Nucleo-144 SPI?

[Dick Lin]

Posted on June 07, 2018 at 17:49

Hi,

I am using STM32L496 Nucleo-144 connecting SPI1 to external Adafruit MAX31865 PT100 temp sensor using breadboard with Saleae.

For some reason (I have no clue at all), I got nothing from Saleae.

One thing I kinda of curious is the the pinout. From UM1279 User manual, the pinout are

- SPI_SCK PA5

- SPI1_MISO PA6

- SPI1_MOSI PA7

- SPI1_CS PD14

While from Cube the SPI1_NSS is PA4. 

I am assuming SPI1_CS and SPI1_NSS are the same (either Chip Select or Slave Select). Just not sure which one should I follow. I did try both, no luck at all.

Why I am not seeing anything from Saleae? I would think should be a very straight forward thing.

Thanks,

Dick

[Dick Lin]

Posted on June 07, 2018 at 20:26

Figured out we need to use PA4 for NSS.    PA4     ------> SPI1_NSS

And need to change the BaudRatePrescaler to 32.

  hspi3.Init.BaudRatePrescaler = SPI_BAUDRATEPRESCALER_32;

[Dick Lin]

Posted on June 11, 2018 at 18:21

Now the issue become can't get any data using HAL_SPI_TransmitReceive() function. The in_data always 0 while I am able to see the data from analyzer.

Any suggestions?

Thx

[Tesla DeLorean]

Posted on June 11, 2018 at 19:37

Got super low visibility into your code/functionality

Perhaps you're not configuring the pin properly? Clocking on the wrong edge?

SPI3?

PD14 should be viable as a GPIO chip select

[Dick Lin]

Posted on June 11, 2018 at 19:45

I am using PA4 connecting to slave CS and SPI_NSS_HARD_OUTPUT. That's the

only way to make it working instead of using PD

GPIO_InitStruct.Pin = GPIO_PIN_4|GPIO_PIN_5|GPIO_PIN_6|GPIO_PIN_7;

GPIO_InitStruct.Mode = GPIO_MODE_AF_PP;

GPIO_InitStruct.Pull = GPIO_NOPULL;

GPIO_InitStruct.Speed = GPIO_SPEED_FREQ_MEDIUM;

GPIO_InitStruct.Alternate = GPIO_AF5_SPI1;

HAL_GPIO_Init(GPIOA, &GPIO_InitStruct);

On Mon, Jun 11, 2018 at 10:37 AM, Clive One <st-microelectronics@jiveon.com>

________________ Attachments : image.png : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006HxWu&d=%2Fa%2F0X0000000azL%2Fd0v3zzeOZFgetzuDogXJp9wwPzDdOj7cxXw._.tYpc4&asPdf=false

[Tesla DeLorean]

Posted on June 11, 2018 at 20:20

Not sure of the SPI_NSS_HARD_OUTPUT functionality, STM32 parts don't do that super well.

Doing SPI on L4 using PA5/PA6/PA7 on a couple of other boards without problem, so shouldn't be inherently broken.

[Dick Lin]

Posted on June 11, 2018 at 20:37

I think we might see the issue. From the slave chip MAX31865, the CS should

go active low all the time until the transaction completed.

For the HAL_SPI_TransmitReceive() seems the CS go high in the middle of

transaction. Thx

On Mon, Jun 11, 2018 at 11:22 AM, Clive One <st-microelectronics@jiveon.com>

[Dick Lin]

Posted on June 11, 2018 at 21:03

When I sending 2 bytes to slave, the CS pin go high in between.

How come it not active low all the time until over? I even manual set the

pin, doesn't seems have any control.

HAL_GPIO_WritePin( GPIOA, GPIO_PIN_4, GPIO_PIN_RESET ); // NSS1 low

HAL_StatusTypeDef status = HAL_SPI_Transmit(m_spi, (uint8_t *)out_data,

len, 10);

while( m_spi->State == HAL_SPI_STATE_BUSY )

; // wait xmission complete

HAL_GPIO_WritePin( GPIOA, GPIO_PIN_4, GPIO_PIN_SET ); // NSS1 high

[Tesla DeLorean]

Posted on June 11, 2018 at 21:20

You should be able to control placement manually, not sure what your issue is. Must obviously wait for last bit of last byte to transit the bus first.

[David Littell]

Posted on June 11, 2018 at 21:21

Don't configure the NSS output as an alternate-function pin (which puts it under control of the SPI module), just make it a simple GPIO.