• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 46 subscribers
  • Views 524 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.

ADS7052: Confusion about ADS7052 datasheet

user5802412
Intellectual 300 points
Part Number: ADS7052
Other Parts Discussed in Thread: ADS7056

Hello TI member:

I have some questions about the datasheet of ADS7052, could you help me to clarify them ? Thank you !

Q1

The typical INL of ADS7052 is announced about +/- 2 LSB in the spec table, but why Fig.16 seems it can achieve such a great performance about +/- 0.7 LSB ?

Furthermore, Fig.19 shows that INL with temperature through -40°~125° is about +/- 2 LSB, but in Fig20.  it shows only  +/- 0.7 LSB at 25°C?

Q2

There are two well – known methods to estimate INL ( best-fit & end-point ), which method is recommended or used for data sheet ?

Thank you!

B.R

Chen

  • 0
    TI__Mastermind 24940 points

    Hello,

    If you look at figure16, the y axis has an interval of 1.2lsb, thus the minimum point on the graph would actually be closer to -1.2 rather than -0.7.

    Granted it is still less than the specification given in the datasheet. The datasheet shows the performance used when testing devices, meaning that as along as a device meets the maximas, it is shipped and sold. The performance in figure 16 is empirical data from devices tested. Also, need to point out that figure 31 show how many devices showed what INL performance.

    The specification table needs to bound all these variables.

     The End point fit is used to find the INL performance.

    Regards

    Cynthia

  • 0 in reply to Cynthia
    Intellectual 300 points

    Hello Cynthia :

    Thank you for your reply, for Q2, may I share you with my linearity measurement raw data and ask your help to analyze them about DNL, INL, offset, gain error ?

    I think my DNL plot seems almost to meet the target spec , however INL is still a big problem…. Thank you!

    B.R

    Chen

  • 0 in reply to user5802412
    TI__Mastermind 24940 points

    Chen,

    Yes, feel free to share any information you have, and I can take a look at it.

    If your DNL is meeting criteria, then your INL will likely also meet it.

    Regards

    Cynthia

  • 0 in reply to Cynthia
    Intellectual 300 points

    Hello Cynthia :

      It’s very appreciated that you can use your time to help my request, since this request was not the original request about ‘confusion of datasheet’ Should I choose a best answer and close this thread then post a new request with my raw data ?

      Sincerely, may God allow us to meet again. Thank you !

      By the way , my raw data format has purely 399999 counts, and the digital sine wave with full code hit, If this format will limit or affect you to use histogram analysis , please let me know!

    B.R

    Chen

  • 0 in reply to user5802412
    TI__Mastermind 24940 points

    Chen,

    Lets just keep goin gin this thread. 

    There are video training on this subject that I think you will find helpful. as part of TI Precision Labs there is a video about DC specifications for ADCs, which most directly addresses your concerns here.  the slide deck is also available for download

    Regards

    Cynthia

  • 0 in reply to Cynthia
    Intellectual 300 points

    Hello Cynthia :

      Thank you for sharing me about TI video training! This series are masterpiece for analog IP! I had watch them a few years ago, and I have notice that one of lecturer about ' Driving SAR ADC without Amplifier ' also named Cynthia ! Hmm... please forgive my small talk if it made you feel rude.. Sorry.

      The attached is my raw data. If this format will limit or affect you to use histogram analysis , please let me know!ADC_1649_3501_33621A_0201.txt

    B.R

    Chen

  • 0 in reply to user5802412
    TI__Mastermind 24940 points

    Chen,

    I am glad you find the videos helpful, the series began a couple years ago but new videos have been added since on new topics we find are helpful. I have helped with a couple videos, which is why my name is included on some of the latest videos.

    From the data you provided, it looks as if you are collecting the incorrect data to measure INL.

    You should use a DC input to measure DNL and INL, not a sine wave. The DC input will need to sweep from 0 to the Full scale range, and you will need to measure the ADC output and compare it to the expected output at the DC level. the expected output is the ADC transfer function. 

    Below is an excerpts of the End Point INL:

    End-point INL: The End-Point method provides the worst case INL. This measurement passes the straight line through the origin and maximum output code of the ADC.  As this method provides the worst case INL, it is more useful to use this number as compared to the one measured using best fit for DC applications. This INL number would be typically useful for error budget calculation. This parameter must be considered for applications involving precision measurements and control.

  • 0 in reply to Cynthia
    Intellectual 300 points

    Hello Cynthia :

    I agree with that sweep from 0 to the full range is a way to measure DNL and INL, but if we reference the document 'SBAU285', a user guide for ADS7056 EVM. We can find that a 2-kHz sinusoidal input signal is recommended to measure DNL and INL on page 20 and that is the reason why my raw data was a sine wave. Could you help me to check about them? Thank you!

    Sorry! I can't find how to edit a hyper-link with 'SBAU285' on e2e, so I list it below directly.

    https://www.ti.com/lit/ug/sbau285/sbau285.pdf?ts=1612952958570&ref_url=https%253A%252F%252Fwww.google.com%252F

    B.R

    Chen

  • 0 in reply to user5802412
    TI__Mastermind 24940 points

    Chen,

    The ADS7056EVM is paired with a software GUI. Within this GUI, there is an elaborate algorithm to measure the INL and DNL with a 2khz wave. This formula is complicated and is used to alleviate the work from the user, and put it on the software. This is only used when using the EVM and GUI, as the GUI is doing all the work. Do you have the ADS7056EVM-PDK?

    If the DNL and INL are being measured manually, then the 2khz input signal is not recommended. In this case, I would suggest DC input signals across the input range. 

    Regards

    Cynthia

  • 0 in reply to Cynthia
    Intellectual 300 points

    Hello Cynthia:

    Thank you for your reply. Sorry, I don't have ADS7056EVM-PDK...  So the GUI have a powerful algorithm to measure INL, could you kindly share your time to discus this algorithm with me? I know this need some probability background.. but I hope that I can study it. Thank you!

    B.R

    Chen

  • 0 in reply to user5802412
    TI__Mastermind 24940 points

    Chen,

    This algorithm used in that GUI is not a prominent one, and is not recommended. It was a result of a necessity, and we do not commonly use it.

    I suggest to use the more widely known methods instead.

    Regards

    Cynthia