• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 46 subscribers
  • Views 307 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.

ADS131E06: Attenuation in harmonics of a 60 Hz sine wave

Thomas Jacobson
Intellectual 310 points
Part Number: ADS131E06
Other Parts Discussed in Thread: TM4C129DNCPDT

Hello TI Community!

I have an ADS131E06 connected to a TM4C129DNCPDT. We're sampling a little under 8 kSPS. After collecting 1536 samples, we run an FFT over the samples to get harmonics up to the 50th order. Our nominal voltage is 120 Vrms at 60 Hz (so the 50th harmonic would be at 3000 Hz).

In doing some tests, I've discovered that the measured harmonic magnitude decreases linearly as the harmonic order increases, even though the actual magnitude remains the same. What happens, then, is the error increases linearly from one harmonic to the next. The 10th harmonic has an error of 5%, the 11th harmonic has an error of 6.2%, and the 12th harmonic has an error of 7.2%. By the time we reach the 50th harmonic, we are expecting a magnitude of 6 Vrms, but measure a magnitude of 1.8 Vrms, an error of about 70%.

I'm trying to determine the root cause of this attenuation, and the A/D converter is one part of the investigation. Is there anything inside of the ADS131E series that could cause this phenomenon? My understanding of analog circuits is limited, but I haven't seen anything in the datasheet that would suggest a low-pass filter inside the A/D converter is causing the higher frequencies to be cut off (or I just didn't understand it).

Thank you for your time.

- Tom

  • 0
    TI__Guru* 93765 points
    Tom,


    I'm not sure you've considered this, but the ADS131E06 is a delta-sigma type converter, which oversamples the input to make each data point that comes out. With this oversampling, you use a different digital filter to get the data at a particular data rate. The consequence of these digital filters is that there is a particular frequency response for each one.

    Go to the datasheet and look up the section on Digital Decimation Filter on Page 24. There's a brief description of the idea, but on the following page there's a set of frequency response plots for the different data rates.

    If you are running at about 8kSPS the DR[2:0] is set to 011. If the input frequency is at 3kHz, then your normalized frequency is 0.375. From Figure 25, I'll admit it's hard to tell which one is 011, but it's the 4th largest. With the sinc3 filter, I would have guessed that a 3kHz signal would have been very small, so I'm not sure what else is going on, unless you're at a different data rate.

    From your question, I wasn't completely sure that was what you were seeing. If I've misunderstood your question, let me know and we can discuss it further.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Intellectual 310 points
    Hi Joseph,

    Thanks for your response. My knowledge of signal processing and filters is quite limited, I still have a lot to learn. Your reply, however, has been very helpful to me