Receiving 16 bits from TWI

[Gabriel Bastos]

Hello,

I'm trying to receive a 2-byte data from a sensor from Maxim Integrated MAX31875 but the nrf_drv_twi_rx function only receive 8 bits data. Does this function to receive a 16 bits data exists or how can I change this function to receive a 16 bits data ? 

[Laurent Bourassa]

Put your received data in a int buffer of size 2 and in the lenght field of the read function, just write sizeof(rx_buffer).

I have used a similar function like so :

nrfx_twi_rx(&adc_twi,ADC_ADDR,&rx_buffer[0],sizeof(rx_buffer));

[Gabriel Bastos]

Hello,

I've tried but without sucess, I got zero as result. In fact, I was in doubt why you have used int as type of the variable, because this sizeof(int variable) is 4 byte and I want to have a 2 byte variable.

[Laurent Bourassa]

Hello, I actually used type uint8_t to avoid confusion, this will specify the length of the variable to one byte exactly. Your variable should be an array of size 2.

uint8_t rx_buffer[2];

nrfx_twi_rx(&adc_twi,ADC_ADDR,&rx_buffer[0],sizeof(rx_buffer));

Hence, once read, your data will be stored like this :

rx_buffer[0] = first byte of data

rx_buffer[1] = second byte of data

Then you should be able to concatenate both 8 bits data into one 16 bits data like so :

uint16_t complete_data = ((uint16_t)rx_bufer[0] << 8) | rx_buffer[1];

I hope this better explains how I did it.

[Gabriel Bastos]

Awesome Laurent, thank you very much, this trick solved my problem! Do you know how can I change the bit LSB to have the (2^(1/256)) value ? , for example the normal LSB is 2^0 = 1 but I want the LSB values (2^(1/256)).

[Laurent Bourassa]

What if you made your complete data a float type instead? I don't know a lot about how floating points work but I think I've found something that might help you :

https://ryanstutorials.net/binary-tutorial/binary-floating-point.php