• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 3 answers
  • Subscribers 31 subscribers
  • Views 1470 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.

TMP116: Is the TM116 temperature shiftted to upper when Continuous Conversion mode?

Kazuya Nakai59
Guru 20130 points
Part Number: TMP116
Other Parts Discussed in Thread: , TMP117

 Hello guys,

 One of our customers is evaluating TMP116 using the device EVM (TMP116: http://www.tij.co.jp/jp/lit/ug/snou145a/snou145a.pdf). In their evaluation, they found TMP116 output temperature data is higher 0.21C than the data of very high accuracy thermometor they provided when Continuous conversion mode is used. Then TMP116 temperature data is 20.035C and the thermometer is 19.825C. But when one shot mode is used, the temperature data is same as the thermometer data (19.825C).

 Is the cause of the temperature difference in the continuous conversion mode TMP116 self heating?

 Also the device datasheet says that the temperature error with Continuoue conversio mode is about 60mC at 5V. But the error temperature the customer measured is 210mC. It is 3 times error compared with Figure 4 on page 4 of the device datasheet.

 Could you please tell me your idea why the 3 times error is occurred?

 Your reply would be much appreciated.

 Best regards,

 Kazuya Nakai.

  • 0
    TI__Mastermind 31600 points
    Hi Kazuya-san,

    Thank you for reaching out to us. Could you please provide me your test setup? Are you using the oven chamber or you placed the board on the lab environment? How close the high precision probe and the TMP116?

    Aaron
  • 0 in reply to Aaron Heng
    Guru 20130 points

    Hi Aaron,

    I got several information from the customer.

    They don't use temperature chamber. Instead of that, they prepared a very dark room with no window and small box in the room for the evaluation of TMP116. all sensors are in the box.

    Also the reference temperature is measured by high accuracy thermometer developed by themselves.

    The distance between TMP116 and the thermometer is about 10cm(100mm).

    They use TMP116 sensor portion of EVM with dividing at the breakouts point of EVM and connect the PCB TMP116 mounted to USB control PCB with 10cm wires.

    The following graph is  their measurement result with Continuouos Conversion Mode. The X axis is Time (seconds) and Y axis is Temperature measured by each sensor. According to the graph, the temperature measured by TMP116 is about 0.2C higher than the reference temperature at 300 seconds.  The customer knows that the temperature data shows higher temperature than actual one when the continuous conversion mode. But they'd like to know whether 0.2C higher is reasonable or not?

    Could you please give me your answer or comment?

    Also the following is the data measured by one shot mode. You can see the temperature data is higher than the references at the measurement starting point. Is TMP116 default conversion mode continuous conversion mode? Is this phenomenon reasonable?

     Could I ask you a few questions more?

     1. Does the customer need to insert any material between TMP116 and PCB board for higher accuracy measurement?

     2. Do you have any recommended resist on copper pattern for higher  accuracy measurement?

     3. Do you have any Long-term stability and drift data more than 300hours? For example 1000hous case, 3000hours case or more.

     Could you please give me your answer for each question?

     Thank you and best regards,

     Kazuya Nakai.

      

     

  • 0 in reply to Kazuya Nakai59
    TI__Mastermind 31600 points

    Hi Kazuya-san,

    Thank you so much for providing the detailed information. There are two possibilities that could cause the temperature offset. According to your description, customer was using the TMP116EVM to perform the temperature correlation.

    • I realized that in the TMP116EVM, the TMP116 thermal pad is soldered down. . We learned that soldering the thermal pad creates a lot mechanical stress on the die since the sensing element is in the corner of the die. So, when the DAP is soldered, the edge of the 4 sides are pulling and stretching downward. For this reason, it effects the temperature accuracy for ±0.1°C, but still meets at ±0.2°C.
    • In the EC table of page 5, the Standby current and Shutdown current are 1.25µA and 250nA. Normally, one-shot mode performs better temperature accuracy than continuous mode due to less self-heating effect. This due to the time delay between one-shot conversion and temperature data read is long enough to allow self-heating to be reduced. However, in continuous mode self-heating effects are usually much larger since the default delay between conversions is typically very small.  The self-heating effect can be reduced somewhat by increasing the conversion cycle time as indicated in table 7 on page 26 using the 3 CONV bits in the configuration register. I would suggest to increase at least 500ms or 1 second to see the effect.
    1. Does the customer need to insert any material between TMP116and PCB board for higher accuracy measurement?
      • No
    2. Do you have any recommended resist on copper pattern for higher accuracy measurement?
      • No
    3. Do you have any Long-term stability and drift data more than 300hours? For example 1000hous case, 3000hours case or more.
      • Unfortunately, we do not have any long-term stability and drift data more than 300 hours; however, we are planning to do this for TMP117 since they are the same package and P2P replacement. At this moment, I do not have the data to share with you. If you are interested in 300 hours report, I can send you the data offline for your information.

    Hope this helps. Let me know if you have further questions.

    Aaron

  • 0 in reply to Aaron Heng
    Guru 20130 points
    Aaron,

    Thank you very much for your prompt replies and I'm sorry for my late response. Your replies are very helpful for the customer and us.

    But could I ask you a few addtional questions?

    1. Do you have any measurement data for temperature error using the EVM with separated TMP116 portion?
    2. Do you think the customer measurement data is not strange?
    3. Do you recommend that the power pad of TMP116 package is no soldering for high accuracy?
    But I think delay time between actual temperture change and measured temperature change would be larger when the power pad is not soldered. Also some noise may be injected. Is my understanding correct?
    4. Could you please tell me what the different feature of TMP117 is compare with TMP116? Is it higher accuracy? When is the device RTM timing?

    Thank you and best regards,
    Kazuya Nakai.
  • 0 in reply to Kazuya Nakai59
    TI__Mastermind 31600 points
    Hi Kazuya-san,

    I will take this offline. I will send you a friend request notification then we will take from there.

    Aaron
  • 0 in reply to Aaron Heng
    Guru 20130 points
    Hi Aaron,

    I sent you friend request notification yesterday. Could you please accept it?

    Thank you and best regards,
    Kazuya Nakai.
  • 0 in reply to Kazuya Nakai59
    TI__Mastermind 31600 points
    Hi Kazuya,

    Thank you for reminding me. Please check your private session for the response.

    Aaron