My Vcc-gnd.com STM32F407VET6 Mini issues

[alex31]

Hi,

After you hit reset and the board hangs, do you have connected a probe to see with gdb (you can attach to a running board) , where, in the init process, it is stuck ?

I have had a couple of boards, with out of spec crystal oscillator (or capacitor) which where stuck in the init process in the loop that check for HSE or LSE stability, and have found it using a gdb attach to the running board.

Alexandre

[Polux]

Hi,

Probably not related, but have you read the issues here https://wiki.stm32duino.com/index.php?title=Vcc-gnd.com_STM32F407VET6_Mini ?

Regards

Angelo

[russian]

I like the osc start-up issues theory! I had some osc startup issues on my other project it was painful

Another weird fact: chip is marked stm32f407vet6 and this is supposed to mean 512K flash size, eBay lot description says "512K" while all the tools report flash storage as 1M with the actual board. We were able to read and verify 1M of flash. Is this a fake stm32f407vet6? Is this a questionable stm32f407vet6? Is the whole board very very questionable ?

Probably not related, but have you read the issues here https://wiki.stm32duino.com/index.php?title=Vcc-gnd.com_STM32F407VET6_Mini ?

USB-related R8 is fine on my board I've checked, weak 3.3v source is a concern but I hope it should not be an issue while powered by ST-LINK 3.3v

[Giovanni]

It is possible it does have 1M and ST sells it as a 512kB, the only different thing could be a letter stamped on the package.

Test it.

Giovanni

[russian]

Status update: we are deep in the weeds and still missing something about virtual USB serial driver: some ST non-ChibiOS example works but chibiOS virtual serial does not work.

We have already tried swapping stm32 chip itself just in case and things are STILL not good, at least we know it's about the board not about a fake chip.

[Giovanni]

Is BOARD_OTG_NOVBUSSENS defined or not in your board.h?

Giovanni

[russian]

Between myself and forum member andreika we now have 7 of these boards.

I cannot explain it, but mine have all started to behave more reliable in terms of firmware start-up - I get reliable LED bilking all the time, and reliable USB serial console in 99% of the cases

andreika who uses same Windows 7 64 bit and same version of ST VCP driver reports "unrecognized USB device" behavior unless the following two lines are removed

Code: Select all

        usbDisconnectBus(serusbcfg.usbp);
   chThdSleepMilliseconds(1500);

I've inherited these lines from ChibiOS USB demos. I do not have a great explanation for this but that's my status update. We track this issue under https://github.com/rusefi/rusefi/issues/705

[Giovanni]

Those lines force a USB device disconnection and reconnection, are there mainly for debug.

How is the disconnect signal handled on that board? is it similar to discovery?

Giovanni

[russian]

https://wiki.stm32duino.com/images/6/69 ... ematic.pdf shows USB Vbus routed into 5vin and not routed into stm32 pa10.

[Giovanni]

Hi,

Try defining BOARD_OTG_NOVBUSSENS in your board.h.

Giovanni