STM32CubeExpansion_LRWAN_V1.1.1 uses stale L4 HAL

[Tesla DeLorean]

Posted on August 26, 2017 at 01:17

The LRWAN tree contains a very old L4 HAL tree, doesn't support L496, also lacks support for NUCLEO 32 and 144 boards.

Ported in the V1.8.0 L4 Cube supplied HAL libraries.

Got about 90% through a port of the SX1272MB2xAS shield on to a STM32L496 Nucleo-144 board, currently stuck on the LPUART1 VCP connection.

#lora #lorawan

[Tesla DeLorean]

Posted on August 27, 2017 at 16:45

Ok, so have the LRWAN code running on the NUCLEO-L496ZG (80 MHz, 1MB FLASH, 320KB RAM)

Created a V1.8.1 L4 HAL tree by merging the V1.8.0 and V1.8.1 downloads. Add the salient parts into the LRWAN V1.1.1 tree replacing the V1.6.0 (28-October-2016) code provided. Pulled over Nucleo-144 board support files.

The HAL_Uart_Init() on LPUART1 throws an error at 9600 baud, 115200 baud worked. Need to review that in the context of the APB clocks and the Reference Manual. Sanitized the VCOM.C code.

Ported the shield pins from the NUCLEO-64 mappings to the NUCLEO-144 ones, added code to support extended GPIO banks.

Need to get an 32L469GDISCOVERY to port LRWAN to that.

http://www.st.com/en/evaluation-tools/32l496gdiscovery.html

 

[Tesla DeLorean]

Posted on August 28, 2017 at 06:31

As I actually have a handful of IoT boards, I ported to code over to that first.

http://www.st.com/en/evaluation-tools/b-l475e-iot01a.html

 

[Tesla DeLorean]

Posted on September 01, 2017 at 05:02

Ok, so LRWAN successfully ported to assorted STM32L4 boards using the SX1272 shield

NUCLEO-L476RG

NUCLEO-L496ZG

B-L475E-IOT01A