• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 47 subscribers
  • Views 567 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.

multiple ADC counts observed for a DC signal

Meera Dasog
Prodigy 40 points
Other Parts Discussed in Thread: ADS130E08, LM7705

Hi,

I am using the ADS130E08 eval kit to sample DC signals from a frontend circuit. The DC output of the frontend circuit varies from 0 to 4V, so I have configured the register to use 4V reference. Everytime I acquire/sample the DC signal I see a set of different ADC counts at the output. The ADC count is sent to a processor. So I would like to know if there is a way to determine the correct ADC count for a voltage value and why am I seeing so many jumps in the ADC count ?

Thanks,

Meera

  • 0
    TI__Mastermind 20845 points

    Dear Meera,

    welcome to our forum and thank you for your interest in our ADS130E08.

    I am afraid I have to disappoint you right from the beginning. With ADS130E08 it is unfortunately not possible to measure a signal where one input is held at GND when using a unipolar power supply (I assume you are using an AVSS=0V and AVDD=5V?). Your input signal will have to meet the common-mode voltage range of the PGA as described in the datasheet on p. 16.

    In principle you have two options to make your measurement work:
    1. Use a negative AVSS of -200mV or less or
    2. level shift your input signal to within the common-mode voltage range of the PGA

    To calculate the expected ADC count you can start by calculating the LSB size which is:
    LSB = (2 x Vref) / (Gain x 2^16) = (2 x 4V) / (1 x 2^16) = 122uV
    You then simply divide your applied input signal by the LSB size. This should then give you the expected ADC code range.

    Please note that the ADS130E08 outputs data in twos-complement format.
    Also note that due to INL, offset, gain and Vref errors your code will not exactly match the code that you calculated.

    What data rate are you running the ADS130E08 at? And how many codes do you see changing?
    At the lower data rates you should not see any code changes due to noise. At the higher data rates you will see some codes flipping because the ADC does not offer a noise free 16bit performance at the highest data rates.

    Regards,

  • 0 in reply to Joachim Wuerker
    TI__Expert 6260 points

    Hello,

    Regarding the suggestion: "Use a negative AVSS of -200mV or less or"

    please have a look at the LM7705 which can easily accomplish this creating a negative bias rail of -232mV.

    LM7705: Low Noise Negative Bias Generator
    Product Folder: http://www.ti.com/product/LM7705
    Datasheet (PDF): http://www.ti.com/litv/pdf/snvs420b

    Thank you, Keith

  • 0 in reply to Keith Keller
    TI__Mastermind 23235 points
    Hello Meera -
    Please ensure that if you are using AVDD=5V that you keep AVSS at 0V or really close. Confusion can arise to this part having a unipolar and bipolar operating mode. It is intended to operate as unipolar (+5V and 0V) or bipolar operation (+/-2.5V). The analog inputs are not intended to go below ground unless you are using bipolar mode.

    When you are referring to 'jumps' are you talking about a constant value with noise or discrete voltage 'steps'? For DC measurements, there will always be some noise on output value, it will not be a constant digital value. This is expected behavior. However, if you are seeing voltage steps, this may be a different problem that we can help you with.
  • 0 in reply to Joachim Wuerker
    Prodigy 40 points

    Hi Joachim,

    Thank you for your response.  For the supply, I have AVDD=5V and AVSS =0V. I am using the sampling clock on the board i.e. 2.048MHz for the data rate. So for a DC input 1.62V, the expected ADC count would be ~13271 and Hex value 33D7. But I see ADC counts varying from 0 to the expected value - please refer attached file, channel 1 data.

    Device_1_Codes.xls

    As suggested, I will try both options and see if it works.

    Thanks,

    Meera

  • 0 in reply to Greg Hupp
    Prodigy 40 points

    Hi Greg,

    Thanks for the response.  By 'jumps' I meant different ADC counts (discrete voltage steps) from the expected value. I am giving a DC signal 1.62V from the frontend circuit (the input range is 0 to 4V), Vref is set to 4V, AVDD = 5V and AVSS = 0V. The expected ADC count would be 33D7h, but I see other random values. Channel data file attached  for reference.

    5270.Device_1_Codes.xls

    Thanks,

    Meera

  • 0 in reply to Meera Dasog
    TI__Mastermind 23235 points
    Hello Meera -
    Can you send your register configuration that you are using? When you are connecting the DC value, how are you connecting the analog inputs? For example, DC value to the + input, what to the - input?
  • 0 in reply to Greg Hupp
    Prodigy 40 points

    Hi Greg,

    I am driving the ADS130E08 in single ended configuration. The negative input is connected to ground. But as Joachim suggested, I am going to try connecting to common mode voltage value. Attached file has the register configuration.

    Device_1_Analysis.xls

    Thanks,

    Meera