• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 3 answers
  • Subscribers 46 subscribers
  • Views 233 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.

ADS8028: How to calculate the ADC error and compare the ADC accuracy with AMC7836?

Zhizhao Niu
Intellectual 2315 points
Part Number: ADS8028
Other Parts Discussed in Thread: AMC7836,

Hi Team, 

I have a question that how to calculate the ADC error and compare the ADC accuracy? I found several error index in datasheet, such as INL, DNL, gain error, offset error. How to calculate the ADC error and accuracy based on those index? I want to compare the ADS8028 ADC accuracy performance with AMC7836. Thanks for your help!

B.R.

Lucas

  • 0
    TI__Mastermind 22950 points

    Hi Lucas!

    It sounds like you are trying to calculate the Total Unadjusted Error (TUE). I have linked a an application note that talks about how to calculate this error. The one thing to note is that the units should be the same values. So for example, if INL is in LSB and gain error is %FS, you need to convert one of the values to the other. 

    https://www.ti.com/lit/an/slaa587/slaa587.pdf?

    Additionally, we have the analog engineer's calculator that has a TUE calculator build it and will convert the units for you. A link to the calculator is below. The PGA gain can be set to 1V/V and the error analysis does not need to include the amplifiers unless you have an external amplifier driving the ADC that you would like to include. 

    https://www.ti.com/tool/ANALOG-ENGINEER-CALC

    Regards,
    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2315 points

    Hi Aaron, 

    Thanks for sharing! After I look through the app note. I have below questions? 

    1. What the TUS is the RMS value of all the error instead of directly add all the error together? 

    2. What we still need to bring DNL into error calculation? Because I think the INL is the integral of DNL, so it can show the max error of each code. 

    3. Sometimes I can see the datasheet shown with offset error and offset error match. What is the difference between them? Like shown in AMC7836 datasheet as below. 

    Thanks.

    B.R.

    Lucas

  • 0 in reply to Zhizhao Niu
    TI__Intellectual 2315 points

    Hi Aaron, 

    Could you please help check the above Q2? Thanks!

    B.R.

    Lucas

  • 0 in reply to Zhizhao Niu
    TI__Mastermind 22950 points

    Hi Lucas, 

    Apologies for the delay. 

    1. Root sum squared allows us to approximate the variation in the errors themselves due to the fact that the errors can drift both positive and negative. You can  directly add the absolute values of the errors but you will get a very large total error approximation that wont be representative of the error drift. 

    2. Yes, you can add DNL to the equation. However, if you do not plan on calibrating the offset error (OE) and gain error (GE), then INL and DNL will dominate. If you do not calibrate OE and GE, you will find that these two error sources tend to be the dominant source of error. 

    3. Offset Error Matching should be the channel-to-channel matching of the offset error. 

    Regards,
    Aaron Estrada

  • 0 in reply to Aaron Estrada
    TI__Intellectual 2315 points

    Hi Aaron, 

    Thanks! Sorry I type a wrong word in Q2. I just want to know can I ignore the DNL error? Because I think the INL have already represent the DNL in each step. 

    B.R.
    Lucas

  • 0 in reply to Zhizhao Niu
    TI__Mastermind 22950 points

    Hi Zhizhao,

    Sorry I believe I was misunderstanding your DNL question...You do not need to include DNL as DNL essentially causes INL. You just need INL in your TUE calculations. 

    Regards,

    Aaron