Pro Mini sleep mode only 3.45 mA

I am trying to reduce the current draw on a Pro Mini 3.3V. I'm powering the bare naked board using a 4-AA battery pack hooked up to RAW. The current draw is 6.24 mA upon startup, and 3.45 mA after going into sleep mode (10s after power is turned on). My understanding is that SLEEP_MODE_PWR_DOWN mode should give a much lower current draw. What am I missing?

Here is the simple test sketch:

#include <avr/sleep.h>

void setup() {
     delay(10000);
  sleep();
}
void sleep(){
  set_sleep_mode(SLEEP_MODE_PWR_DOWN);
  sleep_enable();
  sleep_cpu();// puts Pro Mini to sleep
}

void loop() {

}

Forgot to mention that I already scraped off the power LED.

With the schematic you did not post I can only take a swag and say it is the regulator on the board that is drawing the current Plus what else you may have connected.

If this does not answer your question please post your annotated schematic not a word salad.

I use this code when measuring sleep current.

// This puts an ATmega328P to sleep so you can measure current draw.
// Sleep current should be 1uA or less.
// Some counterfeit chips have sleep current over 100uA.
// This is for a Pro Mini with power LED and voltage regulator removed,
//    so the processor is the only thing drawing current.     

#include <avr/sleep.h>
#include <avr/wdt.h>

void setup(){

  ADCSRA = 0;                         // disable ADC for power saving
  wdt_disable();                      // disable WDT for power saving
  set_sleep_mode (SLEEP_MODE_PWR_DOWN); // Deep sleep
  sleep_enable();
  sleep_bod_disable();                // disable brownout detector during sleep
  sleep_cpu();                        // now go to sleep

}

void loop(){
}

So if you've removed the power indicator LED, then with this code anything over about one microamp would have to the the voltage regulator.

However, there is a history of fake 328P chips on Pro Minis, and the fakes draw more current during deep sleep - around 140uA. But that's still not 3.45mA. So it's probably the regulator. You might try three AAs in series to the Vcc pin. That would bypass most of what the regulator would draw, but not all.

There is no schematic because as my word salad indicated, the board is bare - really bare - literally just lying on the desk with the power wires attached. When you say it may be the regulator, do you mean the regulator may be defective?

The unit was bought on Amazon - brand Coliao. Is that a reputable brand?

the cheap regulator is not swirtching , so the surplus go out , so ideal is to cut regulator and led if you want have low current in deep sleep , arduino is this invention, first born atmega328 , but after someone think to put atmega328 on the board with led, serial usb and power regulator , and in this mode is born arduino , so arduino is not a real invention , if not think arduino to put atmega328 on the board , shure that another man make this after, because for use atmega328 alone is difficult, and this consuming power, because atmega328 is not alone in the board

I tried a 3-AA battery pack (4.7V measured). Weird result: Current is now higher.

4.7V thru RAW gave 6.2mA (before sleep), 3.4mA (sleeping).

4.7V thru VCC gave 11.2mA (before sleep), 7.14mA (sleeping).

Should I just conclude the board is messed up, and move on?

There isn't much on a Pro Mini board other than the 328 and the voltage regulator/LED. You can power it up to 5V5 even though it is a 3V3 model through the VCC pad.


This is the power supply section for a genuine Pro Mini. If you can find the trace where SJ1 is on your version and cut it you can completely eliminate the regulator and then power it at VCC.

Since you are powering your board from batteries you don't need the regulator at all. It will operate down to 2V7 at 8 MHz and will draw the least current at the lower voltage VCC.

Getting the lowest current requires you to not have anything else in the way of peripherals connected including the USB to serial programmer.

If that doesn't give you very low current consumption then you may want to order a genuine Pro Mini from Sparkfun.

Could be a new type of fake. @kentm please post a close-up, focused photo of the board.

Maybe by looking at the photo, you can offer advice on how to cut the correct trace?

I just ordered one from SparkFun. It will be interesting to compare.

Seems very weird that powering thru VCC about doubles the current draw vs. RAW.

It has a Mega 328PB instead of the 328P that an authentic one would have, for one. Many of the PBs are not Kosher.

As far as the correct trace to cut, the anode of the LED was probably connected to the output of the regulator.


The regulator is the chip in the red circle. I would guess that the output is that pin on the upper right. You could just remove it altogether.

Getting one from Sparkfun just makes good sense. You'll always have one to test sketches on that you won't have to worry about it being a shoddy clone.

Clones are cheap for some good reasons and they are cheap enough that some don't mind having to throw some away or that they behave strangely from time to time.

Got it. Now when I power thru RAW, nothing happens, so I think I successfully cut the trace. Powering with 4.7V through VCC gives the same current as before, however (6.7mA sleeping).

Maybe the sleep command isn't working?

I was curious what current a bare 328P would draw with your sketch. No regulator, no LEDs, just the bare minimum. The IC, a crystal, some caps and a pullup on /RESET.

I burned the fuses in a DIP 328P for 8MHz external, 1.8V BOD, no bootloader, everything else default, using Minicore:avr:328.

I put your sketch into it and put the it into a breadboard with a 8MHz crystal, a couple of 22pF caps, a 10K resistor on /RESET, a 0.1uF cap from AREF to ground and another 0.1uF cap across Vcc and GND.

On startup, the circuit used 3.84mA. After 10 seconds, it went down to 108uA.

After @van_der_decken's experiment, I wouldn't spend too much energy trying to figure it out. If you're lucky, it will be usable for projects that don't require low power if it doesn't have some other weird problems.

One caveat, the Sparkfun Pro Mini has an LED on PB5 of the 328 that you may have to remove to get the absolute lowest sleep current.

Thanks, all, for your help. I will do the same experiment with the Sparkfun unit and let you know what happens.

Here is the schematic for the Sparkfun unit.