Questions regarding a comparator circuit coupling to an analog switch

Question

This is directly related to my previous question, so it requires to first have a look at that. And the below LTspice implementation is based on Andy Aka's answer. The following circuit works in LTspice. But I have some observations and questions before I solder it on a board.

First of all, I would like to mention that since I couldn't find the SPICE model for DG419 (analog SPDT switch IC), I made an ADG419 from an ADG411. I read they use the same components inside. So I just needed to invert one input and tie two outputs and it works and mimics ADG419.

A 10k potentiometer Rth will set the threshold voltage, and another 10k pot Rset will set the pre-set voltage Vset.

(left-click to enlarge the schematics)

Here is the 0-10V DC sweep:

And here is a transient analysis plot for 0-10V in 1 sec (notice the glitch during the comparator action):

And here is the change at Vth during the comparator action (voltage at the node between R2 and Rth):

My questions are:

1. Is interfacing the comparator LM311 output to the ADG419's input fine this way as in the schematics?
2. Is that glitch the product of the change in Vth in comparator action? And can that be mitigated? (I don't see the glitch in DC sweep)
3. Is there any fundamental problem with this circuit or resistor values? Or any peculiarity. For example I'm not sure whether R4 is necessary? (I found it from a circuit I made years ago so I don't remember why it is there.)

Edit:

Answer 1

1-) Is interfacing the comparator LM311 output to the ADG419's input fine this way as in the schematics?

Nope. Pay more attention to the ADG411 data sheet. Pin 5 is the logic supply voltage, and should be 5 volts, not ground. Then the LM311 output should swing from zero to about 5 volts. You can do this with a voltage divider to +12, or tie your pullup to your 5 volt supply. Likewise, your inverter will need to be powered by 5 volts.

An additional issue is your lack of hysteresis on the LM311 - a real circuit may not switch smoothly as you pass through the transition point.

2-) Is that glitch the product of the change in Vth in comparator action? And can that be mitigated?(I don't see the glitch in DC sweep)

Nope, it's due to the fact that your switching point level is less than the level you use to provide the clamped output. 4.7k/7.7k does not give the same level as 2.2k/4.2k.

3-) Is there any fundamental problem with this circuit or resistor values? Or any peculiarity. For example I'm not sure whether R4 necessary?(I found it from a circuit I made years ago so I don't remember why it is there)

It ought to work, although as stated in 1) you should consider adding some positive feedback (even a 1M will do nicely).

Answer 2

Your switch is rated for 30mA of continuous current, thus I suggest that it might not be the best way to drive a 7805, depending on what you're trying to power.