[Solved/Fixed] Killed Arduino Pro Mini 3.3V with low power experiment?


I followed this guide (Low Power), step 2 in particular, and cut the traces to and from the power LED. First, I have cut the trace between LED and resistor (see pictures attached). I measured the current and it was above 500 µA while sleeping. Because the guide reports 54 μA consumption, I also cut the trace between LED and capacitor. Current consumption went down to 24µA.
Surprisingly, and unfortunately, I cannot program my Pro Mini anymore and the status LED does not flicker when power is applied, which it should AFAIR. Using my DMM I could not find any suspicious shorts or whatsoever. For example, there is no unexpected voltage drop.

Maybe someone has an idea, what to look for to find the cause of the problem.
Should I try and unsolder the LED, capacitor and/or resistor?

Thanx & best!

Edit: Title


Wouldn't it have been easier to just unsolder the resistor, or the LED? Only need to remove one or the other.
I don't see any components lined up like yours in this photo

Try putting a wire back across the trace to the cap, see if that helps the downloading.

... because circuit boards have a top, bottom, and often layered traces, cutting a trace can easily "break" connections made through the visas vias (plated holes between layers):
(thanks to Paul__B for catching the spelling error.)


I also cut the trace between LED and capacitor.

Well, usually capacitors are not connected to LEDs.

You probably cut the Cap connection from/to the RESET Pin (V11 schematics).
This would explaine You can not programm your mini anymore (no reset).

I don't read the whole instructions from your link but there is something wrong there.

...your Pro Mini should now only use about 54 μA (0.054 mA) in power-down mode
if it is the 3.3 V version and about 23 μA (0.023 mA) if it is the 5 V version...

Usually low Voltage µC's needs less current also in power-down mode.


Hi and thank you all for the feedback!
I thought about cutting a trace on an intermediate layer but I don’t think that this pro mini does have an intermediate layer because I can follow all holes from back to the front (see attached). And, actually, I also wondered why a capacitor would be connected to the LED but it iwas just connected to VCC.
Btw. cutting took me only few seconds in contrast to unsoldering :wink:

I had a look at the ATMega328 schematics and confirmed with my DMM: the cap is placed between the GND (Pin3 | 32 MLF, Top View) and the VCC (no short here, the cap seems to be fine). In addition, the RST pins are still connected to pin 29 of the ATMega328.

Maybe the problem has nothing to do with cutting the traces–just bad luck or a different mistake of mine, I did not recognize.

I attached 2 AA (2.7V) to VCC/GND. I measured the voltage between the other VCC/GND: 138mV. :o When connecting the 2 AA to RAW/GND, the voltage on VCC/GND is 380mV.
Obviously not correct. What have I killed? Voltage regulator, diode, the ATMega itself, or …?

Thank’s a lot & best!

So... you killed the decoupler cap? That could prevent your board from booting.

This would explain the voltage readings on VCC/GND? Any suggestion on how to test this assumption?

Thanks & best

You probably cut the connection between the USB power and the ATMega power.

Hi Issac,

Thanks for your help. What exactly do you mean with "USB power". The pro mini (see my pictures in previous posts) does not have an USB port.


I was thinking the connection between the voltage regulator and the ATMega.

I did some more testing, just because I'm curious what happened.

I used a Arduino Nano as ISP to program the Pro Mini via Pin 11, 12, and 13 plus RST. Funny thing, avrdude reports correct/verified flashing. Unfortunately, the blink example did not work. I tried both a different pin and a different sketch but no luck.
Interestingly, I flashed a blank sketch and a sketch that set all pins to OUTPUT/LOW, respectively. I used my DMM to measure the voltage between pins and GND. All digital and analog pins as well as the RST pin showed approx. 2.5 V. Well, that's a bad sign I guess.

I unsoldered the cap (next to pwr LED), pwr LED and LED resistor. No change in behavior. :wink:

Curiosity or not, I've spent enough time with the pro mini. Unless someone have another idea due to the described behavior.

Can I use the ATMega328P alone? No relevant difference to an ATMega328P DIP-28 which I'm running on a breadboard?
Would it be enough to unsolder everything except the ATMega328P? Maybe I have to cut some traced or short some pins I'm not aware of?

Would be a pity to trash the complete board ...

Thanks & best!

Take everything off and see if it works.

You should have only cut between the resistor and the LED - the other cut is bogus.

Thanks for the replies.
I'll give it a shot and let you know if it worked out. :slight_smile:


I love it! Problem solved, board running.
The problem was a disconnected GND pin from the ATMega. I bridged the GND of the ATmega to another GND (which I noticed were disconnected).
Description (because I don't have a DSLR at hand): On picture pro_mini_p2.jpg I soldered a bridge from the left end of the capacitor to the right end of the LED.
Funny thing, I noticed just now that there is a "third" connection from the LED, it's the missing GND link!

As some of you pointed out: don't cut, de-solder :wink:
Lesson learned :slight_smile: