stm32f429t6, the uart8 dma rx function works on 3 out of 5 boards and does not work on 2

[joseph05]

In stm32f429t6, the uart8 dma rx function works on 3 out of 5 boards and does not work on 2. The same firmware was applied to 5 boards, and if only the uart8 interrupt was used without using dma, the 5 boards mode operated normally. What is the cause of the problem of DMA RX not working on two boards and how to improve it?

 

const uint16_t UART8_RX_BUF_SIZE = 256;
uint8_t uart8RxBuf[UART8_RX_BUF_SIZE];

extern DMA_HandleTypeDef hdma_uart8_rx;

/* USER CODE END PV */

 /* Initialize all configured peripherals */
 MX_GPIO_Init();
 MX_DMA_Init();
 MX_USART3_UART_Init();
 MX_USART2_UART_Init();
 MX_UART8_Init();
 MX_LWIP_Init();
 MX_ADC3_Init();
 MX_USART1_UART_Init();
 MX_TIM1_Init();
 MX_TIM2_Init();
 MX_TIM3_Init();
 MX_TIM4_Init();
 MX_TIM5_Init();
 /* USER CODE BEGIN 2 */
 ups.initializeRingBuffer(); 

 HAL_UARTEx_ReceiveToIdle_DMA(&huart8, uart8RxBuf, UART8_RX_BUF_SIZE);

void HAL_UARTEx_RxEventCallback(UART_HandleTypeDef *huart, uint16_t Size)
{ 
 if(huart->Instance == UART8)
 {
 ups.handleUartData(uart8RxBuf, Size);
 HAL_UARTEx_ReceiveToIdle_DMA(&huart8, uart8RxBuf, UART8_RX_BUF_SIZE);
 }
}
/* USER CODE END 4 */

 

 

[Sarra.S]

Hello @joseph05 

This seems to be a hardware issue? 

[TDK]

What does "not working" mean here exactly? You're the only one with the board and code, step through and understand what isn't happening and why. Use those results to guide your investigation.

Look at the hardware for any defects. In particular at the UART_RX pin for solder bridges or cold joints.

[Tesla DeLorean]

That would be a very odd state of affairs. Assume the problem is not with the DMA, but the reception of data.

Not working in the sense that the DMA or UART report some error or failure status you're not addressing? Check what info UART->ISR / SR reports in these conditions. Clear sticky errors preventing reception. Check manual.

Your code looks to ignore errors

[joseph05]

The problem has not been resolved yet and the cause is being investigated. I will update any additional information found.