TMS570LS1224: Issue of SPI

Hello QJ,

Thanks a lot for your correction. That's I misunderstood the pins of SPI. After changing the pins, the sample works well.

Then I enabled spiEnableNotification to generate requests of RXINTENA and TXINTENA. I think the two interrupts of transmitter empty and receiver full would be generated shown as below.

However it seems 0x0100 of RXINTENA enabled both RXINTENA and TXINTENA. 0x0200 and/or 0x300 will cause master's receiving in poll mode incorrectly as below.

Here are my functions of main and spiEndNotification. The line underlined is the way I enabled RXINTENA and/ or TXINTENA. Please correct my mistake ?

int main(void)
{
/* USER CODE BEGIN (3) */
    int i;
    spiDAT1_t dataconfig1_t;

    dataconfig1_t.CS_HOLD = FALSE;
    dataconfig1_t.WDEL    = TRUE;
    dataconfig1_t.DFSEL   = SPI_FMT_0;
    dataconfig1_t.CSNR    = 0xFE;

    gioInit();

    /* Enable CPU Interrupt through CPSR */
    _enable_IRQ();

    /* Initialize SPI Module Based on GUI configuration
     * SPI1 - Master ( SIMO, SOMI, CLK, CS0 )
     * SPI3 - Slave  ( SIMO, SOMI, CLK, CS0 )
     * */
    spiInit();

    spiEnableNotification(spiREG3, 0x0100U);//0x0100 R buffer is full; 0x0200 T buffer is empty

    while(1)
    {
        /* Initiate SPI3 Transmit and Receive through Interrupt Mode */
        spiSendAndGetData(spiREG3, &dataconfig1_t, 16, TX_Data_Slave, RX_Data_Slave);

        /* Initiate SPI1 Transmit and Receive through Polling Mode*/
        spiTransmitAndReceiveData(spiREG1, &dataconfig1_t, 16, TX_Data_Master, RX_Data_Master);

        for(i=0;i<DELAY_VALUE;i++);

        if(RX_Data_Master[15]==0x20)
        {
            RX_Data_Master[15]=0;
        }
        else
        {
            while(1);
        }
    }
/* USER CODE END */

    return 0;
}

/* USER CODE BEGIN (4) */
/** @fn void spiEndNotification(spiBASE_t *spi)
*   @brief Interrupt callback for End of TX or RX data length.
*   @param[in] spi   - Spi module base address
*
* This is a callback that is provided by the application and is called upon
* an interrupt at the End of TX or RX data length.
*/
void spiEndNotification(spiBASE_t *spi)
{
    if(SPI_COMPLETED==SpiRxStatus(spi))
    {
        if(RX_Data_Slave[15]==0X10)
        {
            RX_Data_Slave[15]=0;
            gioToggleBit(gioPORTB, 1);
            return;
        }
       else
       {
           gioSetBit(gioPORTB, 1, 0);
           return;
       }
    }else if(SPI_COMPLETED==SpiTxStatus(spi))
    {
        gioToggleBit(gioPORTB, 2);
        return;
    }
    else
    {
        gioSetBit(gioPORTB, 2, 0);
        return;
    }
}
/* USER CODE END */

Kind Regards

Datïan

Hello QJ,

Thank you very much. I'll try it.

Warm Regards

Datïan

Hello QJ,

I use SpiRxStatus(spiREG1) to check whether the receiving of master is finished. But the receiving buffer is not as I expecting of { 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F, 0x20 }.

Best Regards

Datïan

Hello QJ,

No problem. The high lighted part is TX_data_Slave, which was received incorrectly by Master.

Here is the snapshot of incorrect RX_Data_Master observed by a break point, in the case of spiEnableNotification(spiREG3, 0x200u).

And here is the correct RX_Data_Master observed by the other break point, in the case of spiEnableNotification(spiREG3, 0x100u).

Warm Regards

Datïan

Hi QJ,

Many thanks for your very helpful information of spiEnableNotification should not be called in the case of calling spiSendAndGetData. After I removed spiEnableNotification, it works very well as spiEnableNotification(spiREG3, 0x0100U).

Normally I will reference "Help Topics" in HalCogen, when an issue is emereged. The "Help Topics" is very good. May I ask you recommend more materials about this kind of topics for me ?

Kind regards

Datïan