I'm trying to build a circuit to control a contactor that brings online more available power during times of supply droop from our existing source. I decided that an asymmetric inverted Schmitt trigger might do the trick, but am having trouble selecting values for a pretty basic resistor network feeding back into the non-inverting Vref side (for Vupper and Vlower thresholds.)

If for example I wanted a Vref of 24V, a Vupper of 23V and Vlower of 20V, could someone provide formulas for solving for the resistor network (calculating voltage at node A) in the attached graphic? I do not understand how they derived the R.eq of this network in the example I'm referencing.

Thanks,
M

 

I could be way off base, but that looks to me like it's showing a comparator (or op amp) with externally set hysteresis, not what I normally think of as a Schmitt trigger.

I'm still a relative novice, and I first learned about hysteresis and how to set trip points from one especially good datasheet. Here are a couple screenshots from the MCP6542 datasheet, showing how to set high/low trip points (and therefore also setting the hysteresis.) For the example you showed, I think you'll need the formulas in "Equation 4-2" in the inverting circuit section.


The full datasheet is here:
http://ww1.microchip.com/downloads/en/DeviceDoc/21696H.pdf

 

Cool. Thanks for the info. In the literature it seems to transform the resistor network to a Vthevenin equivalent and proceed from there. Useful either way. I am apparently a lot worse than I thought at characterizing the node voltage there -- wasn't quite remembering my formulas.

Not sure about the Schmitt bit. From my understanding it is one, with the inversion controlled by whether the feedback is applied to reference voltage (inverted) or input voltage (for non-inverted).

My problem right now is finding one of these spec'ed in non-CMOS / non-TTL logic levels! Particularly 12-36V

 

Basically a Schmitt trigger is any digital/comparator circuit that has hysteresis in its input switching point, so whether the hysteresis is done internally or externally makes no difference in that designation.

 

Basically a Schmitt trigger is any digital/comparator circuit that has hysteresis in its input switching point, so whether the hysteresis is done internally or externally makes no difference in that designation.

Thanks for the clarification. I was a little off on my understanding there.

 

What's the practical approach for a 24V system? Dropping Vref and Vin through a volage divider prior to?

 

I'm trying to build a circuit to control a contactor that brings online more available power during times of supply droop from our existing source. I decided that an asymmetric inverted Schmitt trigger might do the trick, but am having trouble selecting values for a pretty basic resistor network feeding back into the non-inverting Vref side (for Vupper and Vlower thresholds.)

If for example I wanted a Vref of 24V, a Vupper of 23V and Vlower of 20V, could someone provide formulas for solving for the resistor network (calculating voltage at node A) in the attached graphic? I do not understand how they derived the R.eq of this network in the example I'm referencing.

Thanks,
M

You only need to calculate the voltage at the non-inverting input for each output voltage. If you have a rail-to-rail opamp that isn't loaded too much, then the output voltages are the supply rails. Otherwise, figure out what Vmax and Vmin are an use those.

What you'll get it is that the thresholds are a linear combination of the output and reference voltages with the resistors setting the scaling coefficients. Depending on what you want, you can pretty easily transform this into the center of the hysteresis window and the width (or half-width) of it.

 

If you do the analysis, what you will discover (if I did it right, which I think I did) is that if the comparator has symmetric bipolar output swing then the center of the hysteresis window is controlled by Vref while the width is controlled by the voltage swing. As for the resistors, by choosing R3 to be sufficiently smaller than R1 and R2, R2 can be used to adjust the width while having limited impact on the center. Similarly, R1 can be used to adjust the center while having limited impact on the width. But choosing R3 small enough to make that happen limits the range of either, which may or may not be an issue.

Asking for thresholds of 20 V and 23 V with Vref being 24 V is going to be tricky.

Without knowing what the limits on the opamp output voltages are, I can't tell you any more than that.