0
\$\begingroup\$

I designed a PCB with a microcontroller and a load that draws 3A on average with current spikes of up to 9A. The PCB is powered by a high-drain Li-Ion battery. I want to add a simple undervoltage protection feature to the firmware that cuts the load from the power supply when the battery voltage falls below a certain threshold. The microcontroller can measure battery voltage, but I'm unsure under which conditions to cut the load from the power supply. There are in principal two different approaches:

  • cut the load from the power supply when the battery voltage falls below 2.7V (or another threshold), independent from how long the battery voltage stays below this value
  • cut the load from the power supply when the battery voltage falls below 2.7V (or another threshold) for a certain time

I'm unsure which approach would be better. If the second approach would be better, what would be an appropriate time during which the battery voltage would need to stay below the threshold?

\$\endgroup\$
2
  • 1
    \$\begingroup\$ I would lean towards at least low pass filtering (averaging) the voltage reading of the ADC measuring the battery voltage, so you would not react to short voltage drops caused by current spikes. Maybe over one second - but this is just a gut feeling. \$\endgroup\$ Nov 30, 2019 at 19:13
  • 1
    \$\begingroup\$ " what would be an appropriate time" - how long are your current spikes? \$\endgroup\$ Nov 30, 2019 at 19:47

2 Answers 2

1
\$\begingroup\$

I would go with the second approach, as you don't want to cut off for momentary spikes in current. However, towards the end of discharge, Li-ion battery tends to drop the voltage quickly. So you cannot wait for long, a delay of 2 seconds should be good enough.

The threshold voltage will depend on the chemistry and the undervoltage protection built into the battery.

The battery terminal voltage is a function of current as well. If the current is higher, there will be higher drop in battery voltage. If you are expecting no load on the battery for some time, you can consider 2 different threshold voltages, one for no load and one for full load.

\$\endgroup\$
1
\$\begingroup\$

I think second approach would be better because 9A load can lead voltage drop for each cell as a result of internal resistance*current you applied. You may also consider voltage curve of li-ion cell at the end of discharge. It goes to zero volt rapidly due to chemistry and internal resistance increase. Therefore 2-3 second delay is suitable for yor application.

\$\endgroup\$

Your Answer

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

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