STM32L412: State of pin PB9 doesnt change, no matter the configuration

[VKuma.7]

Hello,

Thank you in advance for any answers and your time.

I am using STM32L412CBT6P for my application. There are 2 custom boards with the same part and the following problem exists in both of the boards. Also attached is the connections diagram to the MCU.

The problem: PB9 of the MCU is clamped at exactly 1.2V despite of any configuration. The configuration I tried include output high, low, open drain; input - no push pull, pull up and pull down; external interrupt. I also tried putting an external pull down and; lastly, I tried lifting the pin to not have any connection at all, but with no results whatsoever.

Even without any firmware the pin stays at 1.2V despite the reset state of GPIOs being input(?).

Any hint towards the solution would be appreciated.

Regards,

Vivek

[TDK]

Pin PB9 is pin 45. Pin 46 is VDD12, which is indeed clamped to 1.2V, or at least should be.

My guess is you will find the same issue with PB11 on your current board.

 

Perhaps you want a STM32L412CBT6 chip instead. They have different pinouts.

[VKuma.7]

Hello,

Thank you for the reply.

You are right about STM32L412CBT6 being the desired device and not STM32L412CBT6P.

They have the same datasheet (DS12469 Rev 7) and they have not mentioned the SMPS variant for the LQFP-48 footprint. Compared to proper details and pinout diagrams (page 50 - 54) and pin details table (page 56 onwards) for all the other versions with different pinouts.

Now I have manufactured about 100 boards with the wrong controller and it would all be a loss or require a lot off manual patch work.

Regards,

Vivek

[MFass.1]

Hi,

I fell into the same trap when I tried to use STM32L412CBT6. Here I'm using PB11 like USART3_RX.

In the datasheet (DS12469 Rev 8) there aren't explicit differences between versions with and without P for the LQFP48 package.

CubeIDE has the same problem, there aren't any pinout differences between these two kinds of Part Numbers.

Thanks.

[TDK]

> CubeIDE has the same problem, there aren't any pinout differences between these two kinds of Part Numbers.

Sure seem different to me in STM32CubeIDE:

 

> In the datasheet (DS12469 Rev 8) there aren't explicit differences between versions with and without P for the LQFP48 package.

The datasheet is missing the SMPS LQFP48 package pinout, but it's there for the LQFP64 and others.

[MFass.1]

You are right.

I was choosing the device by the generic Part Number, that automatically selects the version without P.

I didn't realize that was possible to choose the device by the reference.

My mistake. 

Thanks.

[bobaruni]

I too have manufactured a batch of boards and none of them work for the same reason.

[Tesla DeLorean]

Which part was the board designed for, and which did you use to build them?

Right now with the shortage of parts a lot of less popular versions are available from secondary sources, need to be very careful about part numbers and suffixes.

The -P ones being particularly troublesome due to pins moving around or being eliminated.

[Piranha]

After 2 years they finally added LQFP48 SMPS pinout in DS12469 Rev 9.

[bobaruni]

Thanks Tesla, the board was designed for STM32L412CBT6 but I manufactured a batch using STM32L412CBT6P thinking they are just a different packaging SKU as in tape and reel vs tray.

There are several variants of the L412CBT6 at Mouser and they all list 38 I/O for all of them which I now know not to be the case as 2 of those pins are used for the 1.2v supply in the P suffix.

Unfortunately my design used both of the supply pins as GPIO so it doesn't work, luckily I have been able to get some stock of the STM32L412CBT6.

An of course there is not a single mention of the STM32L412CBT6P or STM32L412CBT6 for that matter in the latest datasheet.