• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 30 subscribers
  • Views 968 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

AFE4404: AFE SNR measurement - getting started

Rudi Hechfellner
Prodigy 30 points
Part Number: AFE4404

I am trying to measure my systems SNR - can someone get me started on the:

1) lab bench test setup

2) data analysis and SNR calculation

Thank you in advance.

  • 0
    TI__Genius 12575 points
    Hi,

    You can have a setup with the optical loopback.
    For this, keep a reflective surface in front of the sensor module and capture the data after configuring the AFE.
    Please note that using this method, you will capture AFE noise + Optical noise ( due to LED and PD).
    To calculate the noise, take the standard deviation of the captured data. The calculated noise can be referred to 2V to get the SNR in dBFS.

    Regards,
    Prabin
  • 0 in reply to Prabin Yadav
    Prodigy 30 points
    Prabin,

    Thank you for your response - this is very helpful. Now I understand the test setup a bit better.
    Please allow me to ask a couple more clarifying questions on the data analysis.
    1) I am using:
    SNR = 20 * log (mean ADC count/Std. Dev. ADC count)
    would the resulting values out of this formula be comparable with the data sheet values - is this an apples to apples comparison?
    2) I found that the data collection time (I am using 3 min.) increases the Std. Dev. (the longer I collect data the higher the Std. Dev. of course). What time window should I be using to find a comparable value with the data sheet?

    Thank you in advance.
  • 0 in reply to Rudi Hechfellner
    TI__Genius 12575 points
    Hi Rudi,

    For apple to apple comparison with datasheet, you have to calculate the SNR as
    SNR(dBFS) = 20*log10(2/noise), where noise is standard deviation in V.

    Do you see some kind of setting behavior in the raw data? May be can you share the plot of data. We had never seen increase in noise with time.

    Regards,
    Prabin
  • 0 in reply to Prabin Yadav
    Prodigy 30 points

    Prabin,

    Thank you very much for your detailed information - very informative and helpful!

    On the topic of increasing noise over time. The more data I collect the higher the noise gets - this is simple math and a statistical increase of noise relative to the test time. My questions is how long of a test time was used to calculate the data sheet values. I just need to come up with an apples to apples comparison between my setup and the data sheet.

    Thank you in advance.

    Rudi

  • 0 in reply to Rudi Hechfellner
    TI__Genius 12575 points

    Hi Rudi,

    I totally agree with your statement, however the increment will be negligible within the time window for the measurement.
    Usually we take data for ~ 40S.

    Regards,
    Prabin