DIY Sony VFET pt 1

Anyone already working on a tube front end yet?

Something like this?

attachment.php


dave

Nelson, hope it is alright to use your X?
 

Attachments

  • VFET-Evo-X-concept.png
    VFET-Evo-X-concept.png
    11.1 KB · Views: 1,036
I think the VFET is a design that already has a “tube like” sound quality already, adding a tube front end is perhaps like a bit too much sweetener in your favorite recipe. Also, making it a BTL (“x” or balanced) output amp will cancel out even order distortion and make it more odd-order dominant distortion. If that’s the sound one likes, I guess it’s fine.
 
However, will not placing such a large electrolytic capacitor in the output signal path negatively offset some linearity this transistor is known for? Again, I'm assuming this is to keep costs down, as designing a push pull class A version using these devices would easily triple the cost.

Not looking to start a debate, Just wanting to get some feedback from the group regarding this.


The short answer is “because the amp uses a single-rail power supply.”

It doesn’t affect the linearity of the Vfet at all.

This is great information -

SYclotron Audio | Why are people obsessed with coupling caps?
 
I think the VFET is a design that already has a “tube like” sound quality already

I hate the term "tube-like". I take it to mean a badly designed tube circuit. A good triode is one of the most linear amplification devices we have. A well designed stage should be very neutral, have lots of swing, and high DDR.

No "tube-like" please.

dave
 
In my mind, “tube like” means dominant second harmonic, a bit of third, not much else and THD around 1%. I think the VFET amp is sort of described by this. So adding a SET tube front end will just be adding more THD and more H2 and H3 and it may be laying it on a bit too much.

I don’t hear anyone else referring to “DDR” of a tube amp. If you mean resolving detail and having low distortion, a state of the art solid state opamp like OPA1656 will be tough to beat.

There is the nano vacuum channel transistor (NVCT) (a filament-less SET that is like a vacuum tube but operates by having electrons fly across a 40nm gap (smaller than mean free path) so that even ambient pressure is considered a “vacuum”). The NVCT can operate up to 400GHz so is very high speed. Unfortunately, still researchy “one of a kind” prototypes.

Nanoscale vacuum channel transistors fabricated on silicon carbide wafers | Nature Electronics
 
Last edited:
"In my mind, “tube like” means dominant second harmonic, a bit of third, not much else and THD around 1%"

this is a perfectly fine description of "tube like", I like it.

Planet10- you criticize "tube like" then use a term like "DDR", hmm...
 
I hate the term "tube-like". I take it to mean a badly designed tube circuit. A good triode is one of the most linear amplification devices we have. A well designed stage should be very neutral, have lots of swing, and high DDR.

No "tube-like" please.

dave

agree 100%

OK, it is not outrageosly linear ...... I mean - with care it is

point is - when I'm content with tube stage I made, nobody could tell is it tube or something else, if not being able to peek under the lid
 
....

After 10 days of digesting I got to admit defeat. My understanding is that a P-channel common drain square law device loaded with a current source should have a negative phase H2, but Papa said in diy_part_1 the J28 was doing positive. 😕

see- that's why I'm not even bothering with phase

as Homer would say, after head is starting to ache - "hm....... Beer!"

:clown:
 
Folks:

A dumb question, perhaps, but how much capacitance would you recommend for a linear power supply for this VFET amp? I am contemplating a CLC or CLCRC supply and was imagining something on the order of 100,000 uF, but I have no experience with VFETs and limited experience with low-powered amps.

Thanks for the input!

Regards,
Scott