[time-nuts] Zero-Crossing Detector Design?

[Gerhard Hoffmann]

Am 16.09.2012 21:46, schrieb Richard (Rick) Karlquist:
>
> Again, a really high speed comparator necessarily has a really broad
> bandwidth, meaning its noise bandwidth is large.  This means more
> noise and more jitter than a lower speed comparator.  The comparator
> cited is much faster than necessary for 30 MHz, by orders of magnitude.

I did not suggest it for this 30 MHz system.

And having orders of magnitude more noise bandwidth goes along with
orders of magnitude less time where it contributes noise at all.

Setting rise time / 3dB corners with NP0/PP caps, stable resistors
and close to zero contributions of semiconductors ( in absolute ps)
is much more desirable than leaving it to the mood of a LM139.

Precise fitting of the BW  vs. the  required slope, in multiple steps,
is the essence of O. Collins' paper.

73, Gerhard


ps  in the usenet group sci.electronics.design is a discussion
about ovens with Phil Hobbs. I did inject a pointer to your paper.