Code working on STM32F429I-Disco rev.MB1075C but not STM32F429I-Disco rev.MC1075E

[SSmit.0]

Hi,

I've been working with the STM32F429I-Disco board, rev.MB1075C for a while and recently purchased another which I notice is now at revision MC1075E.

The code which works fine on the older revision boards fails at the CubeMX generated SystemClock_Config(); function. This function is left exactly as is and has not been modified in any way.

I tried some other projects and they also fail at the same function call.

Could someone please clarify what is wrong here and why code written for an older generation board is seemingly incompatible with a newer one? I have briefly checked the solder bridge defaults and they seem to be the same.

Thanks

[waclawek.jan]

>fails at the CubeMX generated SystemClock_Config(); function

What does it mean? What are the symptoms, exactly, and how do they differ from expectations?

Do you use the MCO output from STLink? What else is different between the board revisions? There should be a list of changes, usually it's on the schematics.

JW

[SSmit.0]

Specifically this function within SystemClock_Config();

 if (HAL_RCC_OscConfig(&RCC_OscInitStruct) != HAL_OK)

 {

  _Error_Handler(__FILE__, __LINE__);

 }

where the debugger never moves past a break point placed at the line below, within the HAL_RCC_OscConfig call.

if((HAL_GetTick() - tickstart ) > HSE_TIMEOUT_VALUE)

I thought the problem may somehow relate to the ST-link firmware but I have downgraded this to the same version as working boards and it still isn't working.

The only things left I know of are bootloader and physical configuration of the board? I have checked and compared the datasheets and there doesn't appear to be any difference apart from an upgraded IC used for the gyroscope which I do not use. All solder bridge configurations are the same.

I have a blank CubeMX template I saved when I started this project in 2018. The blank default will also not even boot past these lines of code.

Thanks

[waclawek.jan]

I do see differences in solder bridges settings between C and E revisions.

Appending UM for rev. C , as ST removed it from the product page.

@Imen DAHMEN​ , can we please have the documentation for older revisions of the Nucleo, Disco and EVAL boards (and I mean all of them) kept available forever? People do have these boards, they are still useful and they may want to return to them from time to time. Shall I open an "idea" for this, will it increase the chance these requests remain ignored?

Thanks,

Jan

[SSmit.0]

Many thanks I had only looked at the solder bridge table, though not looked carefully enough. After bridging SB18 and removing SB20 if I remember, it it now working fine.

This sort of thing really should be mentioned somewhere as an NB within the manual.

[waclawek.jan]

> This sort of thing really should be mentioned somewhere as an NB within the manual.

+1

Thanks for coming back with the solution.

JW

[Imen.D]

Hello @SSmit.0​ and @Community member​ ,

Thank you for your proposal and for helping us improve our document !

I raised your request internally to the appropriate team, and it's under review.

I'll make sure to post updates when I have more information from them.

Imen

[waclawek.jan]

Thanks, Imen!

Jan

@Imen DAHMEN​ 

[Imen.D]

I've escalated internally this requirement to keep on st.com older revisions of a document, and I will give you an update here as soon as I hear back from our team.

So, thank you for your patience while we work on this.

Note that when removing the schematic from the UM, it is ensured that all schematics are available into the CAD resources section of the web page. 

For the STM32F469I-DISCO, there is an issue as only the latest schematic is available. I’ve thus asked the appropriate team to fix this problem as quicly as possible.

However, to find a way out for that situation and while waiting for the publication of the missing schematics, I suggest you to review in other packages (board design project file, BOM and manufacturing files) which included all the different board revisions. 

Imen

[waclawek.jan]

Hi Imen,

Thanks for working on this.

Just one remark, it's not just the schematics (although that helps too). It's also the example code, both sources and binaries (does the one published binary for 'F429 Disco work for both C and E revisions?). And also the manuals should inform the user of the changes, and should not remove old relevant information - see this very thread, where the default solder bridges changed significantly - I admit there's changelog in the manual, but IMO the description of this particular issue is not sufficient.

And it's not just the 'F429 DISCO - for example the 'F746 Disco had seen a quite significant revision, too; yet the corresponding UM1907 fails even to mention it, maybe except "Removed product references..." in the changelog. Okay, maybe the new parts were carefully selected to have the same behaviour; but what's the point in hiding this information from the user? These boards are used by engineers, information is our primary work tool.

Maybe the old manuals could be kept, too. I'd suggest a simple files repository, it's easier maintained than the website, and it would present one single link from the website. Again, we are engineers, please have some faith in us being able to cope with this sort of information.

I know this is hard work. Please try to look at it from the users' point of view, these boards represent a known-working use case. ST might want to try to lower the entry threshold into using them.

IMO, ST generally underestimates and underfunds the ecosystem around these boards (while probably spending significant sums by selling them at a loss) - but that's one of those long-term rants of mine you already know.

Jan

@Imen DAHMEN​