• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 519 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.

TMP117: Accuracy, offset and slope

Lorenzo Merlo
Prodigy 10 points
Part Number: TMP117

Hello,

we are looking to TMP117 as possible replacement for thermistors.

Graphs at paragraph 6.9 of the datasheet already provide useful info about accuracy but I have a few additional questions, specifically for temperature range 10-40°C. 

If I take one TMP117 sensor with its own error characteristic, how much of the error could be an offset and how much could be a slope error?

In other words, if a sensor has 0.08 (for example) accuracy, can it have error of  +0.08 at 10°C and -0.08 at 40°C?

Would the slope be somehow constant or could the sensor show many ups-downs within the accuracy bandwidth? (always in range 10-40°C)

And if I take more TMP117, how much could  the slope error be spread among them?

Any clarification related to these questions would be welcome.

Thank you, Lorenzo Merlo

 

   

  • 0
    TI__Guru* 91206 points

    Dear Lorenzo - 

    Thanks for the post and welcome to E2E! 

    With the TMP117 line of devices, to answer your question directly, you can reference the image on the front page of the datasheet to see the average and min/max lines that are based on package usage, then you can go to table 6.5 on page 6. In this table, you can see in the temperature range you are concerned with, the min/max for this device is ±0.1ºC, and typically, with "8 averages" averaging turned on (inside the part, which is default) you will typically see ±0.05ºC. The averaging takes out any "peaks and valleys" from the result naturally, and it can be turned up or off (Section 7.3.2 of the datasheet goes into detail on this topic)

    Let us know if you have more questions! 

  • 0 in reply to Josh Wyatt
    Prodigy 10 points

    Thank you for your answer.

    I understand that the "8 averages" mode takes out peaks and valleys, but what about overall slope?

    Can I expect a sensor to have, for example, -0.05 error at 10°C and +0.05 at 40°C?

    It would be nice to see example error curves of single sensors, to understand how the slope of a single sensor can be and how diffferent sensors compare.

    Thank you

  • 0 in reply to Lorenzo Merlo
    TI__Guru* 91206 points

    Dear Lorenzo -

    This part by default is averaged, you can see Figures 6-2 and 6-3 in the datasheet which represent what can be observed from each package type under those conditions. We do not publish data or specifications in our datasheets which is/are made from single parts, because that would be irresponsible of us to do. The min/max and typical values listed in our datasheets are made from large sets of data, from multiple lots of parts, and you can therefore depend on them. 

    If you want to place single devices into mode where there is no averaging and test, that is certainly something you can do with an EVM or your own hardware, if you like. 

    EVM is here ==>  https://www.ti.com/tool/TMP117EVM 

  • +1 in reply to Josh Wyatt
    Prodigy 10 points

    Dear Josh,

    thanks for the explanation.

    I still have some question marks in my head about slope, even in case of averaged readings; but I think we will have to investigate on that by our own.

    Regards, Lorenzo

  • +1 in reply to Lorenzo Merlo
    TI__Guru* 91206 points

    Lorenzo - sounds good - another way to look at this is that the averages readings should result in any slope being almost zero, meaning the output of the device will be close to a flat line and would appear even more so if you zoomed out from the graphs shown in the datasheet to reduce resolution. Do let us know if you have more inquiries.