I am working with the STM32 MCUs and I wish to step up a 5V pin to a range between 7V and 12V. In summary I wish to power the second Board from the first board. Thanks for any reply.

  • \$\begingroup\$ What is your question? \$\endgroup\$ Jan 15 '19 at 14:15
  • 1
    \$\begingroup\$ How much current does your second board draw? How much current can your 5V supply? \$\endgroup\$
    – DoxyLover
    Jan 15 '19 at 14:22
  • 1
    \$\begingroup\$ The second board drawsmax 800mA and the I think the 5V supplies close to 500mA. \$\endgroup\$
    – McSteve
    Jan 15 '19 at 14:29
  • 3
    \$\begingroup\$ No, you can't take 5V at 500mA and step it up to a higher voltage at 800mA. That would violate conservation of energy. \$\endgroup\$ Jan 15 '19 at 14:33
  • 1
    \$\begingroup\$ You need to provide much more details and your attempts to solve the problem so far. Step-up regulators is a big topic. \$\endgroup\$
    – Lundin
    Jan 15 '19 at 14:59

I am working with the STM32 MCUs and I wish to step up a 5V pin to a range between 7V and 12V.

I see several major problems here.

First: the "5V pins" on a STM32 MCU are 5V tolerant. They can accept a 5V input signal without damaging the MCU, but they will only output VDD (usually 3.3V) as a high output, just like all the other general-purpose output pins on the device.

Second: The GPIO pins cannot source or sink a useful amount of power. The absolute maximum current draw from a GPIO pin on a representative STM32 device (STM32F103C8) is 25 mA, with an additional maximum of 150 mA for the entire device. Attempting to draw 500-800+ mA from the chip (per your comments) will destroy it.

Given the constraints you've described in your comments, you need a secondary power supply for your 7-12V board. The power supply available to your STM32 is insufficient, and the STM32 is not capable of producing or transforming power in the way you believe it to be.


Per your comment that your second board requires 800ma at 7-12V but your 5V only supplies 500ma, you can’t do it. You have 2,500mW of power but you need at least 5,600mW. This just doesn’t work.


Assuming that the power requirements at the higher voltage are modest, you can program the microcontroller to output a square wave at one pin, Vsq, and use a voltage-doubling rectifier stage afterwards. Look at this Wikipedia link and the Greinacher circuit.

A similar idea, with some improvements, is Philips Application note AN10218 which suggests a circuit like this:


simulate this circuit – Schematic created using CircuitLab


  • The original application note uses an astable circuit to generate the square wave. I suggest that your microcontroller code does this additionaly to its usual operation. The waveform does not need a 50% duty cycle. In fact, it does not even need to be strictly periodic.
  • The frequency of Vsq does not need to be 1 MHz.
  • Circuit element values can be tuned accordingly. Schotky diodes would be better, but I have not bothered to search any of them
  • \$\begingroup\$ Thanks alot for the explanations and all the Tipps. \$\endgroup\$
    – McSteve
    Jan 15 '19 at 15:00
  • 1
    \$\begingroup\$ "In summary I wish to power the second Board from the first board" So this doesn't answer the question. In particular, where does the clock come from when there is no power? \$\endgroup\$
    – Lundin
    Jan 15 '19 at 15:00
  • 2
    \$\begingroup\$ This answer isn't applicable to the OP's situation, and following it is likely to cause hardware damage. The "5V pins" on their microcontroller are 5V tolerant -- they do not output 5V -- and are limited to sourcing/sinking 25 mA. \$\endgroup\$
    – user39382
    Jan 15 '19 at 19:25
  • \$\begingroup\$ How do you expect to get 800mA out of this? I don't think it has any chance of meeting the OP's needs. \$\endgroup\$ Jan 16 '19 at 0:44
  • \$\begingroup\$ @Lundin: The clock comes from the first board, which is the one that has power. Elliot: The 800 mA requirement came after my response... Hence, the assumption of "modest power consumption". And duskwuff is right, I did not bother to check if there are 5V versions of the STM32. It seems there aren't. In any case, the proposed circuit does work for "modest comsumption" and 5V pins :) \$\endgroup\$
    – Petrus
    Jan 16 '19 at 12:54

Not the answer you're looking for? Browse other questions tagged or ask your own question.