SPI Data Register is Always 0 on my STM32F302CBTx I have a code working on STM32L476VETx, it reads SPI data from AFE4400 Chip. Now I want to use this code on STM32F302CBTx (SPI2) but it does not work since I always get zero on data register.

[SLONK.1]

Here is the code:

uint32_t AFE4490Read (uint8_t address){

uint32_t data = 0;

uint8_t buf[3];

HAL_GPIO_WritePin(AFE_STE_GPIO_Port, AFE_STE_Pin, LOW);    // enable device

HAL_SPI_Transmit(&hspi2, &address, 1, 1000);      // send address to device

HAL_SPI_Receive(&hspi2, buf, 3, 1000);         // SPI.transfer (data);

data |= (((uint32_t)buf[0])<<16);            // read top 8 bits data

data |= (((uint32_t)buf[1])<<8);            // read middle 8 bits data

data |= ((uint32_t)buf[2]);               // read bottom 8 bits data

HAL_GPIO_WritePin(AFE_STE_GPIO_Port, AFE_STE_Pin, HIGH);    // disable device

return data;                      // return with 24 bits of read data

}

[Amira]

Hello @SLONK.1​ 

For STM32 HAL SPI driers there is no a noticeable difference between L4 and F3 series.

Did you verified the configuration you used for F3 from datasheet ? ( Clock, GPIO Pins, ..)

Regards,

Amira

[SLONK.1]

Hello,

Thanks for your response.

Yes I verify It. In fact I just generate it with STMCubeMx software. I generate the code on STM32CubeIDE for programming and debugging.

[Amira]

Hello and thanks for sharing these details.

I have some doubts about the clock configuration. You can refer to the provided examples (SPI) within STM32F3Cube package and follow the same clock config.

It can help you!

Regards,

Amira

[SLONK.1]

Sir,

I'm using STMCubeMx to configure the clock. STM32F3Cube is a package as I know, How can it help me ? How to follow what you are calling "same clock config".

Regards.

[TDK]

Monitoring DR has consequences. If your debugger reads data, it won't be read by your program.

How do you know the data on the line isn't 0? Can you show a scope plot of it?

[SLONK.1]

Yes I know but when I monitor the same code on STM32L476VETx, I got data and it works fine. So I don't think it is the reason. You can check on the image the value of DR (not 0) when using it. So I think it is not the problem. For the data on the line, I don't know. May be I should find a scope to plot it; I don'h have a scope.

[TDK]

Seems very likely to be a hardware/wiring problem rather than a software issue.

[SLONK.1]

I see, I check with multimeter and the mean voltage on SPI_SOMI of the STM32F302CBTx is less than the one on the STM32L476VETx where the code works well. I will try to find a scope to cofirm that.