Yet another Szekeres variant - good option for a balanced build?

I've been looking for an elegant (read: simple) Class A line-buffer design to build into a balanced headphone driver to match to my 1212m balanced source. The 1212m outputs plenty hot--and has a pro-level +4dBu output level so a line-driver should suffice for my HD555's. Just found this, the "Szekeres VE:"

RJM Audio - Szekeres VE Headphone Buffer

Would two of these work in a balanced build, substituting V+ and V- for L and R on each side?

I'm planning on playing around with a Mogami recable for my 555's, and have really been wanting to "play" around with balanced setups, anyway.

Thanks! I was kind of wondering about that.

I actually just wired up one channel of the original version...but I think I fried the IRF513 at some point when I was firing it up. *sigh* ...back to Radio Shack for another Mosfet.

I'm very interested in trying something like this, but while my basic electrical knowledge is reasonable, my practical application knowledge is fairly sophomoric. I can tell you what individual electronic components do and even to some extent how they actually work, but I can't explain how to put them together--so designing a balanced Class-A line-driver from scratch is a little bit beyond me. I need to be able to analyze other designs to learn more...and I figured adapting a Szekeres or one of it's variants would be a good departure point since it's a relatively simple design.

Problem is, it seems few people have really experimented with simple (but truly double-ended) balanced line-driver or amplifier designs. I've found a lot of circuits that show how you split a signal and invert it (like yours), or else convert a balanced signal into an unbalanced signal, but nothing that's designed truly balanced end-to-end.

Thanks for the link--this is more the kind of info. that would help me work out a basic design...though I need more!

One of the things that has worried me about "cobbling together" a balanced build, as you say, is how to keep the two phases perfectly mirrored through the circuit. I could totally see myself building a Szekeres, liking how it sounds, and then adding the stages for the - phases and having it sound like crap because the phases don't mirror correctly due to non-equal signal changes introduced from non-matched components or tolerance issues. Is this that much of an issue? Do truly balanced circuits utilize some kind of feedback to keep the phases perfectly matched? Maybe that's what you're trying to show me with your neat DAC project circuit?

EDIT: Well, I did run across this design--the study of which was very illuminating. Unfortunately the complexity of adapting it to drive headphone impedances, fine-tuning it so it would work "right," and biasing it into class A (if possible) would prove to be beyond my limited capabilities, I'm afraid.