• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 495 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.

Original question:

ADS1263: 3-Wire RTD Low-Side Reference

ADS1263: ADS1263: 3-Wire RTD Low-Side Reference Noise Source

Steven Park
Prodigy 60 points
Part Number: ADS1263


Hello,

I asked similar question couple months ago regarding unknown offset in the measurements. The issue has been resolved by running custom calibration, since the offset was uniform at multiple temperature points. Now I'm having an issue with some noise in the measurements.

The RTD setup is as follows,

- IDAC1 and IDAC2 are setup with 500uA 

- AIN7 connects positive lead of RTD, AIN6 connect negative lead of RTD

- AIN5 and AIN 4 are selected as external reference

- Reference resistor is 2000K ohm

- ADC2 data rate is selected as 10SPS, PGA = 1

- PT100 RTD is used to measure the temperature, the conversion factor is 0.385Ω/°C. Using the pk to pk noise provided by the data sheet (32.6uV), I was expecting around 0.1693°C noise; 0.385Ω/°C * 500uA = 192.5uV/°C and 32.6uV/192.5uV/°C = around 0.1693°C

The noise I'm seeing is around 0.8°C = around 154uV.

There are few components that I am suspecting as a source of noise but before dive into troubleshooting, I'd like to clarify some questions I have regarding the data sheet.

- The IDAC error spec on page 8 defines the accuracy of the IDAC not the noise

- ex) absolute error of +/-%0.7 means the current output will be within the error

Thanks,

Steven

  • 0
    TI__Guru 53516 points

    Hi Steven Park,

    Theoretically, the noise of the IDAC should cancel out, as it should be seen in the input and reference equally because you are using a ratiometric reference configuration

    How are you determining that this is noise, and not offset or gain error (or some other error)? Are you actually calculating the standard deviation of a group of samples e.g. 128? Or are you just comparing the measured value to the expected value from the RTD, which is actually measuring accuracy?

    The RTD itself will have noise and error - have you calibrated your system without the RTD to understand how much noise/error the sensor actually contributes? Usually you would use a very high precision fixed resistor e.g. 100ohm, to understand the accuracy of your system, because you could just be measuring the error from the sensor.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 60 points

    Hi Bryan,

    Honesty, I haven't collected large number of samples as I did in other post regarding noise. However, I've collected around 1000 samples at 10SPS. The difference of Min and Max of the sample group was around 0.8°C.

    The RTD was in a high precision dry-well from Fluke with precision +/-0.01°C. The error spec of RTD is +/-0.5°C but doesn't specify the noise spec. I'd assume the error would be constant. 

    I will try your suggestion of testing with a precise 100Ω resistor.

    Thanks,

    Steven

  • 0 in reply to Steven Park
    TI__Guru 53516 points

    Hi Steven Park,

    Typically, just looking at the min and max values is not very helpful. Especially if you are taking data over a long period of time.

    In your case, you took 1000 samples at 10SPS, or 100 seconds worth of data. There could be temperature shifts in this data, or 1/f noise. If your data is varying because of these phenomenon, the min/max data will appear to be worse even though this is not necessarily the result of "noise".

    I would encourage you to look at the complete data set in the time domain to see if the noise is purely Gaussian (broadband / white) or if there are other components to the noise spectrum

    Let me know what happens with the precision resistor (0.01%, 5 ppm/C or similar)

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 60 points

    Hello Bryan,

    I've ordered the precision resistor with the spec you mentioned. I was still seeing a wider pk to pk noise range than I expected. I think this is related to other post that I opened regarding noise I'm seeing on regular ADC measurements. I will mention this in the other post

    Thanks,

    Steven

  • 0 in reply to Steven Park
    TI__Guru 53516 points

    Hi Steven Park,

    Thanks for your feedback, I will address this in the other post then since there is more information there

    -Bryan