5
\$\begingroup\$

I've recently finished programming the STM32F429ZI DISC1 board. I'm fairly new to this and now want to move away from connecting it to the PC via USB and power it up via a power supply.

At present I have connected a 5V supply to the 5V pin. This seems to power the board fine in that the LCD screen comes on (blank white), the PWR LED(LD2) lights up. However my program fails to initialise and stays blank, the COM LED(LD1) also flashes red at 1hz approx.

Has anyone experience with powering this board or other STM32 boards without a USB connection? And if so which Pin did you connect to and with what voltage.

\$\endgroup\$
  • \$\begingroup\$ AFAIK you need to put power before regulator so your board has all the necessary voltages such as 3.3V \$\endgroup\$ – Sean87 May 24 '16 at 10:30
  • \$\begingroup\$ There is a jumper, JP3 on the board which connects the 3V net with the VDD net. Is this jumper is ON on your board? It is also named as Idd and it is above the blue button. As far as I see in the schematic the MCU is only connected to VDD net. \$\endgroup\$ – Bence Kaulics May 24 '16 at 10:54
  • \$\begingroup\$ Also you could measure the voltage on the pin 22 on the P1 header (the one near the BOOT0 pin). \$\endgroup\$ – Bence Kaulics May 24 '16 at 11:00
  • \$\begingroup\$ @BenceKaulics thanks for your reply. JP3 is connected, and the voltage measured on pin 22 (Vdd) is just under 3V. Do you have a link to schematics, struggle to find anything on the ST website. \$\endgroup\$ – Pop24 May 24 '16 at 11:10
  • \$\begingroup\$ st.com/content/ccc/resource/technical/document/user_manual/6b/… \$\endgroup\$ – Bence Kaulics May 24 '16 at 11:11
3
\$\begingroup\$

From this answer: STM32F429I-DISC1 can not be powered by using ext. USB Battery

There is/was a known issue with the ST-LINK firmware shipped with the F429I-DISC1 and F407-DISC1 boards, and perhaps others. The STM32 is held in reset by the debugger when not attached to a host computer, so charger/battery use cases failed. This can be remedied by updating to current firmware revisions, at the date of posting I'd recommend downloading the ST-LINK Utilities v3.9.0 and using that to update the firmware, this has solved issues I've seen with this.

I believe this sets grounds for what you did and why it worked.

\$\endgroup\$
  • 1
    \$\begingroup\$ Hello, Please avoid posting answer that are only links. The purpose of this site is to provide a repository of answers that could help others. Links will be broken eventually. \$\endgroup\$ – Blup1980 Mar 28 '18 at 15:30
  • 2
    \$\begingroup\$ from the link, the solution provided: There is/was a known issue with the ST-LINK firmware shipped with the F429I-DISC1 and F407-DISC1 boards, and perhaps others. The STM32 is held in reset by the debugger when not attached to a host computer, so charger/battery use cases failed. This can be remedied by updating to current firmware revisions, at the date of posting I'd recommend downloading the ST-LINK Utilities v3.9.0 and using that to update the firmware, this has solved issues I've seen with this. \$\endgroup\$ – Pop24 Mar 28 '18 at 15:49
  • \$\begingroup\$ In revised form this seems like a useful answer. While the reset issue has been mentioned before with the suggestion to disconnect it, that has consequences and a software fix as mentioned here seems more advantageous. \$\endgroup\$ – Chris Stratton Mar 28 '18 at 17:12
  • \$\begingroup\$ Upgrading using ST-Link Utility v4.3.0.0 from firmware V2.J27.M15 -> V2.J32.M22 worked for my STM32F407-DISC1 board. \$\endgroup\$ – phoenix Dec 6 '18 at 18:22
1
\$\begingroup\$

I've figured this out (or rather I've found a solution), after pouring over the schematics and some general diagnostic work (measuring voltages, shorts etc)

In the user manual UM1670 (by ST) it has a page (19) on solder bridges

"SB10 (STM_RST) OFF No incidence on NRST signal of STM32F429ZIT6. ON NRST signal of STM32F429ZIT6 is connected to GND."

The solder bridge is not connected by default but as it was related to a STM reset I decided to short this to see if I could get the whole system to reset. It worked.

I've since soldered SB10 together permanently and the board runs as expected. (I'm not sure exactly how this fix works I need to read into it and then edit this response)

IMPORTANT NOTE: YOU WILL NO LONGER BE ABLE TO FLASH CODE TO THE DEVICE THROUGH USB ONCE YOU HAVE SOLDERED SB10, I'VE NOT YET CHECKED BUT IMAGINE DESOLDERING WOULD ALLOW FOR YOU TO PROGRAM THE MCU AGAIN

\$\endgroup\$
  • \$\begingroup\$ I suspect either you or what you are quoting from has mixed something up here, because as literally stated you are saying you connected NRST of the target MCU to ground. Doing so would prevent it from operating at all. Unlike some other ARM Cortex implementations, an STM32 does not normally require manipulation of the NRST line to achieve SWD programming, though it can be needed to recover from garbled states or a program which repurposes the SWD lines for some other purpose immediately on startup. \$\endgroup\$ – Chris Stratton May 24 '16 at 16:53
  • \$\begingroup\$ The SB10 solder bridge is connected to the NRST pin of the STM32F103CBT6 (not to the STM32F429ZI), which is the on-board ST Link actually. That's why he cannot flash the MCU anymore. \$\endgroup\$ – Bence Kaulics May 24 '16 at 19:03
  • \$\begingroup\$ @BenceKaulics yes you're correct. ChrisStratton I'm not sure why the user manual links SB10 with the main processor on the board but Bence is correct, its actually linked to the ST-link processor on board \$\endgroup\$ – Pop24 May 25 '16 at 8:04
  • \$\begingroup\$ It appears there is not an updated firmware for the on-board STLINK which fixes this issue. \$\endgroup\$ – Chris Stratton Mar 28 '18 at 17:11
1
\$\begingroup\$

A reddit post I read (https://www.reddit.com/r/stm32f4/comments/4f6uaq/stm32f4_discovery_external_power_supply_not/) suggested that you update your firmware. I did this and my programs run now. However LD1 still flashes.

\$\endgroup\$
  • \$\begingroup\$ Can you specify which version you updated to? \$\endgroup\$ – skrrgwasme Jul 8 '16 at 10:33
  • \$\begingroup\$ LD1 will probably flash until the programmer is enumerated. It's annoying, and has actually caused me problems \$\endgroup\$ – Scott Seidman Jul 8 '16 at 11:06
  • \$\begingroup\$ The same one as in the reddit link. V2.J27.M15 \$\endgroup\$ – Spu Jul 8 '16 at 12:52
0
\$\begingroup\$

LD1 will flash slowly as long as there is no USB connection to the programmer. It has actually caused me problems in the past, forcing me to remove a resistor to turn it off.

You are powering correctly, or at least to the correct header pin.

\$\endgroup\$
  • \$\begingroup\$ Yeah I had seen a post of yours regarding this. Whether the LED flashing is causing issues for my program remains to be seen as it doesn't actually start. Ill check out your post and see which resistor it is and try remove that perhaps. thanks \$\endgroup\$ – Pop24 May 24 '16 at 11:23
  • \$\begingroup\$ Thanks for the heads up on powering correctly, aprreciated \$\endgroup\$ – Pop24 May 24 '16 at 12:11
  • \$\begingroup\$ @Pop24 the flashing LED probably doesn't even involve the microcontroller, just the programmer \$\endgroup\$ – Scott Seidman Jul 8 '16 at 23:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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