LINEUP Input-VAS, idea

[ontoaba]

For better thermal, R8, R9, R10 could be split into four. See my failed VL83 schematic that I posted about few month ago at "Its 2010 what your favourite VAS" or something like that, also attached at my VL85 cheapamp my lonely thread.

The center pic at first post VAS is very good and very low gain if Re is split into two with input from center. For facing my crazy supply ripple, using crossed current feedback input with only 1.6x gain, it work great.

[lineup]

A nice example. All JFET.
All 2SK170BL, 2SJ74BL runs 6 mA. Output single end, too.
Gain is 1 = follower.
Level for test is +/-4 Vpeak, 1 kHz.

All 4 JFET in VAS runs with 12 Volt across each.
THD is very low.
2nd at -131dB
3rd at -138dB

4 caps for compensation gives a nice AC analysis.
See image 2.
-3dB at around 5 MHz

Enjoy

[destroyer X]

But will they sound fine?.... only building to listen and compare to a reference and test it using non biased evaluators.

Looks great..no doubts....but there's a lot of things that looks great but are not great.... sometimes sound ordinary as Hugh told.... well..he can say things about as he built.

regards,

Carlos

[kenpeter]

What exactly was the purpose of redundant stage U1(Active)-U2(CCS)?
Weren't U4(Active)-U5(Active) already able to swing twice that much current?

[peranders]

The stage is an output buffer. U4, U5 has a very high output impedance. U1 and U2 are both essential here.

[MiiB]

Can it work floting in the center like that...??

[thorstenlarsen]

Hi Lineup!
As far as I can see, you wont get double gain by using complement pairs. Look at it this way: One transistor will need a certain drive to pass a certain current. The other transistor will need an equal amount of drive just to pass the same current. ( which it will have to , since they are in series. )
You might get some linearization of the FETs though.
Thorsten Larsen

[kenpeter]

Originally Posted by peranders

The stage is an output buffer. U4, U5 has a very high output impedance. U1 and U2 are both essential here.

U4, U5 has very LOW output impedance if you were to close the loop here.
And twice the potential current swing into a real load as U1 + CCS.

U1 and U2 are absolutely redundant and non-essential if they aren't going
to slew any more current than the stage that drives them.

[artu]

Originally Posted by kenpeter

U4, U5 has very LOW output impedance if you were to close the loop here.
And twice the potential current swing into a real load as U1 + CCS.

U1 and U2 are absolutely redundant and non-essential if they aren't going
to slew any more current than the stage that drives them.

I don't agree, if you do so, you will degrade OLG, rise the output impedance and rise the distortion significantly. With the same OPS, slew depends to a great extent on how weak or hard is the drive, that means how high or low is the output impedance of the driver.

Cheers
Arturo

[kenpeter]

OLG (voltage) is not increased by a source follower. Go figure...
And unfortunately neither is this specific example giving us any
current gain...

I agree about slew, thats why the driver stage is better here:
His output stage abusing same JFET at exactly same current.
Yet pulling SE against fixed CCS, is only half as capable as the
push-pull circuit that is driving it. Defeats its own purpose...