AN5094 comments

[waclawek.jan]

https://community.st.com/s/question/0D50X0000CBn2icSQB/nucleog431kb-help-with-dma-gpio-output

JW

[waclawek.jan]

The waitstates table in 4.5.1 Performance versus Vcore ranges also does not reflect the current state of affairs.

https://community.st.com/s/question/0D53W000004JX9fSAG/stm32g474-flash-wait-states-at-170mhz-cpu-clock

JW

[waclawek.jan]

Comparator chapter might mention that the 'G4 COMP is much faster (propagation delay typ. 16.7ns vs. typ. 90ns on 'F303xB/C at the fastest setting), but also has much higher power consumption (typ. 450uA vs typ.1.2-75uA on the 'F303xB/C depending on setting).

JW

[Imen GH]

​Hello,

I take your feedback into consideration and I will raise this internally.

Thank you Jan,

[waclawek.jan]

Thanks, @Imen GH​ .

JW

[hamdi.najoua]

​Hi Jan,

You are totally right! The migration AN should be updated to consider the new G4 WS specification.

This will be considered in next AN5094 version.

Thanks for the input.

Regards,

Najoua.

[hamdi.najoua]

​Hello Jan,

I think we can provide simply a link to the F3/G4 comparator application note:

https://www.st.com/resource/en/application_note/dm00074240-getting-started-with-analog-comparators-for-stm32f3-series-and-stm32g4-series-devices-stmicroelectronics.pdf.

FYI:  There are other STM32F3 devices coming with a faster comparator than the one we have in the F303xB/C.

Best regards,

Najoua.

[waclawek.jan]

Hi Najoua,

> I think we can provide simply a link to the F3/G4 comparator application note:

Yes, but then it would be appropriate to give at least the typ. values in that AN4232.

And I still believe it's also appropriate to put the comparison and typ. values to the migration document - after all, the analog and semi-analog functions including comparators are supposed to be the major points of 'F3 and 'G4, don't they.

Imagine you are an engineer who recommends migration based on superficial study of the differences (eg. being impressed by the 'G4 benchmark calculation power); just to find out after the fact, the comparator draws significantly and quite surprisingly much more power.

Your documents should serve by providing straightforward and fair information, not hide inconvenient facts.

> FYI: There are other STM32F3 devices coming with a faster comparator than the one we have in the F303xB/C.

So, you are revisiting the 'F3 range, or is this a typo?

JW

@hamdi.najoua​ 

[hamdi.najoua]

​Hi jan,

Ok, we will add typical values in the migration application note.

Regarding the second point, it is not a typo, I mean STM32F3: STM32F303xD/E ; STM32F303x6/8, STM32F302x6/8 are coming with a faster comparator than F303xBC. In these products, the comparator speed is as follow:

[waclawek.jan]

Ah, I see, thanks. I wasn't aware of that; having used only the 'F303B/C, it didn't occur to me there may be such a difference *within* the 'F3 family.

So, it's the '303BC which is really different, not the 'G4... Still IMO worth mentioning, in both abovementioned appnotes.

Thanks again,

Jan