Comparator hysteresis using high resolution input signals

[MRW]

Hi all!

I really don't know how to word this question properly. From reading
various comparator articles, it seems like any talk about hysteresis
mentions a signal being affected by noise in the low millivolt range
(1 mV or so). If I have a 100mV input signal and the trigger point is
at 100 mV, then noise may get the input signal to bounce between 99mV
and 101mV allowing the comparator to trigger on and off.

If an input signal has a resolution of 1mV or 500uV, would adding a
hysteresis to a comparator still make sense if the resistors in the
hysteresis network contribute noise anyway?

Thanks!

 

[Andrew Holme]

Hi all!

I really don't know how to word this question properly. From reading
various comparator articles, it seems like any talk about hysteresis
mentions a signal being affected by noise in the low millivolt range
(1 mV or so). If I have a 100mV input signal and the trigger point is
at 100 mV, then noise may get the input signal to bounce between 99mV
and 101mV allowing the comparator to trigger on and off.

Hysteresis means there are *two* trigger points: one for rising inputs and
one for falling inputs. After the input rises above the higher trigger
point, it has to fall all the way down to below the lower trigger point
before the output will change again. Hysteresis is the margin between the
two trigger points. As long as the hysteresis window is wider than the
peak-to-peak noise amplitude, the output will not chatter.

 

[MRW]

Hysteresis means there are two trigger points: one for rising inputs
and one for falling inputs. After the input rises above the higher
trigger point, it has to fall all the way down to below the lower
trigger point before the output will change again. Hysteresis is the
margin between the two trigger points. As long as the hysteresis
window is wider than the peak-to-peak noise amplitude, the output
will not chatter.

Thanks, Andrew! I guess I'm still confusing the term.


--