• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 80 subscribers
  • Views 995 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CCS/MSP432P401R: MSP432 Multiple SPI Slave Devices in one SPI bus

Vincent Lai
Prodigy 170 points

Part Number: MSP432P401R

Tool/software: Code Composer Studio

Hi All

I want to use a MSP432P401R to control two SPI slave devices in one SPI bus with separate chip select pins (other SPI buses on my 432 are occupied).

Question/Problem

I have no trouble to control these two devices when deploying my debug code from a launch pad  to my final target (MSP432 on my PCB).

However, when I stop debug in CCS, reboot the target MSP432 and my PCB, only one SPI slave device work.

Trouble shooting steps

1. At the first, I wonder if the failed slave device takes long time after reset, I add some delay before calling the failed slave device, my problem still exists.

2. I swap the SPI calling order; the problem still exists.

3. I wonder if the code burning process does complete successfully; while all other devices controlled by the 432 on my PCB work.

Part of My code

const eUSCI_SPI_MasterConfig spiMasterConfigADF4351 =
{
        EUSCI_A_SPI_CLOCKSOURCE_SMCLK,             // SMCLK Clock Source
        3000000,                                   // SMCLK = DCO = 3MHZ
        500000,                                    // SPICLK = 500khz
        EUSCI_A_SPI_MSB_FIRST,                     // MSB First
        EUSCI_A_SPI_PHASE_DATA_CAPTURED_ONFIRST_CHANGED_ON_NEXT,
        EUSCI_A_SPI_CLOCKPOLARITY_INACTIVITY_LOW,  // CLK low when SPI not enabled for ADF4351
        EUSCI_A_SPI_3PIN                           // 3Wire SPI Mode
};

const eUSCI_SPI_MasterConfig spiMasterConfigAD8366 =
{
        EUSCI_A_SPI_CLOCKSOURCE_SMCLK,             // SMCLK Clock Source
        3000000,                                   // SMCLK = DCO = 3MHZ
        500000,                                    // SPICLK = 500khz
        EUSCI_A_SPI_MSB_FIRST,                     // MSB First
        EUSCI_A_SPI_PHASE_DATA_CHANGED_ONFIRST_CAPTURED_ON_NEXT,    // Phase
        //EUSCI_A_SPI_PHASE_DATA_CAPTURED_ONFIRST_CHANGED_ON_NEXT,
        EUSCI_A_SPI_CLOCKPOLARITY_INACTIVITY_LOW, // High polarity
        EUSCI_A_SPI_3PIN                           // 3Wire SPI Mode
};

// Typical 3 pin SPI initialization
GPIO_setAsPeripheralModuleFunctionInputPin(ADF4351_SPIPort, ADF4351_SPICLK | ADF4351_SPISIMO | ADF4351_SPISOMI, GPIO_PRIMARY_MODULE_FUNCTION);
SPI_initMaster(ADF4351_SPI, &spiMasterConfigADF4351);
SPI_enableModule(ADF4351_SPI);


GPIO_setAsOutputPin(ADF4351_CSPort, ADF4351_CSPin);
// set chip select high as default value
GPIO_setOutputHighOnPin(ADF4351_CSPort, ADF4351_CSPin);
GPIO_setAsOutputPin(AD8366_CSPort, AD8366_CSPin);


adf4351_initialization (ADF4351_SPI, ADF4351_CSPort, ADF4351_CSPin);
SPI_disableModule(ADF4351_SPI);

// re-config SPI since SPI master configurations of the two devices are different    
GPIO_setAsPeripheralModuleFunctionInputPin(AD8366_SPIPort, AD8366_SPICLK | AD8366_SPISIMO | AD8366_SPISOMI, GPIO_PRIMARY_MODULE_FUNCTION);
SPI_initMaster(AD8366_SPI, &spiMasterConfigAD8366);
SPI_enableModule(AD8366_SPI);

AD8366_Set_Gain (4500, 4500, AD8366_SPI, AD8366_CSPort, AD8366_CSPin);

PS.

Software platform CCS 8.3.0.0009

Hardware: MSP432P401R launch pad -> JATG ->  target MSP432P401R

  • 0
    TI__Guru 57566 points

    Can you see any difference in the actual SPI waveform with and without debugger?

    What is the actual SPI port you are using?  AD8366_SPI?

    Thank you,

    Chris

  • 0 in reply to Chris Sterzik
    Prodigy 170 points

    Hi Chris

    Yes.  I did check the SPI waveform with an oscilloscope both with and without a launchpad.

    The SPI waveform is correct.  Now I suspect there could be something wrong in hardware.

    While if you have any idea, please share your insight with me.

    BTW, The SPI port I use is UCA2 (P3.1/3.2/3.3)

    Thank you

    Vincent

  • 0 in reply to Vincent Lai
    Guru 95350 points

    > GPIO_setAsOutputPin(AD8366_CSPort, AD8366_CSPin);

    I don't see where you set the AD8366 CS pin initially low (inactive). The contents of the OUT register are undefined after Reset.

    -------------------

    When something succeeds in the debugger, but not without, the (first) Usual Suspect is timing. Are you making sure your transactions are complete (UCBUSY==0) before calling disableModule?

  • 0 in reply to Bruce McKenney
    Prodigy 170 points

    Hi Bruce

    >> I don't see where you set the AD8366 CS pin initially low (inactive). The contents of the OUT register are undefined after Reset.

    Thank you for pointing out.  I set default value for AD8366 CS pin but the problem still exists. 

    Regards

    Vincent

  • 0 in reply to Bruce McKenney
    Prodigy 170 points

    Hi Bruce

    >> I don't see where you set the AD8366 CS pin initially low (inactive). The contents of the OUT register are undefined after Reset.

    Thank you for pointing out.  I set default value for AD8366 CS pin but the problem still exists. 

    Regards

    Vincent

  • 0 in reply to Vincent Lai
    TI__Guru 57566 points

    Be cautious on how the chip select is activated and deactivated relative to the spi communication.  If you set it low after sending the last byte, it is possible that the CS is deselected before the actual last byte is transmitted.  The TX IFG indicates that the TX buffer is empty and not that the byte was transmitted.  Use the receive IFG to determine when the last byte is actually sent.

    Regards,

    Chris

  • 0 in reply to Chris Sterzik
    Prodigy 170 points

    Hi Chris

    Thank you for the explanation.  My problem is due to PCB design power supply circuit.  When rebooting, its power sequence cause some unexpected long delay that leads to some chips not ready before MCU able.

    I add some delay in my code as a workaround.

**Attention** This is a public forum