VL53L8CX SPI communication

[bastl]

Hi all,

I'm trying to setup the SPI communication with a VL53L8CX which is connected to a STM32F401RE-Nucleo. I took some code snippets from the STSW-IMG040 package and I wonder why the bytes to be send in VL53L8CX_WrMulti and VL53L8CX_RdMulti are treated with a special mask instead of just separating the bytes like shown below:

#define SPI_WRITE_MASK(x) (uint16_t)(x | 0x8000)  // 1
#define SPI_READ_MASK(x)  (uint16_t)(x & ~0x8000)

data_write[0] = SPI_WRITE_MASK(temp) >> 8;
data_write[1] = SPI_WRITE_MASK(temp) & 0xFF;

       respectively

data_write[0] = SPI_READ_MASK(temp) >> 8;
data_write[1] = SPI_READ_MASK(temp) & 0xFF;

     instead of just

data_write[0] = (temp >> 8) & 0xFF;
data_write[1] = temp & 0xFF;

Is there a special reason for doing this?

Regards
Bastian

[John E KVAM · Accepted solution]

On an i2C, if the sensor has a base address of 0x29, one shifts the base address one bit and uses the LSB as a write/read. So write to address 0x52 and read from address 0x53.

According to Bing A/I, "In SPI communications, there is no specific "read/write" bit. Instead, the first byte contains the SPI address, and the following bytes contain the data. The first bit of the first byte indicates whether it is a read (1) or write (0) operation¹²³."

So that mask sets (or clears) the MSB.

and the rest is to swap the bytes as SPI (and I2C) are big-endian and the STM32 is little endian. 

- john