STM32: original or fake?

[moffet]

Can you figure out from the attached photos is it the original STM32F103C8T6 chip?
Maybe someone can compare it with their version.

I don't like the poor STMicroelectronics logo, but maybe it's supposed to look like that in newer markings.

Thank you

[ataradov]

Unless ST changed how they mark the devices, it looks fake. Legit ST devices have ST logo on the bottom and I have never seen one have this matrix code. I think they only put this on modules, not individual ICs.

ST has TN1433 document that details their package markings. I don't see why they would deviate from that.

Molding quality also looks like shit, but that may happen with different assembly sites or equipment changes.

[wek]

Unless ST changed how they mark the devices, it looks fake. Legit ST devices have ST logo on the bottom and I have never seen one have this matrix code.

https://www.mouser.com/PCN/STMicroelectronics_24_14718.pdf

@OP: Datecode of the chip on picture is week 20 2023 though. And as @ataradov said above, lasering and molding looks sloppy.

And if it's a 'F103 and you have an urge to ask if it's fake, it's most probably fake.

Try to find the "blue pill most likely fake" thread here for more info, if you feel so.

But probably the best option is to avoid untrustworthy sources and also avoid the 'F103.

JW

[ataradov]

So, those are either qualification samples that somehow got out or fakers got ahead of themselves and used the new marking a whole year ahead of time.

And I don't see a reason for them to fake the new markings. So, I would assume those are just some random test devices that were not intended to be shipped out.

[Sacodepatatas]

Unless ST changed how they mark the devices, it looks fake. Legit ST devices have ST logo on the bottom and I have never seen one have this matrix code. I think they only put this on modules, not individual ICs.

A couple of.weeks ago i received some STM32F401RET that also have a QR code on the marking. It is normally assumed that F4XX aren't being faked, but it might be the case as everything has a beginning.

[ataradov]

It is possible that they started to implement that PCN sooner than expected, but it is really strange.

I wonder how is the quality of the engraving? I can't read the data matrix code from the pictures here. It seems to be pointless with quality this crappy.

[ArdWar]

Laser marking getting worse seems to be rather universal. TI even forego printing their Texas logo altogether due to how bad it looks. It must be really much cheaper than previous marking process.
ADI/LTC also start to stop painting their marking.

[5U4GB]

Could it be to speed manufacturing, so you're taking 0.1s per device rather than 0.5s?  The linked TI doc looks like a fair bit of of detail has been dropped, and the ST data matrix code looks like someone tried to run at 2X what was designed for a speed of 1X.

[Ice-Tea]

You forgot the most important question: where did you get it from.

If not from Digi/Mouser/Farnell/... then you pretty much have your anwser...

[daisizhou]

There are too many counterfeit chips, so I bought the old chip directly. Although it is old, it can be guaranteed at least.

It does not affect amateur players. If it is a manufacturer, you can only buy from a regular seller

[moffet]

Thank you everyone for your thoughts.
I will just add, the MCUs came really well packed. In the sealed bag with big moisture absorber sachet and moisture tester inside.

[ataradov]

Came from where? What is this company?

Those devices are likely produced by ST, since there is no reason for fakers to use new marking scheme, it just attracts attention. But it is not clear if those devices were ever supposed to go to the customers. And if they are from some test run, there is a chance they were not calibrated properly, for example. And maybe ST ran out of devices with the old making and started shipping new devices earlier than promised. Or maybe distribution was supposed to hold them until that date and they still sold them.

Ultimately, only ST support would be able to answer what those devices are and if you are supposed to have them.

[mino-fm]

The yellow label looks like Schukat-Germany: https://www.schukat.com/schukat/schukat_cms_de.nsf/index/CMSDF15D356B046D53BC1256D550038A9E0?OpenDocument&wg=Y4773&kb=STM32F103C8T6#STM32F103C8T6

A serious seller since tens of years.

[wek]

At a different packaging site, the 2D code has been introduced by the end of 2022 so this can be legit.
https://www.mouser.com/PCN/STMicroelectronics_13318_merged.pdf

JW

[josip]

JLCPCB SMT from 2024-03-25

[ataradov]

Engraving quality is much better, but the code is still not readable.

[SiliconWizard]

Not that I want to trigger a flamewar of make false accusations, but I have suspected on occasion some parts from LCSC/JLCPCB not to be genuine, not just from their markings, but also from their performance (in my case, that were LDOs and opamps). Of course I don't have evidence that those were not genuine. Just suspicion.

[daisizhou]

This is normal. In China, LCSC/JLCPCB is about 20% more expensive than the market average price.
If you are a perfectionist, you can choose LCSC/JLCPCB。
Ordinary people usually choose Taobao shopping,It is inevitable to encounter fakes，Or second-hand refurbished items, sold as new

[fchk]

I suggest migrating to a newer chip like STM32F303 as soon as possible. I don't know of any fakes of the '303.

[SiliconWizard]

This is normal. In China, LCSC/JLCPCB is about 20% more expensive than the market average price.
If you are a perfectionist, you can choose LCSC/JLCPCB。
Ordinary people usually choose Taobao shopping,It is inevitable to encounter fakes，Or second-hand refurbished items, sold as new

Yes, but that wasn't my point. My point here was that precisely, I suspect LCSC/JLCPCB to also ship occasionally some non-genuine parts. Which is also what we can suspect from josip's post. But again, something I have suspected also on simpler parts like voltage regulators.

[DELTA67]

Have you tested it with this?
https://mecrisp-stellaris-folkdoc.sourceforge.io/bluepill-diagnostics-v1.6.html