Query: STM32F4 programming with USB

[Mike xx]

I can program my bespoke board using a repurposed USB port. It would normally be used by a mouse.

I was under the impression that if I powered up the board with BOOT0 pin held high the MCU would be in a state to connect to USB and do it's enumeration. However, using USB Device Tree Viewer I can see the 43 error.

If I then hold the NRST pin low for a few seconds, and then release enumeration follows and I am able to program the device.

Is this expected operation? I thought simply powering up the PCB with BOOT0 high would place the MCU in the correct state ready for programming?

[mƎALLEm]

Hello,

You need to reset the MCU after holding high the BOOT pin .

[Mike xx]

Thanks for confirming. I can't find any programming guide or application note that says that, hence the question.

Furthermore, I find for reliable connection I have to power up in reset, remove the reset link and then attach the USB cable to the socket.

In fact the whole process of connecting is a little unreliable depending on port and / or the USB hub.

[mƎALLEm]

Check you cable and your hardware.

You can either use reset or power cycle the MCU after pulling up the BOOT pin.

You can also refer to these articles:

DFU mode with system bootloader is not working, while USB does

FAQ STM32 boot process

How can I use STM32CubeProgrammer to access the USB-DFU bootloader on my STM32 board?

Hope that helps.

 

[TDK]

Powering up with BOOT0 held high should be sufficient. Might be running into an issue with the crystal having a startup time.

[Mike xx]

I find this using this bootloader process unreliable.

I can put a frequency meter on the 8MHz crystal. Whether it bursts into oscillation is hit and miss. With the application running, ie BOOT0 low, it always starts up.

If I hold the BOOT0 pin high with NRST high after power-up and the USB cable disconnected I can see the 8MHz crystal running

If I apply power with BOOT0 pin high and NRST low, then allow NRST go high there is no 8MHz oscillation until I connect the USB cable. It seems the longer the PCB is powered the process becomes more reliable.

I see a comment in AN2606 regarding temperature:

  "When (because of temperature variations or other conditions) the internal oscillator precision is altered above the tolerance band (1% around the theoretical value), the bootloader can calculate a wrong HSE frequency value. In this case, the bootloader DFU/CAN interfaces can malfunction, or not work at all."

That doesn't instil much confidence in reliable programming. I was hoping to remove the JTAG connector from the build, but JTAG seems the most reliable way to go.

[TDK]

Perhaps the crystal selection or design could be improved.

Guidelines for oscillator design on STM8AF/AL/S and STM32 MCUs/MPUs - Application note

What you're doing should work on ST boards.