• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 34 subscribers
  • Views 948 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.

LMT70: Temperature readings is off by 0.5 C to 1.0 C

Arif Quronfleh72
Prodigy 80 points
Part Number: LMT70

We're not sure why our device is reading the temperature value off by  0.5 C to 1.0 C.  We are comparing it to other accurate and calibrated devices placed in the same conditions. 

We get the same error if we connect it to the sensor on our board OR the sensor part of the LMT70EVM (We hard wired it to the sensor). 

Our temperature range is from -30 C to +70 C

We are interfacing it to the Nordic nRF52832 

This is an urgent request as we just finalized the firmware and are trying to resolve this issue. We have purchased several thousands of these sensors and are ready for production. Not sure why we're getting this issue now. 

Is there anything we can do in the calculations? 

If it's easier and faster I can contact customer support. 

Thanks very much.

Arif 

  • 0
    TI__Expert 5695 points

    Our support team will contact you with updates in the next 14-16 hrs due to the expertise team location in the east cost of USA.

    Please share with us the other "reference" temperature systems descriptions that you are using to baseline the LMT70 behavior. Block diagrams and description of ADC/ Gains  you are using will be very helpful. Also share your calculation method and temp range you are verifying.

    P.S. Sorry for the various edits.

  • 0 in reply to Jose Arteaga4
    TI__Expert 5695 points
    Please share with us the other "reference" temperature systems descriptions that you are using to baseline the LMT70 behavior. Block diagrams and description of ADC/ Gains you are using will be very helpful. Also share your calculation method and temp range you are verifying.
  • 0
    TI__Mastermind 31600 points

    Hi Arif,

    Thank you for bringing my attention to this thread. I am sorry that you have to go through the issue. I checked the specs of the nRF52832, and the 12-bit ADCs should be enough for the LMT70. Could you please provide me the following information?

    1. Did you use the LUT table in section 8.6 for your temperature calculation?

    2. What is the reference temperature equipment you are using? What is the accuracy of the probe?

    3. How do you test the LMT70 with the reference temperature?

    4. Did you put the LMT70 and reference temperature in the enclosure? Or outside the lab environment.

    5. Did you use the oven or oilbath to test the LMT70?

    Aaron

  • 0 in reply to Aaron Heng
    Prodigy 80 points

    Hi Aron,

    Thanks for your quick response. I was able to confirm what we use this morning. Here are the answers:

    1. We are NOT using the LUT table. We're using a third-order calculation formula which gave us 0.1-0.2C lower values from the LUT calculations. 

    2. We're using standard equipment with similar accuracy. We're measuring ambient temperature. 

    3. Placing it in ambient. 

    4. We're doing both with similar results. 

    5. No. 

    The interesting part is that the temperature is consistently above other values. 

    I also noticed that on the LMT70EVM the sensor has a coating of some sort so not sure if that contributes to accuracy or reading delay etc.

    Thanks

    Arif  

  • 0 in reply to Arif Quronfleh72
    TI__Expert 6475 points

    "Standard equipment with similar accuracy"?  Please clarify the accuracy.  Is your reference accuracy +/-0.27C from -40C to 90C?

    Just based on the LMT70 and your reference accuracy spec (not including error from the μC's ADC) your worst case error w/ respect to your reference can be as high as +/-0.54C.  Your ADC gain error temperature coefficient is 0.02%/C. From 25C to 70C, you can expect to get 0.9% gain error due to temp.  The typical gain error w/o temp drift is specified to 3-4% in the datasheet.  You will have to take all of these into consideration.

    -Kelvin