ucd and unfiltered deltasigma dac chips

[Alex_J]

The UcD amps are quite insensitive to HF noise. You can use them as post filter for a deltasigma DAC chip (ie. skip the usual op-amp based filter).

At first I want to said The BIG THANKS to the Bruno for his increadible Ucd amps.

I wonder by the possible expluatation of UcD as a post-filter for the D/S DAC (SM5872BS in Marantz CD63mkII) but avoiding some problem with unstability of UcD.

8xfs oversamling (8x DF included in the DAC chip) is 352.8kHz which is near the workin' frequency of UcD (400 kHz as I knew)

Is there any problems out there?

 

[Bruno Putzeys]

wonder by the possible expluatation of UcD as a post-filter for the D/S DAC (SM5872BS in Marantz CD63mkII) but avoiding some problem with unstability of UcD.

8xfs oversamling (8x DF included in the DAC chip) is 352.8kHz which is near the workin' frequency of UcD (400 kHz as I knew)

Trying to get hold of the data sheet but the npc site is down. Whether the images at 8fs are a problem or not depends on the frequency response of the s/c filter that's on the DAC chip. Which is what I'm hoping to find out from the data sheets.

 

[hreith]

SM5872

look at http://www.nacl.de/marantz/

 

[Bruno Putzeys]

Thanx hreith!

This is not a switched cap dac but a PWM dac. If you feed this signal unprocessed into a UcD it'll sync with it and distort (because of reduced loop gain). This is one chip you wouldn't use without a proper reconstruction filter.

(note to a moderator: you might want to move this discussion out of the nuforce thread into one of the ucd threads)

 

[IVX]

Bruno, there is expected 705.6Khz as worst case (Figure 1.Arithmetic operation), maybe some increasing of the C1&C2 (input lowpass UcD400's buffer) will be ok? Or no chance yet?

 

[Bruno Putzeys]

Bruno, there is expected 705.6Khz as worst case (Figure 1.Arithmetic operation), maybe some increasing of the C1&C2 (input lowpass UcD400's buffer) will be ok? Or no chance yet?

I still wouldn't try it. Better candidates are AK4393 and up.

 

[Alex_J]

OK, clear.

So my plans to keep it simpler is radically aborted
Sad.

 

[IVX]

What about R2R NO_OS, e.g. at PCM63 or so (or 4fs OS interesting either)?

 

[Bruno Putzeys]

What about R2R NO_OS, e.g. at PCM63 or so (or 4fs OS interesting either)?

Certainly no "no_os" circuits. The mirrors are wholly unattenuated and there's a limit to the HF energy you can feed any amp without eliciting problems. Also, when I talk about "unfiltered" deltasigma chips, keep in mind it's still got a switched-cap filter inside, so unfiltered is somewhat exaggerated.