[time-nuts] What is a Time-Nut grade Zero Crossing Circuit?

[Bruce Griffiths]

Bob Paddock wrote:
>>  Jitter specs assume a logic
>> waveform input, not a sine wave input.  Many jitter specs refer to
>> pattern jitter of data, which does not apply to clocks.  Also, jitter
>> increases at low frequencies in practice, even though in theory it
>> should not.  Like I said, this topic is very tricky.
>>     
>
> How exactly does zero jitter?  I understand that the logic detecting
> the zero voltage point of a sine wave might not be perfect.  However
> if the circuit is truly a Zero Crossing Detector, things like
> the frequency and amplitude variations of the sine wave are irrelevant,
> as long as the bandwidth is sufficient to the design of the detector.
>
>   
Real circuits have noise.
Thus with a signal having a finite slew rate through the zero crossing 
any noise (internal or external) will cause the switching delay (from 
the noise free zero crossing) to differ for each zero crossing.

The noise increases with circuit bandwidth increasing the noise.
There is an optimum bandwidth which minimises the ratio of the noise to 
zero crossing slew rate and hence the jitter.
Smaller or larger bandwidths increase the jitter.

Sinewave amplitude variations alter the zero crossing slew rate and 
hence the jitter.

There is a minimum input slew rate required to minimise the jitter at 
the output of a logic gate.
When the zero crossing slew rate is less than this a slope amplifier 
consisting of a cascade of amplifier limiters each with a filtered 
output is used to amplify the slope.

The gain and filter bandwidth for each stage are optimised to meet the 
required slope gain whilst minimising the output noise.
There is also an optimum number of gain stages for minimum noise.

Bruce