Voltage Stacking with energy harvesting STM32WB55

methene · ‎2025-02-25

Hello, I am attempting in voltage stacking of 2 STM32WB55 board, but unlike arduino I could not connect Top board's GND to bottom board's 3.3 V pin. I would like to request for help or insights for alternative method I could make it work. I would really appreciate the help.

Problem

1. When I connect the Top STM32 GND to the VDD of Btm STM32, it discharged the voltage in Btm capacitor.

2. On one attempt I rearranged the circuit in similar stacked configuration, it did not discharged the Btm capacitor voltage. Instead the system goes haywire.

Thank you.

methene · ‎2025-02-27

The multimeter I am using does measured uA, but with a slow sampling rate. I do understand your point, I actually observe the value multiple times.

If that is the case I will try to contact my lab to purchase this equipment.

methene · ‎2025-02-27

Currently the STM32WB board will consume voltage from capacitor once it reaches 3V.

I totally agree with you, multimeter might not be the best choice but it is what I had in my hand currently. The oscilloscope in my lab does not have a current probe to measure current. Will contact my lab to purchase this.

methene · ‎2025-02-27

Is this a first year project? You should know parallel vs series by now. That's high school level physics.

This is my bachelor project, I am not familiar with how the circuit of energy harvesting system should work. Therefore, I am trying out anything I can.

What does "it discharged the voltage in" mean? It's not correct use of terminology and probably not grammatically correct either. I have no idea what it means.

It discharged the capacitor or the capacitor could not be charged, as if its positive polar was directly connected to GND.

Define haywire. What happened? Or what didn't happen? I assume one MCU received all the voltage and was being fried while the other one shut down due to too low voltage.

In this particular situation, both board turns on and capacitor started charging, the TOP board was running just fine as intended, but the BTM board capacitor seems to be supplying power to TOP board during operations. Causing BTM board to be not functioning properly. Will be attaching a picture of the result for better explanation.

methene · ‎2025-02-27

Thank you, will be remembering this and make this my motto when doing experiments.

Andrew Neil · ‎2025-02-27

@methene wrote:
The operating voltage of STM32WB series is from 1.71V to 3.6V

Exactly: there is a minimum voltage (1.71V) - it will not work below that voltage.

Therefore you cannot power it from a supply which ramps-up slowly from 0V - such as your PV + capacitor.

@methene wrote:
The STM32WB board monitors the voltage in capacitor and only start the task when it reaches 3V.

That won't work - see above.

At the very least, you will need an external supervisor to ensure that the capacitor has not only reached the STM32's minimum operating voltage, but also has enough charge to start-up the STM32 and then keep it running.

If you start the STM32 as soon as the voltage reaches 1.71V, its startup will cause the voltage to dip back down below the minimum.

Again, this is the kind of thing that proper energy-harvesting power supply chips take into account and manage.

methene · ‎2025-02-27

To deal with this, when it reaches 1.71V the board will enter LP mode without initializing anything until 2.2V only then initialization begins. This is to overcome the dip back down below the minimum problem and to ensure it would not reset due to below 1.71V.

Andrew Neil · ‎2025-02-27

@methene wrote:
multimeter might not be the best choice .

It's worse than that - it will give you completely useless & misleading results.

This is where we hit the difference between theory & real-life practice.

@methene wrote:
The oscilloscope in my lab does not have a current probe

You don't necessarily need a current probe - you can measure across a sense (or "shunt") resistor.

A current-sense amplifier (CSA) can help - which is how I did this.

As @Uwe Bonnes suggested, the limitation with such an approach is the dynamic range - which is what the more expensive devices address...