Aleph-X, higher diff pair bias?

[acaudio]

Hi people,

I do at present the first tests with Aleph-X . I build for first only one. This works well, 12FET´s, 7A Bias, 22V rails, ~51% CCS 100W/8 153W/4 72/2 Ohm. 4V abs. offset cold, ~50mV after warm up, diff. offset under 10mV in all time. I have modified the CCS, see pic, I removed V1 and R11 (G.Rollins schematic) und built these as voltage divider at base NPN at CCS. I use 0,66 Ohm source resistors, this helps by current sharing and I can set the bias up, not down. This works well.
I have a question:
Has someone experience with higher bias of the front end /diff pair? I use at present 50mA, with 221 Ohm by diff pair and ~100 Ohm by CCS.

All Aleph´s have only ~22mA with 392Ohm diff pair source resistor and 221 Ohm at CCS, but the pictures from XA160 http://www.diyaudio.com/forums/showthread.php?postid=59132#post59132 shows diffpair and ccs with nice heatsink. Runs these with higher current? My heatsink heat up to ~30°C.
Normally has Fets less distortion with higher current. Is this here too? I found this here, but I´m no sure: http://www.diyaudio.com/forums/showthread.php?postid=112816#post112816

I haven´t a distortion analyser to measure this, only function generator and oscilloscope.

Thanks in advance for your opinion.

Sorry for my bad english

Adam

 

[acaudio]

a pic from ax (test)

 

[Nelson Pass]

There is an optimal current for the diff pairs, and it has a lot
to do with the gain of the amp, and so on, but it also relates
closely to the transconductance and capacitance of the output
stage.

I have built some prototype XA160's and 200's with IRF250
devices, which have much more transconductance and
junction capacitance. To compensate, I doubled the bias
through the diff pairs, reduced their load resistance, and the
result was very good.

Generally speaking the value we choose is decided based on
the measurements and the sound, with the sound getting the
last word.

 

[acaudio]

There is an optimal current for the diff pairs, and it has a lot
to do with the gain of the amp, and so on, but it also relates
closely to the transconductance and capacitance of the output
stage.

I have built some prototype XA160's and 200's with IRF250
devices, which have much more transconductance and
junction capacitance. To compensate, I doubled the bias
through the diff pairs, reduced their load resistance, and the
result was very good.

Generally speaking the value we choose is decided based on
the measurements and the sound, with the sound getting the
last word.

Mr.Pass, thank You very much!

I forgot to say, I use 12x IRFP254, these have too a much higher input capacitance than IRFP240 and are very similar with IRFP250 (in my opinion), and my thought was the same,
more capacitance->more current.
The sound of this one is already very,very good. Best dynamics, no noise and good bass. What do you think about my CCS (without an DC offset)?

Regards
Adam