MCUs: We want your knowledge base suggestions!

[Laurids_PETERSEN]

Dear community members,

We invite you to provide your suggestions for knowledge base articles in this thread. We strive to create articles that can assist or inspire you in the best possible way.
Anything goes, the only condition is that it is related to STM32 MCUs.
Simply create a comment in this thread or send me a private message if you are more comfortable with that. If you are interested in suggestions that other members have made, please use the “kudos” button. This helps us identify the most popular requests.

The following guidelines can help you think of suggestions that provide value to you or others:

• Implementing a new hardware or software functionality in your design/project
  
  
• Best practice guidelines for specific products in various scenarios to improve efficiency
  
  
• Explanations/guides on frequently asked questions that could provide benefit to all community members

We will review your suggestions and send a PM if any further information is required to avoid bloating this thread. This thread is active until the end of February 2024. Should you discover this thread after February 2024 and you have a suggestion you are always welcome to send me a PM.

Best regards,
Laurids

[Tesla DeLorean]

I don't have PM's enabled, you can use LinkedIn if you want to discuss, or my email.

What I think would be useful is some material on pivoting to different displays via LTDC/DSI, ST tends to stick to displays on DISCO and EVAL boards, but the real world is more fluid than that, so there needs to be some more proactive work in generating drivers for different and common display controllers. And then integrate into TouchGFX, or simpler tools.

Yes, other vendors parts, but by enabling on STM32 you're greasing the rails so the STM32 is the obvious and flexible choice for the customer's platform.

With enough clean examples you provide comfort to those unsure about what they are doing to given them confidence to try. Much of this stuff is more alike than it is different, but with something close enough it's going to be easier to port or triangulate to a solution. If there's already a solution that gets 90-95% of the job done, people are generally a lot less intimidated by the task that's left to them.

Again I think it's important to extend the bounds of what ST builds and ships, and into hardware customers want to integrate in the real world, where the cycles might be 12 to 18 months.

[waclawek.jan]

What's the budget?

JW

[Tesla DeLorean]

ARM/ACORN had two advantages, Hauser says: it had no people and no money.

As someone who grew up in the UK technology business market of the 1980's I can attest to this being the reality for most.

ST has a lot of money, and a lot of people. I'd have thought enough of both to be able to read the tea leaves and come up with a strategy here. I think the most powerful thing I can suggest is to step outside the internal use cases, and look at the real world ones.

I'm comfortable with transoceanic navigation of my embedded world, but there's a vast number who need to see the shore line at all times.

[Pavel A.]

@Laurids_PETERSEN Why not use the MCU wiki? It seems confusing to have both the wiki and KB for the same content?  Also, there should be a place for curated list of projects that the users generously share  on github etc - such as fixed/improved Cube examples, ports to different boards etc.  - the wiki can be good for this as well.

Wiki pages can be commented by others, so errors can be corrected and abandoned articles can be updated and improved by ST people and community.

[waclawek.jan]

> ST has a lot of money, and a lot of people.

That does not mean they genuinely intend to throw them at some REAL stuff. I'm not sure they realize how VERY expensive that is (I've had a "is there money on documentation" post when ST bought Atollic but that was jive and got dumped throughout the times, the reply was reassuring but we haven't seen the promised influx of documentation nor the ST employee who answered me that, ever since). We are talking about man-*weeks* of work on one "article" (should be appnote, really); and also somebody to orchestrate the whole effort to a coherent whole. But I suspect the plan is to throw some ideas on some overworked support crew, a particular case to be whipped up in two spare afternoons. That was the nature of most articles we've seen in the salesforce version under knowledge base; and unfortunately even some of the newer appnotes. Lacking both depth (as in going into the minute details of given peripheral/use case) and breadth (as in discussing portability across families, relationships to other IPs, etc.) This also requires dedication from ST as a whole, again, a tech support engineer who is not given access to the intimate internals and works out of the same DS/RM we do is not likely to produce a thrilling document.

So again, what's the budget?

JW

PS. > strategy ... and seeing shores.

Yes, yes, wholeheartedly agree.

PS2. Roughly one good topic per week comes up here, it's enough to listen. What about agreeing on some way to flag such threads?

PS3. Want example? The 'L0 low power fabric, https://community.st.com/t5/stm32-mcus-products/stm32l010k8-on-off-adc-vrefint-in-low-power-run-mode/td-p/571769 "It's an established product, effort spent in AN would not be justified."

[Pavel A.]

@waclawek.jan  man-weeks... The sort of work that you've described IMHO calls for the AI. Imagine an editor that never loses attention, never forgets, never retires, knows perfectly all needed languages, knows enough of electronics and CS to actually understand the text, has digested all erratas and so on. This is already not a fiction. If ST won't employ this, their competitors will certainly )) Yes this thing can be *** expensive given the demand, but it will be worth the money.

[STOne-32]

Hi @Pavel A. 

Interesting comment , we need to have an FAQ : Knowledge base 

Q: What is the difference between a Wiki and KB and purpose of each for STM32 ?

A : @Laurids_PETERSEN  may help .

Ciao

STOne-22 

[STOne-32]

Hi @Pavel A.  @waclawek.jan 

All it is has a cost either at ST side or at customer side . We are  seeing that current KB are very appreciated by our customers even the pace is not at the level we target, as it has a cost in $ by KB 🙂 instead of  xx/day. your feedback is very appreciated. Example we can give you the opportunity to flag a thread as a “KB Idea” with a label and then periodically to harvest it and publish . Let us know 

Ciao

STOne-32

[waclawek.jan]

@Pavel A.,

No, exactly the opposite. The most valuable knowledge is what the glorified web search engine a.k.a. AI does not have in it's database.

JW