Has someone encountered issues using dual bank flash on random chips?

[YCoqu.1]

Our design and firmware used to work with former batch but a current lot of STM32F427 does not have a proper readout of the flash size and does not switch from bank A<->B.

The STlink reads the following on a working chip:

STM32 ST-LINK CLI v3.2.0.0

STM32 ST-LINK Command Line Interface

ST-LINK SN : Old ST-LINK firmware/ST-LINK already used

ST-LINK Firmware version : V2J14S3 (Need Update)

Old ST-LINK firmware detected!

Please upgrade it from ST-LINK->'Firmware update' menu.

Connected via SWD.

Target voltage = 3.3 V.

Connection mode : Normal.

Device ID:0x419

Device flash Size : 2048 Kbytes

Device family :STM32F42xxx/F43xxx

0xE0042000 : 19 64 01 20

0x1FFF7A22 : 00 08 AE 4D

and reads this on a failing bank swap chip:

STM32 ST-LINK CLI v3.2.0.0

STM32 ST-LINK Command Line Interface

ST-LINK SN : Old ST-LINK firmware/ST-LINK already used

ST-LINK Firmware version : V2J14S3 (Need Update)

Old ST-LINK firmware detected!

Please upgrade it from ST-LINK->'Firmware update' menu.

Connected via SWD.

Target voltage = 3.2 V.

Connection mode : Normal.

Device ID:0x419

Device family :STM32F42xxx/F43xxx

0xE0042000 : 19 64 03 20

0x1FFF7A22 : FF FF D0 A0

The register 0x1FFF7A22 should be factory-programmed but reads 0xffff instead of 0x0800 !

The DBGMCU_IDCODE, Address: 0xE004 2000, suggests a die revision 3 (20 01) in the first working chip and a die revision 5 or B for the second one (20 03).

Could the die revision affect the dual bank features?

Can the flash_size be overwritten?

Or should I think of a clone chip?

Thanks for your inputs!

YCoqu

[Tesla DeLorean]

>>Can the flash_size be overwritten?

More of a case of the System Memory NOT being written, which is often the case for Engineering Samples. This could be the case where a mask-set has been brought up at a new fab. I think some of the F4 that previously came out of a French fab, now come out of one in Taiwan as well.

You should perhaps open an Online Support Request via the support portal.

[Tesla DeLorean]

This looks to be a similar situation with a part out of TSMC 14 Taiwan, there's been at least one other report I've seen within the last few weeks, or month, but the search engine here is dire..

https://community.st.com/s/question/0D53W00001PuyY4SAJ/hello-everyone-this-trying-to-program-some-stm32f427-vit6-through-dfu-but-usb-windows-does-not-recognize-the-board-tried-different-cables-and-computers-i-have-also-tried-to-program-it-using-stlink-programmer-message-revision-id-unknownsome-help

@Amel NASRI​ can you double check what happen with the F427 parts here, if there's a batch or range-of-weeks we should be look out for. Thanks -Clive

2020 Week 44 ?

Related PCN

https://www.farnell.com/datasheets/2626118.pdf

https://www.anglia.com/registration/pcn_ptn/docs/pcn/st_PCN%20MDG_19_10816_Report.pdf

[Tesla DeLorean]

This was the other one

https://community.st.com/s/question/0D53W00001xnbnJSAQ/chinese-stm32f407vet6-fakes

[YCoqu.1]

Thnaks Tesla!

I have a ticket open since 28 november but you are much more reactive.

I will check which batch dat code I can retrieved so far these marking

STM32F427

VIT6

9909X VQ

MYS 99 121 4

ST e3 ARM

04

Or

STM32F427

VIT6

990LD 9R

MYS 99 011 5

ST e3 ARM

01