• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 193 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.

TMP61-Q1: Clarification about Accuracy Performance of the Thermistor TMP61-Q1 in TO-92 package

Neet
Intellectual 542 points
Part Number: TMP61-Q1
Other Parts Discussed in Thread: TMP61

Tool/software:

Hello E2E Community,

We are considering the use of TO-92 packaged version of TMP61-Q1 for one of our remote temperature measuring use case with 3m long twisted pair cables.

I am seeking clarification about the following Figure 8-13 from datasheet about the errors.

Assume following as our configuration:

  1. Do NOT perform 1-point offset calibration
  2. Use Rbias 10kOhm +-0.5% (50ppm) [cannot use current bias configuration]
  3. Install RC Low Pass Filter with 15kHz cutoff
  4. We perform 1024 oversampling on our ADC for improving accuracy

How do we calculate the expected error typical & min/max from this configuration?

Is it a safe assumption the errors (min/max) are instantaneous errors & can be averaged out by oversampling?

Also, will performing 1-point calibration significantly improve the measurement accuracy?

Our expected accuracy is +/-2'C maximum.

Kindly guide with your best insights to speed up our design decisions.

  • 0
    TI__Expert 5825 points

    Hi Neet,

    When it comes to sources of error with TMP61 you must consider both oversampling and 1-point calibration if possible. In your case, you are using a low pass filter so this should help to mitigate noise. However, due to process variation, there will also be an offset with respect to the ideal response curve of TMP61. We have collected data where we were able to achieve accuracy levels below 1°C after implementing post-processing techniques. Please let me know if you have further questions about this, thank you.

    Best regards,

    Simon Rojas

  • 0 in reply to Simon Rojas Avila
    Intellectual 542 points

    Hello ,

    All of my questions are unanswered in your response, I have been very clear about what are our questions - kindly review my posting once again & help us with insights.

    We want to use TMP61 in TO-92 package yet are concerned about the performance given our configuration - which we have configured after going through the given material & application notes by TI.

    Kindly loop in someone from TMP61 product team if it helps speed up the resolution of these queries.

  • 0 in reply to Neet
    TI__Expert 5825 points

    Neet,

    I disagree with your statement. 

    Is it a safe assumption the errors (min/max) are instantaneous errors & can be averaged out by oversampling?

    "In your case, you are using a low pass filter so this should help to mitigate noise. However, due to process variation, there will also be an offset with respect to the ideal response curve of TMP61" --> meaning that oversampling is not enough. You should consider performing 1-point offset calibration. No, the sources of error are not all instantaneous.

    I cannot help you calculating min and max error of your entire system without knowing details about:

    • The ADC--i.e. effective resolution, ratiometric vs. nonratiometric measurement.
    • How much coupling capacitance from the twisted pair affects the measurement. This will depend on how much noise you expect in your system. TMP6 is particularly susceptible to EMI.

    Even then, we have an error budget calculator included in the thermistor design tool which will assist in calculating error with the Rbias topology. 

    For this reason, I would expect the customer to be able to perform an error budget calculation at the system level since you are aware of all of the technical details of your system. What I recommend is, start with the min and max of the thermistor (stated in the datasheet). And characterize your setup by implementing oversampling and 1-point offset calibration. As mentioned before, we have seen thermistor be able to reach accuracy levels higher than 1°C after implementing these post-processing techniques.

    If you would like, I can get you in touch with a field apps engineer to formally forward your requests to me. I am the applications engineer from the product line supporting thermistors, analog temp sensors, etc. 

    Best regards,

    Simon Rojas

  • 0 in reply to Simon Rojas Avila
    Intellectual 542 points

    Hello ,

    The clarification on the Figure 8-3 was unclear to me, I understand your core point - but then we are not asking you for system level accuracy - just the configuration given.

    However, you connecting me with filed application engineer would be helpful - we look forward to including TNP61 in our system, given our concerns are resolved before we invest in testing & EVM.

    Ideally we are also trying to find a low cost integrated solution for Ibias as it offers a higher dynamic range, but then that post is entirely unanswered so far. Kindly help speed up these queries.

    - Neet

  • 0 in reply to Neet
    TI__Expert 5825 points

    Neet,

    I understand. I have sent you the FAEs information to your E2E direct message inbox. Please let me know if you have any issues getting in touch with them.

    Best regards,

    Simon Rojas