[time-nuts] Phase noise & Jitter
bruce.griffiths at xtra.co.nz
Tue Apr 29 05:57:24 EDT 2008
Hal Murray wrote:
>> For digital circuits, it would seem that the potential for
>> metastability & threshold uncertainty, is reduced when the
>> slope-to-noise ratio is optimized. I wonder if this concept is, or can
>> be quantified?
> I don't think noise hurts or helps metastability. For every case where it
> might hurt there is the mirror image where it helps.
> I would expect a slower rise time to cause more metastability. Here is the
> way I think about it. If you have N ns to settle, there is a window in
> voltage that is Y mV wide. (give or take some statistics) The signal with
> the faster rise time will spend less time inside that window.
> There is probably some limit where making it faster doesn't help any more.
> Speaking of rise time...
> I'm somewhat looking for a signal generator - digital rather than sine wave.
> I'm pretty sure I used an HP box 10-20 years ago that had knobs on the rise
> and fall times as well as frequency and whatever. Is there some obvious unit
> is should be looking for? Do they show up on eBay? ...
Perhaps an HP8082A, but it all depends on what range of risetimes etc
There's one currently listed on ebay.
More information about the time-nuts