[Foton]

After thinking about this project I started to have doubts. It looks like I will waste a lot of power with no reason. I do not have 600 Ohms headphones either and this higher supply voltage is for a reason.....
I am slowly getting towards Szekeres then. But, the first step to get parts has been done: I got my transformers last week. So, my new question is.... what is the highest and making sense voltage to power Szekeres? My transformers are 36 Volts @ 1.5 Amps.

 

[Avro_Arrow]

 
I ran mine at about 18 volts

 

[Foton]

On one of the forums, I do not remember which one though, Tomo said the MOSFET ( IRF510?) like to see 20V across. I think, it was in reference to Szekeres. This is what prompted my question.

 

[Avro_Arrow]

 
I had a transformer laying around that put out 18 volts.

Thats the only reason I picked that voltage for mine.
It was just built out of whatever I had laying around.
I didn't even use a regulated supply. That was one
other benefit of the three channel version...better
PSRR.

 

[wakibaki]

Looking at the sims I've done, there's a big problem with the one showing the Szekeres variant, the output is in the microvolt range!!! I don't know how I missed that, just looking too much at the shape of the graphs I guess. Looking at what I've done, I've taken the output at the drain instead of the source.
 
It should be possible to rebias the circuit to run with ~50V, which is what you'll get with 36VAC rectified, you'll have to use something like a BC546 instead of the 2N3904. IRF510 will tolerate 100V
 
I re-jigged the circuit, you can see it's working kind of properly @ 51V, but given my past errors, perhaps this variation needs to be taken with a pinch of salt.. the 50R resistor in this circuit is dissipating 23W.
 

 
I'll have another look at this if you want, but it probably won't be until the day after tomorrow. The circuit is much happier @ 12V, the distortion doesn't go down when you run all that voltage and current, you just get a bigger output voltage swing.
 
w

 

[holland]

  After thinking about this project I started to have doubts. It looks like I will waste a lot of power with no reason. I do not have 600 Ohms headphones either and this higher supply voltage is for a reason.....
I am slowly getting towards Szekeres then. But, the first step to get parts has been done: I got my transformers last week. So, my new question is.... what is the highest and making sense voltage to power Szekeres? My transformers are 36 Volts @ 1.5 Amps.

 
 
You can build the PRR SOL with modifications for lower voltage rails, lower current, etc.
 
Example, drop to about 30V rail, raise the 6ohm resistor to 50ohm or thereabouts.  That should drop the bias to about 100mA.  The resistors along the transistor will drop to about 1W for the 50ohm and 3W for the 120 ohm.  Change the 22M resistor to 15M.  You might be able to go to 24V, but not sure.
 
Just a guess, but that should give you a gain of about 2 or a bit more.  I think it should be able to swing full gain with a source of 2Vrms.  You should be able to drive 600 ohm and 32 ohm headphones without any issues.  Volume might not be loud enough with 600ohms.  Simulate it to be sure.
 
If you want, you can throw in some negative feedback.  Say, a 10K resistor between the input pot and the cap, and a 50K resistor feeding back between the 10K resistor and the cap.  Maybe drop the gate resistor to about 100.  Drop the 150ohm resistor to about 120, and lower the 50ohm resistor to about 20.  Roughly, gain is about 2 with about 6dB of NFB.  Another guess, but should be decent starting points.  Bias might increase a bit, maybe 150mA or so.
 
With Szekeres, put a CCS on it, should work better.  You should be able to go to 12V rails with the Szekeres.
 
Edit: Here's a variant of the SOL, since I mentioned a CCS on the Szekeres.  It's Nelson Pass's Zen amplifier modified to be headphone compliant, using an active current source.  It's basically the SOL with a current source on top. http://headwize.com/?page_id=105