Accurately measuring board current consumption.

[Karthik.ganesan]

Hi, I'm trying to get as accurate a power measurement as possible on an NUCLEO-F103RB board. I set it up using STM-CUBE to run at 1MHz and all peripherals in the default state. The only code I am running (once initialization is complete) is an empty while(1) loop.

I have bypassed all the ST-LINK circuitry and I am providing 3.3V using a Keithley 2450 SourceMeter. I am also performing the current measurement using the same meter. I have soldered off SB2 and SB12 as well.

I am getting different current measurements from different runs, without changing anything else in the setup. The current ranges from 2.43 all the way to 2.89 mA between runs. I just press the on-board RESET and wait for the current to settle.

Is there something else that could be happening in the background to cause this variation in power between runs? Any help/suggestions would be much appreciated!

[Piranha]

Could be floating input pins catching noise. Read AN4899 section 6.

[Karthik.ganesan]

Hi, Thank you! I'll try out those optimizations and see if the problem goes away. I also saw AN3430 when I googled your suggestion so I'll try that out too and report back.