2025-08-03 03:51 UTC
@kenshirriff What's the point of the noise source? Don't you naturally read noise in between channels, is that not sufficient? I'd expect the VCO to be slewing around randomly due to an unlocked control loop.
WRT all the crazy single layer routing chips and exotic BJT layouts, I don't know how you do it. The oldest thing I've done any significant RE on was 350nm CMOS and most of what I've done lately has been 180 to 45.
It's sooo much more readable when everything is standard cells, all your NMOS in a nice row, all your PMOS in a nice row, all of your wiring mostly on orthogonal axes, all of your layers mostly planar, etc. Sure you need a SEM and fancier delayering equipment but once you have the data it's so much easier to make sense of.
Replies (2)
-
@tom_verbeure@mastodon.social 2025-08-03 04:03
@azonenberg @kenshirriff Comfort noise?
-
@kenshirriff@oldbytes.space 2025-08-03 14:30
@azonenberg From the application note, I think the main issue is if the VCO locks onto a sideband, then you get distorted audio. It's better to replace this with white noise so the user knows that the tuning is off.