• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 364 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.

ADS7042: ADS7042 SNR/THD defination and calculation

Eroshoutai Gigi
Prodigy 90 points
Part Number: ADS7042

Dear Sir,
I’m work to verify the ADS7042 performance and check the SNR and THD result by raw data (sample 4096).
I have 2 questions which need to be support.

  1. Can anyone please help to explain how TI definition the SNR and THD, and how to calculate it?
    I cannot verify result from Python to TI GUI.

  2. In GUI calculation, SNR is very closed in 4096, 8192 and 65536 different samples.
    Considering the process gain, the SNR should be better when samples be larger, is that correct?
    or TI had other digital filter process in GUI calculation?

The detail of question 1 is as below:
Input 10KHz Vpp=1.62V sinewave, I get the result in GUI tool.

And export raw data and try to verify result by python.

But I cannot get the same SNR/THD value as GUI tool.
I try to ask the code to community (Stack overflow).
https://stackoverflow.com/questions/76099845/snr-calculation-by-python-verified-code-by-ti-ads7042-evm/76100036#76100036
"Calculating SNR is usually a bit arbitrary thing to do. It depends on the signals you are measuring.
If you want to have 1-to-1 correspondence you could ask Texas Instruments how they calculate it. Maybe this is not a secret."
They suggest me to seeking support from TI for getting correct definition and calculation of SNR and THD.

Thanks

  • 0
    TI__Guru**** 173076 points

    Hi Eroshoutai,

    I don't have details on exactly how the calculations are done in the ADS7042 GUI, however, there are two main reasons why they might be getting a different SNR value:

    1. Often times the first few FFT bins are ignored (exclude them from the total noise) since they can influenced by sample size/length and any offset drift present in the data.
    2. Usually all harmonics are removed from the spectrum in order to compute the total noise. The # of bins around each harmonic peak can be adjusted to account for spectral leakage of the window function. For a 7-term blackman-harris window we usually count and sum +/- 7 FFT bits around each tone as representing the total power of each tone (including the fundamental) since this is the expected spread of that window function.