Page 1 of 1

Rationale for small offsets in output?

Posted: Thu Sep 25, 2014 9:03 am
by Digital Larry
For example, here are the last few instructions of the GA_DEMO_WAH.spn program:

Code: Select all

sof	1,0.02
wrax	dacl,1
sof	1,-0.04
wrax	dacr,0
DACL is shifted up by 0.02, then DACR is shifted down by -0.0008.

As far as I can tell, this puts a very small DC offset in the output. I can't understand why you would want to do this. I thought that perhaps it was to defeat limit cycles, but I don't think there is any recursive path in the given algorithm that would be affected by it.

Posted: Fri Sep 26, 2014 9:13 am
by frank
Good question, I wish Keith was still around to ask him. It may have to do with the first silicon of the FV-1, there was a minor birdie issue and small DC offsets can remove them. The DAC was fixed prior to release so the birdies went away but we forgot to remove the offset from the code. Won't hurt anything to leave it.

Posted: Fri Sep 26, 2014 10:04 am
by Digital Larry
Works for me Frank. Thanks!

I wish Keith were still around too, and I didn't even know him.

DL