I have a device in my flowerbox where I monitor soil moisture. The device requires about 3.3 v (it's an Arduino pro mini with an ESP8266).

The device is running on 3 x AAA-batteries (should be 3.6 v)

The device runs a cycle of:

  • Powering on ESP and Sensor
  • Collecting sensor data
  • Transmitting it via WiFi
  • Powering off ESP and Sensor
  • Sleeping 20 minutes

It's been running fine for almost 75 hours, then it started continuously restarting.

I measured the voltage of the batteries. They were 2.86 v. Then I waited 5 minutes, measured the voltage. Now they were 3.1 v. I thought this problem might be caused by low outside temperature. I stuck the batteries in the fridge for 20 minutes, then tested them. They were 3.1 v.

So question is: How come the batteries voltage change that much?

Best regards Frederik

  • 3
    \$\begingroup\$ The most common reason of batteries having lower voltage than printed on them is that they are empty \$\endgroup\$
    – PlasmaHH
    Jul 10, 2017 at 10:16
  • \$\begingroup\$ What is your average current consumption from your system? \$\endgroup\$
    – winny
    Jul 10, 2017 at 12:31
  • \$\begingroup\$ sounds like the ESP is not actually sleeping; i get a weekend from 3 AAAs w/o deepsleep... \$\endgroup\$
    – dandavis
    Jul 10, 2017 at 20:08
  • \$\begingroup\$ @dandavis, I'm turning the ESP off via a transistor. \$\endgroup\$
    – Frederik
    Jul 11, 2017 at 7:46
  • \$\begingroup\$ if that's the case, your batteries should be lasting a lot longer... \$\endgroup\$
    – dandavis
    Jul 11, 2017 at 14:07

1 Answer 1


They were 2.86 v. Then I waited 5 minutes, measured the voltage. Now they were 3.1 v.

You may be seeing the recovery effect as the energy in the batteries is depleted to a low level. When the load is removed, chemical or physical processes inside the battery (e.g. diffusion) allow it to recover partly.

I would use an analog input to monitor voltage and prevent restart looping.

I would not expect 3 x AAA cells to provide long term energy. At the least I would try to measure power consumption accurately for planning purposes.

  • \$\begingroup\$ Thank you very much for your prompt feedback. I was actually thinking about using an analog input for voltage measurements but I just hadn't come around to implementing it. Can you recommend an alternative rechargeable power source? \$\endgroup\$
    – Frederik
    Jul 10, 2017 at 11:27

This site is temporarily in read-only mode and not accepting new answers.

Not the answer you're looking for? Browse other questions tagged .