STM32f103c8 Blue Pill High Power Consumption

[ButterScotchCookie]

Hi, I'm designing a low power tilt and shake detection sensor module using the STM32 blue pill board. I'm trying my best to attain the datasheet values of 2-5µA. 

I have done the following:

1) Set all GPIO pins to Analog 
2) Disabled all peripherals manually：

3) Enabling PWR_CLK decreased current consumption by a little

4) Break PWR LED and PC13 LED

5) All serial wires disconnected

This is how I enter Standby Mode:

I powered my board using a 3.7V Lipo battery and my multimeter triggers 0.404mA.
Maybe if I remove the pull-up resistor on the NRST pin, I will reduce by another 0.3mA (3.3 / 10kOhm).
So where is the remaining 0.1mA coming from?

1) Could it be capacitor leakages?

2) Does USB leak current even when not in use?

3) The voltage regulator on my board has the markings "LB33 E1A2". According to a schematic I found online:

This regulator has a quiescent current of 90µA. But, I'm powering through the 3.3V pin, so it should bypass the regulator anyways. 

Not sure where the remaining 0.1mA is...

This is all I can think of now, appreciate any help on this. Thanks!

[Peter BENSCH · Accepted solution]

Welcome @ButterScotchCookie, to the community!

As you write, you are using a Blue Pill, for which there are already plenty of problems here in the community. In many cases, more than just a little LED flashing is done, so that a lot of problems quickly arise. However, these are usually caused by the fact that only counterfeits have been used in Blue Pills for years, which of course cannot be expected to comply with ST specifications.

Even if the (counterfeit) ST logo and the designation STM32F103C8T6 can be seen on the chip, you should make sure and read out the chip ID. In my humble opinion, however, you can save yourself the trouble and either use an original STM32F103C8T6 or ask the manufacturer of the chip on the Blue Pill.

Regards
/Peter