Design advice / help for battery low voltage cut off

Thread Starter


Joined Apr 14, 2021

I'm very much a complete amateur at circuit design, but find it fascinating. I am attempting to come up with a simple circuit for a low voltage battery cutoff. I've found several designs that exist (and would work fine, honestly), but haven't found one that is exactly ideal and it's bugging me that I can't seem to figure it out.

In short, I plan to use Dewalt 20v batteries to power a few different things as I have a ton of the batteries and chargers and this would be convenient. The Dewalt batteries contain no BMS of any kind as that circuitry resides in each tool / charger, so it's just a basic 5s Lipo. I do not want to use an off the shelf BMS board as I am adding a few different connections, 5v USB charging, etc. and would like to be able to have it all on a single veroboard or PCB for simplicity of mounting in a 3d printed case that attaches to the batteries.

My criteria for the low voltage cutoff are as follows:
  • As simple as possible.
  • Operates up to 22v and trips at ~16v.
  • Low power draw while on.
  • Deals with the potential hysteresis.
  • Don't care if it utilizes a single momentary button, individual on/off buttons, or a switch.
  • Ideally zero power draw once the low voltage cutoff trips, or extremely low at the very least.

The last requirement is the one that has been giving me trouble. A perfect circuit would trip, and then stop drawing any current at all. It's extremely unlikely that the low voltage would trip, then be left attached to the battery for months, but not inconceivable. It's certainly overkill, but seems to be an interesting design challenge.

A few of the solutions I've found and played with on a bread board:

Dave Jones soft latching power switch circuit - This is very clever, but in practice the momentary switch is a bit finnicky and you have to press it for the "right" amount of time to turn on or off depending on your resistor/capacitor choice. Also, it uses standard NPN BJTs which all seem to have an emitter to base voltage limit of ~6v.

David Watts LVCO - builds on Dave's soft latching circuit, but integrates a TC54 voltage detector for the LVCO portion. Same issues as above.

There exist many 555 timer and TL431 / TLV431 based circuits that I've found on this forum and elsewhere that are effective, but in each case there ends up being one or several paths from the positive of the battery though a few resistors to ground which creates some current draw when off.

So given all that information, my questions come down to:
  1. Is it possible/reasonable to design a circuit that draws no current once it trips but doesn't have the odd single pushbutton behavior? [Again, being fully aware that a draw of a few mA isn't really a big concern, but curious if it's possible.]
  2. What's the best way to deal with the 6v limitation of standard BJTs? Using a voltage divider for each seems awkward, using FETs is much more expensive, is there a simpler solution?

I've got a variety of BJTs, FETs, relays, etc. etc. on hand and can easily order more or less anything from Mauser or Digikey as I am in Canada.



Joined Mar 14, 2008
Here's the LTspice simulation of a TL431 circuit that draws <3mA when active and only leakage current (<1µA) when the output is cut off:
Pot U2 allows adjustment of the cutoff voltage (yellow trace, showing tripping at 16.1V for a 50% pot setting).
The output stays cutoff even if the voltage rises back to normal, so the circuit can be reset back on by momentarily pressing PB S1 (red trace at 0.8s).

If you want to be able to turn it off manually, you can add another push-button between Vb and the MOSFET gate (G).
Thus the circuit can double as an ON/OFF switch.

Last edited:

Thread Starter


Joined Apr 14, 2021
Thank you both very much for the replies. I had not come across that particular circuit nor the ICL7665S in my searching. I'm going to give both solutions a try.

Thanks again!