I'm trying to figure out how to safely implement fast charging using the BQ25756 with a most likely 2s (though up to 4s is possible) layout of physically spaced out cells.

I haven't fully decided on which battery chemistry to go with, though I imagine NMC is the best for longevity at high current draw with fast charging, though I have considered LTO despite its low capacity and lower voltage necessitating more cells.

Anyhow, I've come across what seems to be a general truth that most battery charger IC's have only one thermistor input for a battery pack.

This seems ideal for many battery packs where the batteries are close in proximity and will therefore most likely share similar temperatures, however, for my planned application, the batteries will be mounted end to end, out of physical contact, so my worry is, especially when implementing fast charging at up to 3c or greater, one battery might be in an unsuitable range, while the other battery might still be within a suitable range causing one battery to go outside of the parameters specified as safe by the manufacturers datasheet (the battery that does not have the thermistor attached).

As such, I would like to give both batteries a thermistor of their own. I am wondering if I could potentially use an analog switch and a counter/timer to switch back and forth between the thermistors of either battery, so that the charge IC won't mindlessly push either cell of this 2 cell configuration out of spec (though I don't see any reason this couldn't apply for higher cell counts as well with an appropriately sized analog switch).

The concerns are as followed:

  • I don't know with what frequency temperature is checked (Ideally I would switch with enough frequency that the safety timeouts for each deteriorated state would not be reset in between switching between batteries; such that if one battery is out of the appropriate temperature change, charging wont restart when it samples the battery that is within the appropriate temperature range)

  • What if switching between encounters time segments of no flow where the charger ic could erroneously be tricked into an error state by the lack of thermistor data whilst the analog switch is mid states.


1 Answer 1


It sounds like you are making this hard for yourself with your proposal. What I would do is let one battery pack's thermistor connect up the charger chip (as normal) and, monitor the other battery pack for anomalous temperatures using a 2nd circuit.

When monitoring the 2nd pack (a comparator is all that is needed), you can disable charging if it indicates a temperature that is too high.

  • \$\begingroup\$ That sounds just about as complex though an option if I were to give up on this idea. I'm currently working on a circuit to multiplex using an rc filter with a schmitt trigger going to an analog switch to latch to the hottest thermistor if so for more than ~0.5 to a second or so, just so there isn't constant switching. I'm just assuming the switches will be infrequent and quick, so it probably wont cause problems. I guess I'll see how many legs that idea has. As far as it can tell, it will involve getting a reading from the thermistor with an isolated circuit to avoid disturbing the BMS \$\endgroup\$ Commented Nov 29, 2023 at 18:08

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.