I need a very small UPS just to give an embedded device time to shut down when power is lost. There are a variety of solutions that use LiPo batteries to provide extended runtime; they typically take USB input and manage both the charging and failover to battery on power loss.

Can I replace a single-cell LiPo (nominal voltage 3.7v) in a charging circuit with a pair of 2.7v supercapacitors wired in series and have it mostly "just work" assuming that (a) there is already some sort of current limiting in place on the charging circuit and (b) I include a passive balancing circuit between the capacitors?

In this situation, I would be under-charging the capacitors (to 3.7v, vs the combined 5.4v capacity in this example). What impact does that have on their behavior?

  • \$\begingroup\$ A LiPo charger may refuse to charge a "dead" (substantially undervoltage) cell, so you may need to replace the charger with something (resistor?) if there isn't enough leakage to get the voltage to the point where the charger considers injecting current safe. \$\endgroup\$ Commented Oct 23, 2018 at 16:14
  • \$\begingroup\$ Keep in mind that, unlike a battery, the voltage on the capacitor will begin to drop the instant you start discharging it. You will need to see how much headroom you have voltage-wise, and also calculate the amount of energy you will need to see if you have enough capacitance to keep the thing up long enough. \$\endgroup\$ Commented Oct 23, 2018 at 16:20
  • 1
    \$\begingroup\$ When contemplating caps instead of batteries, always compare the energy stored vs. volume & weight, because that's most often the deal killer. A 1-farad, 5V capacitor can store 12.5J at 5V. That's roughly the energy equivalent of a LiPo battery with a capacity of 1mAh. You can't find LiPo cells that small! \$\endgroup\$
    – TimWescott
    Commented Oct 23, 2018 at 18:21
  • \$\begingroup\$ @TimWescott Note that I am explicitly not trying to achieve the same runtime I would get with LiPo batteries. I'm looking for < 60s of runtime, for which I think the caps should be sufficient. \$\endgroup\$
    – larsks
    Commented Oct 23, 2018 at 18:29
  • 1
    \$\begingroup\$ @larsks: if you've done the math, all well and good. \$\endgroup\$
    – TimWescott
    Commented Oct 23, 2018 at 20:01

1 Answer 1


You could do this, but will it work? The charging circuit will be confused as it is built to charge/discharge at certain voltages. Because lithium ion batteries are considered dead below a certain threshold (like 3.0V) the UPS charging circuit may not switch on at all because a supercapacitor will have no voltage initially.

As far as the capacitor goes, it doesn't matter if it is charged below a threshold, the main thing to worry about is the breakdown voltage (max voltage) and that you don't exceed that rating. It can hold any value between the max voltage and zero.

enter image description here Source: https://learn.adafruit.com/li-ion-and-lipoly-batteries/voltages

enter image description here
Source: https://www.tecategroup.com/ultracapacitors-supercapacitors/ultracapacitor-FAQ.php

  • \$\begingroup\$ After you posted the OP said: The device draws around 400mA. The batteries/capacitors feed into a boost converter that can in theory take input as low as 1.8v, but because the charging circuit is targeted at LiPo batteries it may cut off around 3v. \$\endgroup\$ Commented Oct 27, 2018 at 20:35

Your Answer

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

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