stm32f407vg spi1 data register stay 0xff

[Jalal Sadigli]

Posted on May 02, 2017 at 10:53

hi, i am new to community.....

i have a stm32f4 discovery board with sym32f407cg microcontroller i tried to test spi communucation of the microcontroller with onboard accelerometer

but there is one or more problem that i can not hande

the problem is whatever i write on data register of spi1 , the register only shows 0xff when i read it with the help of pointer

how can i solve this problem

thats what i did as settings of spi1..

spi_cr1 |= 0x30; // select speed

spi_cr1 |= 0x03; // select cpha and cpol spi_cr1 |= 0x00; // select dff spi_cr1 |= 0x00; // select msbfirst

spi_cr2 |= 0x10; // select frf

spi_cr2 |= 0x04; // set ssoe spi_cr1 |= 0x04; // set mstr spi_cr1 |= 0x40; // enable spi##discovery ##spi #stm32

[Tesla DeLorean]

Posted on May 03, 2017 at 17:30

You'd want to make sure the correct chip select GPIO pin is being used, and that you can read back the content of the CHIPID or WHOAMI type register.

[Jalal Sadigli]

Posted on May 05, 2017 at 07:43

the board exactly the same with the image you upload ... the version is MB997D 

[Jalal Sadigli]

Posted on May 05, 2017 at 07:44

it the Pin E3 that i am using for chip select for spi1

[Jalal Sadigli · Accepted solution]

Posted on May 07, 2017 at 19:22

i have solved the problem....

the problem is spi does not communicate correctly because of reasons

***the ti frame format has its own cpha cpol settings which controlled by hardware and can not controlled through cr1 register

***rxne and txe flag should be checked with busy flag . because txe flag set after data transfered to shift register but transmission is not finnished

and also before setting cs pin  low the busy flag must be checked

[Andrei Chichak]

Posted on May 07, 2017 at 20:14

Each slave chip manufacturer is free to choose the cpha and cpol settings. Each master manufacturer is free to choose how to express the phase and polarity. 

Beware, there is a bunch of documentation problems with this accelerometer.

I wrote a series of articles on SPI in general and the LIS3DSH on the F4Discovery in specific. See 

http://embedded.fm/blog/2017/4/26/i-apply-spi-good-vibrations

 and the two previous articles.

Andrei