Designing a simpler tube opamp from scratch.

[brad]

Sorry if I sound like a broken record about freakin' tube opamps, but I've gone completely Captain Ahab on this White Whale 

I've spent all afternoon reading about opamps and trying to figure out if it would be possible to make something simpler than a Philbrick opamp that uses 12v 12U7 tubes and doesn't need a bipolar supply.  Franky, it's done my head in!    This is all I got:

[zachary vex]

well, if you give up and just want a couple, there's some on ebay right now.  two of them for about 26 bucks.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=5820851251

[The Red Rooster]

Just my 5 cents;

Looking at your sch I noticed that you're using 1 1/2 tube - if you use the last 1/2 of the second tube as a follower, you'll have the advantage of a lower Z-out instead of the higher Z-out of the output stage you're using now. Another thing is that using a follower won't affect phase since it's got no phase shift - this means you'll stil be able to use feedback as intended.

Regards

Kurt

[puretube]

be aware of the hi voltage on plate V1b: probably too high for grid V2a...

differential input circuits are quite common in hi-fi amps...



http://www.tubecad.com/

[brad]

Thanks for the replies guys.

From this point, it looks like adding the final triode as a follower would basically make it into a unipolar Philbrick that's even more complex than the real deal!  Perhaps the only way to keep things simple is to tweak an original Philbrick design to make it suitable for 12U7s being run off a bipolar 12v supply from a 24v wall-wart.

[lovekraft0]

I'm confused! The Philbrick only uses 4 triodes, a differential input pair, a gain stage and a follower - what's to simplify? Is it the split high voltage supply you're trying to work around, or is there something else that needs to be addressed? Are you trying to get a linear response, or are you looking for "tube warmth"? Are you trying to make a drop-in replacement for, say, a 741? 

[puretube]

he wants: Ub = 12V

[brad]

Yep, by a "simpler tube opamp" I mean more simple to apply in stompboxes (i.e. unipolar and low voltage), while still keeping the component count down.  In other words, a basic building block that can be used in a hard and soft clippers instead of 741s ect.

[zachary vex]

unrealistic.  look at the characteristic curves.

[Paul Perry (Frostwave)]

The problem is this: if you want an amp to go down to DC (which an op amp has to do) then, you simply can't do it without a negative high voltae supply as well as positive.
In an audio amp, we get around that with interstage coupling caps, that let the signal thru, while allowing one side of the cap to be at plate voltage and the othe to be at grid voltage.

I'm afraid Moby's going to win this one, Captain

[brad]

So it's no good for low voltage audio applications even with a bipolar supply? 

Damn!

[Joe]

Something like that should be fine for distortion, but not without quirks. In transistor land, things like that work as long as there are feedback/output clamp diodes. Since you have an extra 1/2 tube, try using a mu-amp output stage (like Jack's mini-booster but with tubes). Are you actually going to build this?

[Joe]

This would be a hack, but you might be able to simplify it down to a single tube. Just use the grid as the negative input, and the cathode for the positive input. There would be voltage/impedance differences between the inputs, limiting the usefulness to stuff like distortion pedals. This idea stems from me trying to use a single transistor in a psuedo-opamp mode, in hopes of an ultra-simple TS9. It worked to a degree, but the gain/tone pretty much sucked, so it got scrapped.