1
\$\begingroup\$

I am currently working on a project that involves charging 18650 Li-Ion battery through USB-C using TP4056. Nothing really fancy (or at least have I thought!).

The circuit is pretty simple - Battery connected permanently to the TP4056; the rest of the circuit (two boost converters, ESP32 etc.) is separated by toggle switch (SW3).

Schematic of the charging circuit

Now, very weird thing happens when the user tries powering on / off the rest of the device by toggling the switch - the TP4056 starts drawing approximately 1A from the battery and instantly overheats, even though it is not plugged into the USB (+5VP rail is disconnected). This does not stop until I take out the battery.

This problem happens only from time to time, so it is pretty random. However one interesting thing is that when I tried to power the device with my lab PSU instead of the battery, it used to happen a lot more.

I managed to capture the moment, when I toggled the switch and the problem occurred. The screenshot shows voltage on pin 5 of the TP4056 when toggling the switch:

Oscilloscope screenshot when toggling the switch

but IMHO there is nothing that special, that should "lockout" the IC until it is reset.

The TP4056 that was causing me problems was ordered from LCSC, PN C16581. Then, I swapped it with another TP4056 (again from LCSC, but PN C382139), made by a different manufacturer and now it works flawlessly. But I am still a little bit worried that this problem could come back even with the "good" TP4056, because I highly doubt that same ICs from different manufacturers should behave so differently.

For those thinking that it might have been only one bad TP4056 - it was not. All C16581 that I had at home were having the same problem.

Has anyone ever experienced something like this? Does anybody know what could be causing such problem?

\$\endgroup\$
2
  • \$\begingroup\$ It is naive to think the same type number part from different manufacturers will be identical. Nevertheless, show your pcb layout as this is a critical part of the circuit. \$\endgroup\$ – Kartman Apr 29 at 21:47
  • \$\begingroup\$ The charging circuit is in the upper right corner of the PCB. The layout in PDF also with the original KiCad files is here: PDF layout; I did not mean that they would be identical. I just thought, that they would be functioning. \$\endgroup\$ – Marek Coufal Apr 29 at 22:03
0
\$\begingroup\$

Short answer: Low quality error prone clone ICs have polluted the market.

Long answer: This is a surprisingly well know issue, but the info is scattered. And the prices are low so people are still buying them, as such these unfunctional chips are still flooding the market.

Known problems

  • IC gets destroyed when used with a switch. - Reference
  • Charging problems (see the comments section) - Reference
  • Wrongly placed resistors (for the module version) - Reference
  • Doesn't stop charging - Reference

Clone vs Original

enter image description here

PS: Stay away from these ICs! I've read people trashing hundreds for their unreliability.

\$\endgroup\$
1
  • \$\begingroup\$ I will look for the desoldered ICs when I'm home and check how their marking looks. \$\endgroup\$ – Marek Coufal Jun 20 at 11:24

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.