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.

MSP430I2040: Increasing the sample rate and update rate

Expert 4041 points
Part Number: MSP430I2040


Hi team,

I am working on an MSP430i2040 sub-metering EVM. The design has a sample rate of 8kHz and update rate of 4 AC cycles.

I'd like to ask:

1. What is the maximum sample rate that can be achieved?

2. Is it possible to make the update rate of a single cycle?

Please advise the trade-off if any. Thank you.

  • The sample rate of 8kHz sounds like OSR=128.

    The hardware can supply: OSR (ENOB)=32 (15), =64 (18), =128 (21), =256 (24).

    1) With OSR=32, you can get 1.024MHz/32=32000sps (with ENOB=15).
    2) Based on your numbers, this sounds like OSR=32 gives you 4/(128/32)=1 cycle

    I don't know what the software is capable of. I glanced at it some months ago, and the ADC ISR looked rather heavyweight, so I don't know if it could handle a 4x increase in the sample rate.

    There's also the question of the accuracy of the results -- somebody presumably thought 21-bit ADC precision was important.
  • Bruce,

    Thanks for your advice.

  • Hi Kien,

    Please mark Bruce's reply as correct if it has answered all of your questions, or provide more feedback if you require further support.

    Regards,
    Ryan
  • Hello Kien,

    Thank you for posting to E2E. Bruce's response is correct. The hardware may be able to reach high sampling rates, but the software may not have enough time between the SD24 interrupts at a certain sampling frequency - you would have to test the limit.

    For the sampling rate, we chose 8 kHz for this design to allow the bandwidth to cover harmonics up to the 80th harmonic (at 50 Hz input signal) or 66th harmonic (at 60 Hz input signal). Is there a specific reason or requirement for sampling faster than 8 kHz?

    For the update rate, 4 AC cycles were chosen for this design based on typical customer requirements. It's actually quite an aggressive update rate but can be changed. Keep in mind that this design is focused on measuring power and not energy. I expect that waiting for additional cycles could cause problems (like data register roll-over, additional calculation time) especially if a higher sampling rate was used whereas waiting for fewer cycles may work fine. Again, you would have to test this.

    Regards,

    James

    MSP Customer Applications
  • Hello James,
    Thanks for your answer.
    For basic power quality data and harmonics, 8kHz should be enough.
    The higher sampling rate may be required when customer wants to track the transient response of the current.
  • Hello Kien,

    Thanks for the additional insight. Please let me know if you have any questions during your testing.

    Regards,

    James

    MSP Customer Applications

**Attention** This is a public forum