Looking for the bestest UcD buffer

[IVX]

looking for the bestest UcD buffer
In fact the UcD offers superior sonic transparency vs ordinary amp , so the reason to have finest buffer is does exist.
Well.. I want to discuss two ways (or more):]

A) Fully differential (THS4131 like) discrete opamp with class A (SE) mosfet output stage- simulations show THD around several micro percents, which have slightly frequency dependance- 10k/1k about 2.


B) i've done simulations for the simplest tube buffer for the UcD- THD about 0.003% (it's for the 100v rails, for the >100v it will much lower) absolutely independent of the frequency.

[IVX]

B) picture:






Any comments are welcome.

[IVX]

Ok, Jan-Peter, PCB is done, el. caps isn't SMD is suitable for BG using or so..TOP layer is ground plane only, and aren't necessary. Last simulated version show 0.0006% THD@ 1Khz equal 10Khz@4Vp-p@voltage gain=4@2Khom load (see attachment).

[IVX]

That's PCB 41x31mm.

[tiki]

Hi Ivan,

good work! And for two channels already.

It is not easy to find out, which components are located on top or bottom side. Where are the vias?
If you would shift Q2 a bit to the right, you could place Q1 left besides Q2. C1 and R1 would move a bit up. You don't need the red GND connection between the rail caps really and could place C1 there instead. May be easier for the wiring, to have the outputs not under the rail caps.
These are my thoughts only. Thank you for your work!
If you'd like to produce the PCBs (PCB-pool?), I'd like to buy two of them for sharing costs.

Regards,
Timo

[IVX]

Well, some measurement of the tube buffer is ready. 0.037% THD@1Khz@2Kohm load@~4Vp-p, but voltage gain=6.5 instead 4(simulated), also anodes resistors, which was adjusted for the minimal THD in reality, much less.(about 11Kohm first and 12Kohm second, if both 12Kohm then THD around 0.8%) My measurement and setup aren't precision action generally.. 100hz hum at -60db even, and no any AP just cheap audigy. However one point is clear- the difference of the emission degradation will require regular balancing of the triodes. Quite sad.. i going to knock down my setup.

[IVX]

Ok, Timo,
Good work is never be if just 2 hours spend for it. 10:30am i had wakeup and 12:00am must be starting my day job, BTW today I delayed for 50 minutes. ;-)
New version PCB done, but one of the decouple cap is out of range, and .33mm lines more now. New dimension are 40x25mm.

[stef1777]

Great works!


Ideas!

It can be better to design a "mono" PCB to fit directly it on the UcB module.

We remove the op-amp, the 22UF caps and use the free holes and free space to plug the small pcb directly on the UcD module. We use the UcD PSU and try to still use the mute circuit integrated to the UcD module.

With this solution, no cables, better integration and better SN.

[IVX]

stef1777, your idea is cool too much! :-)
Yeah, imagine- JP will make only this version UcD right now, or from the next week ;-) Actually, this is just a virtual project and nothing more, at least before hearing by advanced listeners, i'm not sure that ad8620 or opa627 is worse..

[stef1777]

Originally posted by IVX
stef1777, your idea is cool too much! :-)
Yeah, imagine- JP will make only this version UcD right now, or from the next week ;-) Actually, this is just a virtual project and nothing more, at least before hearing by advanced listeners, i'm not sure that ad8620 or opa627 is worse..

Don't forget the OPA2134.

Very smooth sound with it.