VoltageDAC with B1 output buffer

[yggdrasil]

Here's a description of my latest DIY project. A voltage DAC (CS4398) with 4 B1 output buffers (balanced).

Nelson: Thank you very much for this incredible little buffer!

[Nelson Pass]

That's very cool.

[yggdrasil]

I have now made 2 variations of this DAC. I have attached the schematics of one single channel B1 buffer that I use.

The DAC's built are equal down to C9, the coupling capacitor.

Version 1 have 3 cap's in parallell: 100uF electrolytic || 1.1uF Solen PA (metallized polypropylene) || Vishay MKP 1837

Version 2 have 2 cap's in parallell: 1.5uF unknown film cap || Vishay MKP 1837

The sound difference was bigger than expected. I have grown used to version 1. Version 2 were much more forward, the bass was harder and there was less space around the instruments and voices.

All of these film-caps were intentionally bought for passive crossover use. I have never really read any much about capacitor choice in active electronics, so my question here is: Could anyone point me in a direction here?

[Nelson Pass]

Why do you need the input coupling cap? Is there DC at the
output of the DAC?



Nevermind - I see on your previous schematic.

[Zen Mod]

why you need input caps ?

just kill gnd reference for input gates , then place caps on buffer output

well - I suppose that's matter of preference , and type of caps you prefer

edit - I have impression that cap on buffer output is somewhat less critical than cap in input

[yggdrasil]

Originally Posted by Zen Mod

well - I suppose that's matter of preference , and type of caps you prefer

What i prefer? Don't know....

With regards to passive crossover cap's I have found this page helpful: Humble Homemade Hifi

But I have not found any helpful pages with regards to cap's in active electronics. Have you?

[Pano]

The outputs of the DAC sit at about 2.5V, so you need to take that into account. Blocking caps make it easy.
But maybe there is some way to bias the buffer so that it deosn't care about the offset.
I know how to do that with an opamp, but not the B1.

[Zen Mod]

as I said - just delete gate gnd resistors

in that case - made lowpass filter differentially on input , or blahlah on output .

without gate gnd resistors , each upper jfet will float up and sit where it feel comfortably

and - yeah - if you are lacking some drive from DAC , instead of double B1 - make Jfet BOSOZ , with all XCCSwhatever suffixes and prefixes ..... , and make it SUSY , what else ........



there are plenty of options .....

[yggdrasil]

Hi Zen Mod.

I'm afraid I don't follow you. My incompetence level have been reached...

I have ordered some Wima MKP10 for comparison. Any other brands I should try? Subjective opinions appreciated.

[R-K Rønningstad]

Originally Posted by Nelson Pass

Why do you need the input coupling cap? Is there DC at the
output of the DAC?



Nevermind - I see on your previous schematic.

Isn't it so that these DACs run off +5 V (and GND), hence each output is more or less midway between. (2.6 V is almost midway....) But as they have differential outputs (Aout+ and Aout-) all will be hunky dory. Or maybe I should have read the rest of the replies before writing... But I had this cup of damn good coffee in front of me and wanted to type something...