test for real stm32f103 chips

Anything not related to STM32
Post Reply
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

test for real stm32f103 chips

Post by dannyf »

Interesting read https://hackaday.com/2021/06/23/test-yo ... 103c8-mcu/.

I was surprised that the project cited https://mecrisp-stellaris-folkdoc.sourc ... -v1.6.html didn't test for the core's version -> CPUID: https://dannyelectronics.wordpress.com/ ... f10x-chip/

One of the links led me to here: https://hackaday.com/2020/10/22/stm32-c ... -the-ugly/

and it stated that STM32FEBKC6 is not a legit ST part, :). So maybe we shouldn't trust those guys too much.
ag123
Posts: 1655
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: test for real stm32f103 chips

Post by ag123 »

my guess is the 'real clones', are actually some arm variant where the io parts are reverse engineered.
the fake clones won't even work correctly. these days for the 'real clones' they would tell you the chip sku e.g. cs32f103 etc
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: test for real stm32f103 chips

Post by dannyf »

(unless they are outright fake chips) clones use the real ARM core - often r2p1, vs. r1p1 by the real thing. so technically the cores on those clones are newer (and better?).

the legality of copying someone peripherals and header files and naming conventons... isn't clear to me. I haven't seen people testing that - for example bringing the i2c bugs over to the clones.
ag123
Posts: 1655
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: test for real stm32f103 chips

Post by ag123 »

i'd guess for the 'real clones', some of them do have variations from the 'real' stm32. e.g. accordingly cs32f103 used the chipid of some stm32f4xx chip rather than stm32f103, that is made known in the issue logs for the open sourced st-link utility, accordingly it causes some conflicts and the support for cs32f103 is revoked at some point. i'm not too sure if it is added back. but otherwise a lot of functionalities do look / work similar. i think someone did mention variances in the flash erase/writing behavior as well.
mrburnette
Posts: 633
Joined: Thu Dec 19, 2019 1:23 am
Answers: 7

Re: test for real stm32f103 chips

Post by mrburnette »

dannyf wrote: Fri Jun 25, 2021 6:50 pm ...
the legality of copying someone peripherals and header files and naming conventons... isn't clear to me. I haven't seen people testing that - for example bringing the i2c bugs over to the clones.
In the U.S. the final statement on "conventions" may come from the Supreme Court's ruling on Oracle vs Google over the use of JAVA syntax and library calling conventions.

But, there are zillion of business-to-business understandings and contracts and licenses; which is to say, that (hypothetically) there is nothing from preventing STM from offering a license to the silicon peripherals used in the STM32Fxxxx chips. The F103 is an old chip, so any additional revenue stream may be beneficial to the bottom line.

I suspect that the functionality of some clone chips is being managed with microcode and that could easily lead to major issues with timing, memory, switching, and even output current capabilities: sink/source.

At some (pain) point, hobbyists will likely just stay away from the older versions of clone chips... not worth the effort or the cost to continue to go down an unknown road.


Ray
ag123
Posts: 1655
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: test for real stm32f103 chips

Post by ag123 »

for 'clone' chips, i think there are a couple of 'camps' using them
- 'hobbyist' who don't really mind the differences and want to experiment with them.
- possible manufacturers who use stm32 chips but wanting a 'cheaper' alternative. these days literally you just get the gadgets and no one knows what is 'inside' (nor bother to find out), just that 'it works'. normally there is a bottom line for this, the manufacturer is probably not concerned about continuity etc but just want a product to sell, and owns no responsibility for the failure of the product. this could be for example 'clone blue pills', apparently some vendors didn't even bother if it works. that has happened, some are outright fakes (not even clones) and don't work.
we'd really not know what is driving all that gadgets these days, if alter all they are really 'actual' chips or some other '3rd party equivalent' chips.
it is quite notable that even for keyboards, some unnamed arm m0 microcontrollers tend to be used for the 'higher end' chips, possibly as part of a contract manufacturing arrangement, the chips are unbranded white labelled.

for beginners etc, it is probably better to stick with a real original product, as after all one'd not be able to tell if that is a 'bug' or *feature* 'standard behavior' if a 'clone' chip is used and you have only a single piece of it and no other reference.

and as for manufacturers apparently those who sell boards at a premium are more than willing to tell you it is a (real) stm32, capitalizing on the st brand itself, this i observed for some 3d printer boards. apparently the 'new' 3d printer boards are dominated by stm32 and some arm based mcus.
they possibly simply run the 32bit marlin variant, common 'well known' names are like those 'reprap' printers from manufacturers like ender, twotrees etc.
recently the 'high end' 3d printer boards are taken off direct sales (including the stm32 ones), i think the chip shortage is partly a reason.
the printer manufacturers needs the chips and boards and won't be willing to spare you just the boards.
mrburnette
Posts: 633
Joined: Thu Dec 19, 2019 1:23 am
Answers: 7

Re: test for real stm32f103 chips

Post by mrburnette »

ag123 wrote: Mon Jun 28, 2021 3:16 pm ...
for beginners etc, it is probably better to stick with a real original product, as after all one'd not be able to tell if that is a 'bug' or *feature* 'standard behavior' if a 'clone' chip is used and you have only a single piece of it and no other reference.
...
+1

... and but 2x boards to facilitate hardware testing... it is easy to damage a uC on a breakout/evaluation board.
Post Reply

Return to “Off topic”