As I was designing a split voltage supply for an opamp supply, I thought that cutting one rail if the other rail wasn't at it's nominal voltage would be a good idea if I wanted a "good" output signal. Turns out it isn't that useful (at least in my particular case). BUT I'm still wondering how to do that. So for my curiosity:

How would one do that?

I have a split voltage (in my case it was +/-30 V and a load of about 25 mA. Although this post is hypothetical so not sure it's useful).

My idea was to use some comparator and the output would activate or deactivate a switch. Maybe a BJT or a MOSFET with a MOSFET driver. If I use a comparator like the LT1017, in it's datasheet the absolute maximum rating for the input voltage is -0.3 V to 40 V. So how can I monitor the negative rail with that? In one application example, V+ and V- are "switched". But can I use a negative Vref on INA- or INA+? And if it's the case, can the output of the negative rail comparator drive a switch (MOSFET or BJT)? Or MOSFET driver?

LT1017 datasheet

  • 1
    \$\begingroup\$ not a good idea; why do you think it's a good idea? \$\endgroup\$ Apr 26, 2021 at 8:16
  • \$\begingroup\$ @MarcusMüller In case the persone doesn't realize there's someting wrong with one of the rails. My question is more general than if it was used with an opamp. \$\endgroup\$
    – Neeko
    Apr 26, 2021 at 8:18
  • 1
    \$\begingroup\$ You can simplify your problem if you realise you don't need to monitor both rails individually. A +/- 30V supply should have 60V between the rails, if any of both rails dips then there won't be 60V between them anymore. One could probably come up with something simple involving a single TL431... \$\endgroup\$ Apr 26, 2021 at 10:35
  • \$\begingroup\$ @Neeko a) this feels like a very unlikely problem. Power rail supervision is something you do by building voltage regulators, then you don't need additional protection (you're not gaining safety, not complexity, that way). b) I think when you're thinking about this, you really should be more used to the concept of looking at potential differences, and about things like voltage dividers; this doesn't look all that complex to me, but I might be missing something. \$\endgroup\$ Apr 26, 2021 at 11:14
  • \$\begingroup\$ @MarcusMüller a) ok I see. I did have a regulator before. LM317. So above my Vcc the voltage was regulated but not below. That's why I was first looking for somthing that can mitigate the "below Vcc" case. b) I don"t undersant what you're trying to say (I do know what a voltage divider) \$\endgroup\$
    – Neeko
    Apr 26, 2021 at 12:15

2 Answers 2


How about something like this?

When the voltage difference between the rails is lower then the zener voltage, no current can flow. This means no voltage drop across R1/R2, keeping both transistors off. Otherwise current can flow trough the zener, when the drop across R1/R2 is larger then 0.7V both transistors turn on. R3 limits base current for both transistors.


simulate this circuit – Schematic created using CircuitLab

If the zener knee isn't sharp enough for your purposes one could replace it with a TLV431 configured for the desired voltage.

  • \$\begingroup\$ I tried this but I didn't get it to work. \$\endgroup\$
    – Neeko
    Apr 27, 2021 at 9:58
  • \$\begingroup\$ @Neeko I tried it on breadboard and it works for me ... \$\endgroup\$ Apr 27, 2021 at 14:57
  • \$\begingroup\$ oh... Maybe it's because I tried to make it wotk with the rest of the circuit I had, jst to see how it would have been behavoring. Thanx for your help! Regarding my post, you answered my question. \$\endgroup\$
    – Neeko
    Apr 27, 2021 at 15:09

Linear Technology (now a part of Analog Devices) has a ton of voltage monitors and hot-swap controllers that can handle positive and negative voltage combinations, have on-board MOSFET drivers, etc.

There is a catch-22 (metastability) issue with the task, that is solved by delays. Since at power-up no two supplies come up at exactly the same moment, there needs to be a power-on delay before the comparator outputs are taken seriously.

Next, if you want the system to restart automatically when both supplies are back within range, there needs to be a way to power the monitor circuit during the unbalanced period.

Last, there probably should be a short delay when power is back to make sure it is there to stay. My home generator waits 1 minute, before switching back to grid power. That probably is a bit long for a small electronic system, but only you can say.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.