Hysteresis For A Window Comparator

[741]

When out=HIGH, want lower comparator's '-' input (TripL) to go lower. This means the output transistor is 'OFF' so the pull-up has a chance of working.
But taking TripL low ALSO pulls TripH low. This could cause a valid input near to TripH to appear invalid, causing the upper comparator's transistor to turn on.

The output is effectively "wired-NOR", with 00-->High for valid input inside the thresholds.

TripL < U   ---> Lower comparator sees valid input
"OR"
U < TripH   ---> Upper comparator sees valid input

So if out=HIGH, we want hyteresis to pull TripH 'upwards', and TripL to move downwards.

Is there a neat, simple solution?

[Zero999]

https://www.eevblog.com/forum/beginners/window-comparator-hysteresis/msg575691/#msg575691

[741]

That is interesting - but I think it assumes the existence of nice low-impedance voltage references?

It looks as if VRefh must be 'solid' in the face of varying feedback via Rf1. Or, possibly one could calculate the effect and allow for it.

So long as I do not need 2x op-amp buffers to supply references.

[Zero999]

That is interesting - but I think it assumes the existence of nice low-impedance voltage references?

It looks as if VRefh must be 'solid' in the face of varying feedback via Rf1. Or, possibly one could calculate the effect and allow for it.

So long as I do not need 2x op-amp buffers to supply references.

How much current can the references source and sink?

Isink = (Vcc-VrefH)/(R1+Rf1+R2)
Isource = VrefH/(R2+Rf1)

Make Rf1 and R2 as big as you can and it won't be an issue.