I have a battery-powered circuit which uses a Li-Ion battery which can be charged over USB using an MCP73831 charging controller. The battery has built-in deep-discharge protection, so disabling discharging is not required. However, Li-Ion batteries shouldn’t be charged below 0°C or above 60°C, which is why I want to be able to disable charging while still powering my circuit over the charging controller (which is why I don’t want to simply turn off the charging controller).

Would the following circuit with Q1 in an unusual, reversed direction work? The CHARGING_ENABLE wire is connected to a microcontroller in the 2.1V domain.


  • 1
    \$\begingroup\$ It appears that the MCP73831 is only intended for charging applications, not charging+load management applications. A different charger is likely recommended to have it properly supply power to the buck converter, too. \$\endgroup\$ Jun 14 '16 at 19:49
  • \$\begingroup\$ This is the one I use, MCP73863, it has a EN pin, for enabling and disabling the charger, you could just use a Mosfet and disconnect the battery from the charger? \$\endgroup\$
    – Alex
    Jun 14 '16 at 20:01

Two tricks...

  1. "The charge management controller can be disabled by allowing the PROG input to float." (Sec 3.5 in the datasheet)

  2. Add a diode to power the step down directly from the supply when the supply is present.

The resulting circuit looks like this....

enter image description here

Anytime the +5V supply is present, it will be higher voltage than the battery voltage which will be maximum of about 4.2V even when charging. This reverse biases the diode, so the step down (or any downstream load) is powered directly from the input supply rather than the battery and charger.

"Charge Enable" now enables the PROG pin to drain though the PROG resistor. Without "Charge Enable", the PROG pin will float and the battery charger will not charge the battery (use this when temp is out of range).

  • \$\begingroup\$ Thank you, I had this idea too. I like its simplicity, but the battery losses on the diode (even if I use a Schottky) are a bit of a disadvantage. Is there a way to use a transistor instead? \$\endgroup\$
    – Michael
    Jun 15 '16 at 5:27
  • \$\begingroup\$ Yes, you can trivially replace the diode with a MOSFET to greatly reduce the forward voltage drop. web.mit.edu/6.012/www/SP07-L25.pdf \$\endgroup\$
    – bigjosh
    Jun 15 '16 at 12:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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