Sons of VHex

[vzaichenko]

A dual fet 2SK146GR as input fet.
Nice balance.
I think that an offset controlling trimmer is no more necessary.

Looks very good. Thanks a lot!

 

[thimios]

Looks very good. Thanks a lot!

 

[thimios]

Sons of Vhex ver.1.2 2sk146GR

Another surprice.
Picture 1 version with unmatched 2sk117
Picture 2 version with dual fet

 

[vzaichenko]

Onother surprice.
Picture 1 version with unmatched 2sk117
Picture 2 version with dual fet

Better overall DC balance leads to lower distortion, especially below 1KHz.
Interesting result One more proof that matching really makes sense, especially in the input LTP.

 

[thimios]

Better overall DC balance leads to lower distortion, especially below 1KHz.
Interesting result One more proof that matching really makes sense, especially in the input LTP.

The other interesting find is that matching input fet ONLY is enough for 0V d.c offset IN THE MIDDLE OF TRIMMER.

 

[jwilhelm]

I've been doing a quick match with a cheap meter for all the VHex input transistors. I rarely need to adjust DC offset at all. They all seem to run nicely within 3mV.

 

[mlloyd1]

i'm guessing the 2sk146 having a lot more transconductance at the operating current probably helps the distortion too.
Folded cascode is helping mitigate distortion increase as frequency increases, in spite of much larger capacitance of 2sk146. It's a large geometry part, famous for large gain, very low noise and high cost, when you can even find them.


mlloyd1

 

[vzaichenko]

The other interesting find is that matching input fet ONLY is enough for 0V d.c offset IN THE MIDDLE OF TRIMMER.

Yes - this is cool.
It's a pity that many good pairs are not manufactured any more.
2SK389 for example.

 

[thimios]

Sons of Vhex ver.1.2 2sk146GR

Scope test

 

[vzaichenko]

Scope test

Wow )) Better than my simulations
Fast enough. Clipping is excellent

 

[vzaichenko]

It's better to see the step response at the higher amplitudes.
It will show some 40-50V/uS.

 

[thimios]

It's better to see the step response at the higher amplitudes.
It will show some 40-50V/uS.

Do you mean ARTA STEPS at the higher amplitudes,or the rise time measurements at a higher level?

 

[vzaichenko]

Do you mean ARTA STEPS at the higher amplitudes,or the rise time measurements at a higher level?

I actually mean this kind of picture for the rise/fall time. If you can see it at higher amplitude - like 50Vpp or so - you will get more precise measurement.
Look at your 20KHz square wave - it makes around 35V in less than 1uS.

 

[rsavas]

It's a pity that many good pairs are not manufactured any more.
2SK389 for example

Do you not think the lsk389 is not equivalent? and now they offer dual p channel lsj689 to go along with lsk489, expensive but available.
Bob Cordell did a talk, (utube video avail) at Burning Amp on his app using a symetrical push pull IPS & VAS... lateral mos ops, so many designs, joining Terry's club

 

[vzaichenko]

Do you not think the lsk389 is not equivalent? and now they offer dual p channel lsj689 to go along with lsk489, expensive but available.
Bob Cordell did a talk, (utube video avail) at Burning Amp on his app using a symetrical push pull IPS & VAS... lateral mos ops, so many designs, joining Terry's club

Ah! You're right - I should try these ones.
So far we're using 2SK246/2SJ103 for symmetric IPS, but they are also getting hard find.

 

[thimios]

I actually mean this kind of picture for the rise/fall time. If you can see it at higher amplitude - like 50Vpp or so - you will get more precise measurement.
Look at your 20KHz square wave - it makes around 35V in less than 1uS.

Ok.this will be a job for tomorrow

 

[thimios]

Sons of Vhex ver.1.2 2sk146GR

Today Sons tested using capacitance multiplier for input and vas stage but i can't see any further improvement.

 

[vzaichenko]

PSRR is good enough
It's like trying to zip a jpeg - it doesn't compress any more

 

[thimios]

PSRR is good enough
It's like trying to zip a jpeg - it doesn't compress any more

 

[thimios]

I see you post these graphs a lot. May I ask what you see in these and how they help you determine whether or not an amp is working well or not? What are you looking for when you look at them?

Thanks, Terry

Terry,look these pictures.
It is from a commercial amplifier to be compared with the sons of vhex