I am choosing a comparator for a clock recovery (frequency about 5...20 MHz); I
liked the Analog Devices's AD8611 and AD8561 specifications, but I would really
appreciate any information about the jitter additions I could expect if I
choose this device. Unfortunately, I did not found any information regarding
this question through the ADI's Web site. Could i find any information about
the AD8611 (I also inquire about the AD8561) jitter and related noise sources?
I am designing a mixed-signal PCB and jitter is rather important for my design.
It is difficult to arrive at a specification for comparator jitter which is
generally useable, because the amount of jitter occurring at the comparator
output depends on various factors related to the input signal and the circuit
This questions frequencly arises in DDS systems when a comparator is used to
generate a square-wave output based on the filtered sinusoidal output from the
In this case, there are two main contributors to the jitter of the output
- the spectral purity of the input analogue signal (i.e. harmonics, spurious
- the noise and random variation in propagation delay present in the comparator
The jitter is measured experimentally with a particular input frequency
(sinusoidal input) and amplitude.
One specification you can use to crudely compare different comparators' jitter
performance is propagation delay dispersion. This is defined as the variation
in propagation delay with different input overdrive voltages. The jitter occurs
at the instant when the input switches, and a low value of propagation delay
dispersion usually indicates a low-jitter input stage.
I think the best way to determine the jitter of the AD8611 and AD8561 is to
order a couple of samples from our website and evaluate them in the
application. To get the best jitter performance
- filter the input to minimise spurious signal and noise
- ensure low impedance ground plane
- ensure clean power supplies
- minimise parasitics (R, L,C) in the layout