Problem with 3.3V external power delivery to my board

MWojt.2 · ‎2023-02-22

Hi,

Any program on my stm32f439zi board does not start when I supply the voltage from the external power supply at 3.3V, unless I connect the programmator via USB cable to the computer as well. When I disconnect the USB cable the program continues its execution, but everything takes way more time than usually. After reseting the board by pushing black reset button the situation reappears, I need to plug the programmator to the computer once again and then if I unplug it everything slows down, but continues execution. I use high-class laboratory power supply and the current drained by the stm board do not get even close to the limits my power supply can deliver. I plugged the "+" of my supply to the CN11 pin 16 as the documentation suggests and GND to GND as well on CN11. I ensured that SB3 (3.3 V regulator) and SB111 (NRST) are OFF and there are no accidental short circuits. Any advice on how to fix the issue?

Regards,

Maciej

Peter BENSCH · ‎2023-02-22

Welcome, @Maciej Wojtyś, to the community!

You have stumbled across a little trap: in some NUCLEO boards there is no crystal for HSE and the clock for HSE comes from the embedded ST-LINK. On your NUCLEO-F439ZI there is still an ST-LINK/V2-1 whose MCO output feeds the 8MHz clock via the two jumpers / solder bridges SB112 and SB149 to OSC_IN of the HSE (see schematics).

If you need a precise clock and the internal oscillators with their lower precision are not precise enough for you, you can now either additionally supply the ST-LINK/V2-1 from the 3.3V input, or open SB149 and populate crystal X3, C37, C38, SB8 and SB9.

Hope that helps?

Regards

/Peter

In order to give better visibility on the answered topics, please click on Accept as Solution on the reply which solved your issue or answered your question.

View solution in original post

Peter BENSCH · ‎2023-02-22

Welcome, @Maciej Wojtyś, to the community!

You have stumbled across a little trap: in some NUCLEO boards there is no crystal for HSE and the clock for HSE comes from the embedded ST-LINK. On your NUCLEO-F439ZI there is still an ST-LINK/V2-1 whose MCO output feeds the 8MHz clock via the two jumpers / solder bridges SB112 and SB149 to OSC_IN of the HSE (see schematics).

If you need a precise clock and the internal oscillators with their lower precision are not precise enough for you, you can now either additionally supply the ST-LINK/V2-1 from the 3.3V input, or open SB149 and populate crystal X3, C37, C38, SB8 and SB9.

Hope that helps?

Regards

/Peter

In order to give better visibility on the answered topics, please click on Accept as Solution on the reply which solved your issue or answered your question.

MWojt.2 · ‎2023-02-23

Hi

You are right, choosing HSI instead of HSE solved the issue. I was wondering about it as well, but for some unknown reason STM32CubeIDE did not allow me to change the PLL Source Mux clock input, so I thought HSE was the only option when using this particular board. In this case placing the only oscillator beyond the main part of the board would be extremely weird. Now when I restarted the computer this problem disappeared, and so did the other problem with startup and execution slowed down when only I switched to HSI. Now I can power the board externally and my board starts up as normal. I appreciate your help, thank you very much.

Regards,

Maciej