I'm trying to put an Arduino Nano in low-power mode using this guide: Arduino Low Power - How To Run ATmega328P For a Year On Coin Cell Battery.

The article suggests 3 things: use the Low-Power library from Rocketscream, cut the power to the power LED, and remove the voltage regulator.

So far I've only done the first two and reduced the clock speed to 8 MHz.

The important part of the source code that puts the Arduino in deep sleep looks like this:

void setup() {
  CLKPR = 0x80; // (1000 0000) enable change in clock frequency
  CLKPR = 0x01; // (0000 0001) use clock division factor 2 to reduce the frequency from 16 MHz to 8

void loop() {
  // Go to sleep until the next scheduled wake up
  int32_t waitSeconds = 60;
  while (waitSeconds > 0) {
    period_t waitFor = SelectWaitTime(waitSeconds);
    LowPower.powerDown(waitFor, ADC_OFF, BOD_OFF);
    waitSeconds -= PeriodToSeconds(waitFor);

period_t SelectWaitTime(int32_t secondsToWait) {
  if (secondsToWait >= 8)
    return SLEEP_8S;
  else if (secondsToWait >= 4)
    return SLEEP_4S;
  else if (secondsToWait >= 2)
    return SLEEP_2S;
  else if (secondsToWait >= 1)
    return SLEEP_1S;

In essence, sleep for up to 8 seconds, remove the seconds from the counter, then sleep again.

So far I've only scratched off the trace to the power LED and haven't removed the voltage regulator. According to the article above, this should reduce the power consumption to around 0.05 mA but I get around 1.7 mA, which is over 30 times more!

Electrically, I only have a MOSFET as a power switch connected to one of the analog pins (I use it as a digital output) and another controller, connected to the I2C pins (the Arduino acts as Slave). The other controller is turned off through the MOSFET when I take the measurements. Physically removing the connections doesn't change the consumption, so I don't think any of that is related.

Is there something I did wrong? What can I do to achieve power consumption similar to the one in the article?


Turns out the voltage regulator was AMS1117 - not MIC5205-KB50/KB33/KBAA like the one in the article. Ripping it off gave me the 0.026 mA idle current which is even better than what I was hoping for.

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.