Please help analyse 807 PPP

[cjcc67]

Is this a feasible build? I like the power output given the valves are wired as triodes. I'd like to make it switchable between triode and tetrode.
I'm stuck at the starting gate looking at the output stage as the bias seems quite (overly) simple. Is this basically a williamson amplifier without feedback?
How can i tell from the circuit diagram what class the amplifier has been biased for?

[jechentau]

34V 680 Ohm so 50mA; class A for 807

the tubes do not balance current as wired such, thus must be matched.

they are not AC bypassed, sacrificing speaker damping and gain. but that may be desired.

[Miles Prower]

Originally Posted by cjcc67

Is this a feasible build?

Yes, it's quite feasible. Whether or not it's worth doing is a whole 'nother story.

I like the power output given the valves are wired as triodes. I'd like to make it switchable between triode and tetrode.

You don't want to do that. Operation as a pseudotriode and a pentode will be a major sonic compromise since each mode is likely to want a vastly different load.

Is this basically a williamson amplifier without feedback?
How can i tell from the circuit diagram what class the amplifier has been biased for?

It's not a Williamson at all. That schemo used an LTP phase splitter, and the Willy uses a cathodyne driving a gain stage/grid driver.

As for that design, I'd say FUDEDABOUDDIT! Using four 807s to get 24W is ridiculous, as you can get 26.5W from a pair if you use 'em like they were meant to be used: as pentodes.

Vixen Main Schemo

Screen Regulator

Power Supply

This will get you 30W / Channel. Adding local NFB and a modest amount of gNFB will sound as good as PP triodes.

[cjcc67]

Thank you both for your replies.
J i now understand the calculation from the 54v @ cathode 79ma.
From this how can i tell it is not AB. this is my data source http://www.retrovox.com.au/STC807.pdf
Miles i agree about the efficiency, my current thinking is triode for the low THD, the idea of switching into pentode increases power with increased distortion thus having the best of both worlds.
Can the OP transformer have some different taps or windings in the primary?

[revintage]

This is not Class-A! The OPT impedance is totallay wrong chosen, it should be something like 4-5k. The amp will give just above 10W in Class-A then clips at ca 25W AB. It will not have a nice distortion pattern.

[aardvarkash10]

stick with miles - he knows of what he speaks...

Running the 807 pentode AB1 will give you 6 watts or more of Class A moving into Class AB for power, at which point you are not hearing any distortion unless its absolutely gross - unlikely with a good design.

In short, 807s in pentode using miles' Vixen or similar is not going to distort such that you will notice - and certainly no worse than the design you originally posted!

[revintage]

This is not Class-A! The OPT impedance is totallay wrong chosen, it should be something like 4-5k. The amp will give just above 10W in Class-A then clips at ca 25W AB if the cathode resistors where bypassed. It will not have a nice distortion pattern. BUT.. it will not work so well in AB as the cathoderesistors are unbypassed. A total disaster!

[jechentau]

indeed it will dip into ab1 driven hard.

the datasheet even describes 3k per pair for ab1, which this is not even yet.

btw i looked at these loadlines for triode; near match given the idle operating points described in the schematic. http://www.mclink.it/com/audiomatica/tubes/807.htm

something could be made of it very well though

[cjcc67]

i thought it had some issues as i couldn't understand the OPT selection either. Putting aside imedpance matching issues for the time being how do we improve the operation, change the cathode resistors to a better location on the load line?
How do i calculate the grid voltage?
thank you all for your interest
Craig

[revintage]

Hey Craig,
You can try my simple www.revintage.se/PPABAMP.xls

This is disastrous:


Purple=Pamax, Red=B Green=Class-A loadline

This method is simplified but works good enough.

But is it worth the effort to correct the circuit? I would go for an amp with only one gainstage. If you haven´t the 807s you could as well go 6L6.