MSP432P401R: Custom BSL hangs randomly after TX interrupt flag gets stuck.

Hello,

 Could you please change your to:

void Radio_sendByteUART(uint8_t data)
{
    /* Wait until the TX buffer is empty, then send */
    while(MAP_GPIO_getInputPinValue(GPIO_PORT_P2, GPIO_PIN4));//Waiting for Radio to be ready to receive data.
//    while(!(MAP_UART_getInterruptStatus(BSL432_RADIO_MODULE, EUSCI_A_UART_TRANSMIT_INTERRUPT_FLAG)));
//    EUSCI_A1->TXBUF = data;
    MAP_UART_transmitData(BSL432_RADIO_MODULE,data);
}

And let us know if you see the same behavior. Also, could you please share the code that you are using to configure the UART.

  Thanks,

   David

Hi David,

I tried your suggestion and it is not hanging as much. I still have the issue every once in a while. 

Now it is getting stuck in the UART_transmitData function:

void UART_transmitData(uint32_t moduleInstance, uint_fast8_t transmitData)
{
    /* If interrupts are not used, poll for flags */
    if (!BITBAND_PERI(EUSCI_A_CMSIS(moduleInstance)->rIE.r, UCTXIE_OFS))
        while (!BITBAND_PERI(EUSCI_A_CMSIS(moduleInstance)->rIFG.r, UCTXIFG_OFS))
            ;

    EUSCI_A_CMSIS(moduleInstance)->rTXBUF.r = transmitData;
}

The code I am using to configure the UART is as follows:

//Using a 24MHz the same settings as the original BSL code.

const eUSCI_UART_Config LBS_RADIO_UARTConfig_115200 =
{
 EUSCI_A_UART_CLOCKSOURCE_SMCLK,
 6,  // UCBRx
 8,   // firstModReg UCBRFx
 32,  // secondMod UCBRSx
 EUSCI_A_UART_NO_PARITY,                  // No Parity
 EUSCI_A_UART_LSB_FIRST,                  // LSB First
 EUSCI_A_UART_ONE_STOP_BIT,               // One stop bit
 EUSCI_A_UART_MODE,                       // UART mode
 EUSCI_A_UART_OVERSAMPLING_BAUDRATE_GENERATION  // Oversampling
};


void Radio_UART_Config(){

   
    UCA1CTL1 |= UCSWRST;
    BSL432_RADIO_MODULE = EUSCI_A1_BASE;

    BSL432_RADIO_PORT = GPIO_PORT_P2;
    BSL432_RADIO_RXD_PIN = GPIO_PIN2;
    BSL432_RADIO_TXD_PIN = GPIO_PIN3;
    BSL432_RADIO_RXD_MODE = GPIO_PRIMARY_MODULE_FUNCTION;
    BSL432_RADIO_TXD_MODE = GPIO_PRIMARY_MODULE_FUNCTION;

    /* Selecting P2.2 and P2.3 in UART mode  */
    MAP_GPIO_setAsPeripheralModuleFunctionInputPin(BSL432_RADIO_PORT,
                                                   BSL432_RADIO_RXD_PIN, BSL432_RADIO_RXD_MODE);
    MAP_GPIO_setAsPeripheralModuleFunctionOutputPin(BSL432_RADIO_PORT,
                                                    BSL432_RADIO_TXD_PIN, BSL432_RADIO_TXD_MODE);

    MAP_UART_selectDeglitchTime(BSL432_RADIO_MODULE,EUSCI_A_UART_DEGLITCH_TIME_200ns);


    /* Configuring UART Module */
    MAP_UART_initModule(BSL432_RADIO_MODULE, &LBS_RADIO_UARTConfig_115200);

    /* Enable UART module */
    MAP_UART_enableModule(BSL432_RADIO_MODULE);

    UCA1CTL1  &= ~UCSWRST;//

    MAP_UART_registerInterrupt(BSL432_RADIO_MODULE,&Radio_Received_IRQ);
    /* Enabling interrupts */
    MAP_UART_enableInterrupt(BSL432_RADIO_MODULE, EUSCI_A_UART_RECEIVE_INTERRUPT);

    MAP_Interrupt_enableInterrupt(INT_EUSCIA1);


}

Thank you for the assistance. Please let me know if I can provide more info. 

Hello,

 I created this example code using your configuration but unfortunately I could not reproduce that behavior.

 Do you have a logic analyzer / scope trace that you can share when the device stops transmitting.

  Thanks,

  David

Hello,

I'm wondering if you had chance to test the example code I attached on my previous post?? Were you able to reproduce the same behavior??

Thanks,

David

Hi David,

I followed your suggestions. I first tried the code change you provide and that significantly reduced the recurrence of the issue. I have been able to run the update software over 500 times without seeing the issue. However, I can still get the issue show up randomly. I also checked the signal integrity for the TX and RX signals, and they seem to be clean (I haven't been able to catch a waveform of the fault since it is so random). I am right now looking at the TX and RX data coming between the radio and the MSP432 to see if I can catch anything.

For the moment, I have also added a Watch Dog Timer when entering the sendByte function. if the WDT is triggered, an ISR function calls a mass erase so I don't lose communication with box. 

Do you have any other suggestions?