• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 1546 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.

ADS1115 SPS vs Resolution

rfengr
Intellectual 720 points
Other Parts Discussed in Thread: ADS1115

I have read the ADS1115 datasheet from cover to cover in order to find any kind of information leading to the relationship between the number of samples taken per second and the effective resolution. 

For example: a competitor device MCP3425 has the following listing:

15 SPS (16 bits)
60 SPS (14 bits)
240 SPS (12 bits)

However, I could not find any similar information for the ADS1115. Not even Google could help me. 

Would someone be kind enough to explain to me how I can calculate this effective resolution vs the number of samples taken per second, or at least provide me with the information I seek.


Thank you!

  • 0
    TI__Guru* 92715 points
    rfengr,


    Effective resolution is a measurement of the noise in the ADC. As an example, if the ADC has a range of +/-2.048V (or 4.096V total), and the ADC measurement has a noise of 100uV peak-to-peak coming from the converted code readout you can calculate the effective resolution as:

    log(4.096V/100uV)/log(2) = 15.3 bits of effective resolution

    Now going to the ADS1115 datasheet, you can find some of this information in Figures 14 and 15. Using Figure 15 as an example with a full scale of +/-2.048V running at 860SPS, let's say that the noise is 30uV rms. Using a conversion from rms to peak-to-peak factor of about 6. You get a noise of 180uV p-p.

    Calculating the noise becomes:

    log(4.096V/180uV)/log(2) = 14.5 bits of effective resolution.

    The figure doesn't cover every combination of range and data rate, but you'll be below 16 bits of resolution for every range with a data rate below 32SPS. If you're interested in a particular combination that isn't given, let me know.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Intellectual 720 points

    Thank you Joseph for the explanation, your help is very much appreciated.

    For my application, I am operating the ADS1115 at two different PGA settings:

    1) PGA = 8 (FS=+0.512V)

    2) PGA = 16 (FS=+0.256V)

    My input supply is approximately 3.3V. I am sampling a relatively slow moving signal similar in nature to a thermocouple sensor. Most of the time there is no activity from the sensor (flat signal response), and the output amplitude is also small (less than 100mV). When I have activity at the sensor, the signal can rise up to 200mV. I would like to achieve a 16bit resolution, and I can most likely achieve that with the slowest SPS available. However, I have to keep in mind that I am operating from a battery, so faster SPS means that I can go to sleep sooner and save energy. 

    Would you be able to provide me with the missing graphs similar to Fig.14 and Fig.15 in the datasheet for (FS=+0.512V) and (FS=+0.256V) with a range of SPS = 8,16, 32, and 64.

    Thank you so much !  

    (P.S. in your previous reply, you said: "but you'll be below 16 bits of resolution for every range with a data rate BELOW 32SPS". I think you meant to say: "but you'll be below 16 bits of resolution for every range with a data rate ABOVE 32SPS". Perhaps the use of slower/faster SPS would be less confusing.)

     

  • 0 in reply to rfengr
    TI__Guru* 92715 points
    Joseph Wu



    You're right about the comment, I went the opposite way on resolution.

    I don't have any specific graphs to show you similar to the one in the datasheet. My information comes out of a table from some basic characterization data.

    However, for all of the settings that you listed, the peak-to-peak noise was about 8uV so the resolution will likely get you 16 bits. The one exception is the +/-0.256 range at PGA=64. The noise is about 12uV peak-to-peak where the noise is about 15.4 bits.

    I don't know the configuration with which this data was taken, so there may be some variance. However, I'd guess that this will be close to what you see.



    Joseph Wu