to full output = can I change the R1 grid resistor with a volume pot and drive the output stage into clipping?

2) What always turned me off on this circuit is the 4-section input-driver-phase splitter circuitry (><KISS); when taking a closer look the second stage (two sections of 6sn7) looks very odd to me; is that stage part of the phase-splitter or what else does it do?

3) Any suggestions for simplifying/improving the circuitry preceding the output stage? (I'm aware of the fact that a redesign probably isn't a beginner's task due to the massive gNfb involved)

Thanks -- Simon

[AKSA]

Klimon,

This should easily drive the output to full clip.
2nd 6SN7 stage uses two triodes to give high Z input, reasonable gain, NO phase inversion. This is necessary to return global nfb to the cathode of the input tube.
The phase splitter circuitry to drive the output tube is accreddited to Julius Futterman, and is something of a legend in OTL designs. This rides the driver off the output rail, while delivering equal, antiphase drive to the two sets of triodes. It's very, very neat, and elegant beyond most circuits you see.

Hugh

[Klimon]

So it's a package deal. Thanks H.

Simon

[Miles Prower]

Originally posted by Klimon
2) What always turned me off on this circuit is the 4-section input-driver-phase splitter circuitry (><KISS); when taking a closer look the second stage (two sections of 6sn7) looks very odd to me; is that stage part of the phase-splitter or what else does it do?

It isn't so odd. V2 forms a noninverting differential voltage amp. It gives the advantage of presenting the driving stage a friendlier load that's less capacitive to improve slew rate. At least it would if he hadn't screwed that up. R6 needs a bypass capacitor connected across it to make the input section of the 6SN7 act more like a cathode follower that doesn't have the Cmiller problem.

Otherwise, it's a conventional Futterman output stage.

I did the same thing in a design I'm working on to make a 6BX7 verticle deflection dual triode present a friendlier load. These may not have much gain, but what they do have makes for a nasty Cmiller.

[carolus]

Hi Simon,
Why not simply built the Trancendent T8 or the "Beast" ...
I did it with PL519 ....
Allez, salukes.
Karel

[desperateaudio]

Originally Posted by Miles Prower

It isn't so odd. V2 forms a noninverting differential voltage amp. It gives the advantage of presenting the driving stage a friendlier load that's less capacitive to improve slew rate. At least it would if he hadn't screwed that up. R6 needs a bypass capacitor connected across it to make the input section of the 6SN7 act more like a cathode follower that doesn't have the Cmiller problem.

Otherwise, it's a conventional Futterman output stage.

I did the same thing in a design I'm working on to make a 6BX7 verticle deflection dual triode present a friendlier load. These may not have much gain, but what they do have makes for a nasty Cmiller.

I am thinking about trying this version as I have tons of NOS NIB 6AS7's. I have 2 questions if you guys can help.

1. Is the 3rd stage 6SN7 splitter using only half the tube or is it paralleling the 2 halfs of the 6sn7?

2. What are the bias voltages? seems like -70 for the top and -220 to -240 for the bottom. Can someone enlighten me.

Thx in advance

[Miles Prower]

Originally Posted by desperateaudio

I am thinking about trying this version as I have tons of NOS NIB 6AS7's.

If you have tons of NOS 6AS7s, I would suggest forgetting all about this particular design. It SUX. Instead, use the 6AS7s to build a Circlotron OTL: similar level of complexity, much better balance since you're not forcing balance.

Just keep in mind that NOS 6AS7s can't be pushed quite as hard as the newer versions.

[desperateaudio]

Originally Posted by Miles Prower

If you have tons of NOS 6AS7s, I would suggest forgetting all about this particular design. It SUX. Instead, use the 6AS7s to build a Circlotron OTL: similar level of complexity, much better balance since you're not forcing balance.

Just keep in mind that NOS 6AS7s can't be pushed quite as hard as the newer versions.

I hear you but building for multi testing purposes so not as big an issue but thx. With that said two things.

1. Do you have the information to help me answer my questions?
2. Have you seen BR's el509 OTL from his book? If so what is your opinion on that design (splitting to the output is done with seperate triode sections).

Thx

[Miles Prower]

The way I'm reading that schemo, both the cathodyne driver and the first preamp are using one section of the 12AX7 and 6SN7 respectively.

Given the Vpp= 150Vdc for the finals, a Vgk= -65 to -70V looks about right.

I've seen other OTL designs that use different types. However, I haven't found anything that works better than the 6AS7. When Ralph designed the Atmasphere line of OTLs, he knew what he was doing.

I still don't like the design here at all. If you insist on going this route, I'd look into the Inverted Futterman topology instead. It avoids that positive feedback to force a cathode follower to look more like a grounded plate amp, but it still has a potentially nasty DC offset problem that can poof speeks. It's still a problem with the Circlotron design, but far less of a potential problem.

[desperateaudio]

Originally Posted by Miles Prower

The way I'm reading that schemo, both the cathodyne driver and the first preamp are using one section of the 12AX7 and 6SN7 respectively.

Given the Vpp= 150Vdc for the finals, a Vgk= -65 to -70V looks about right.

I've seen other OTL designs that use different types. However, I haven't found anything that works better than the 6AS7. When Ralph designed the Atmasphere line of OTLs, he knew what he was doing.

I still don't like the design here at all. If you insist on going this route, I'd look into the Inverted Futterman topology instead. It avoids that positive feedback to force a cathode follower to look more like a grounded plate amp, but it still has a potentially nasty DC offset problem that can poof speeks. It's still a problem with the Circlotron design, but far less of a potential problem.

Thx so much. So you think the -240 is also correct as I suggest?

Also I should clarify that I am using 6080's for the design test not 6as7's sorry got to generic on you there.

I own several NYAL OTL3's and have owned Futterman H3's in the past. Are you saying you would rather go with the Futterman front end instead? Please clarify. THX!!!