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.

TMP75: sensor reading correct or not

Part Number: TMP75
Other Parts Discussed in Thread: TMP1075, TMP461

Hi Sirs,

Sorry to  bother you.

We have some questions regarding TI TMP75 thermal sensor.

Please help to clarify.

 

1. How to make sure that sensor reading is correct? Could you share some experience for us?

2. If the sensor reading is incorrect, how to do correlation on it?

Thanks!!

  • Hi Shu-Cheng,

    To check for accuracy the device can be set to a reference temperature using an accurate and reliable temperature source.  The reference temperature from these temperature sources should be checked with a calibrated high accuracy thermometer and be placed as close to the device as possible.  This can be correlated to what the device is reading.  

    If you are seeing incorrect temperature readings please review below app notes and be sure to follow recommended techniques for accurate temperature measuring.

    Also, please look into using TMP1075 for your design.  This device is Pin to Pin and software compatible part that can be up to 4 times lower power!

    http://www.ti.com/lit/an/snoa967a/snoa967a.pdf

    Thanks,

  • Hi Sirs,

    Thanks for your reply.

    More question as below.

    Please help to clarify.

    To check for accuracy the device can be set to a reference temperature using an accurate and reliable temperature source.  The reference temperature from these temperature sources should be checked with a calibrated high accuracy thermometer and be placed as close to the device as possible.  This can be correlated to what the device is reading.  

    1.     “a reference temperature using an accurate and reliable temperature source” means that we heat up our device and use thermocouple to measure the temperature and then compare with the reading from TMP75 at the same time.

    Is my idea correct?  If there are any misunderstanding ,please let me know.

     2.     In our application, if we use TMP75 to remote access some device temp on PCB.  I think it’s still need to do sensor correlation on this application scenario, right?

  • Hi Shu-Cheng,

    1.

    Yes your idea is correct, you can heat or cold device and use thermocouple to measure the temperature and compare your results.  Be sure allow sufficient soak time to have a stable temperature measurement.

    2.

    What is the application and what do you mean by remote access the device temperature?  If the device you want to measure the temperature has an internal thermal diode/BJT structure you can use one of our Digital temperature remote devices such as the TMP461.

    Thanks,

  • Hi Sirs,

    Thanks for your explanation.

    Regarding your reply in question 2, it means that TMP75 don’t have such function to  remote access device temp, right?

  • Hi Shu-Cheng,

    If you mean remote access a different device's temperature other than the TMP75, then that function is not available on the TMP75. 

    I would recommend if you are trying to monitor a device's temperature other than the TMP75, and that device has access to an internal thermal diode/BJT, to use a Temperature Remote device such as the TMP461.

    Thanks,

  • Hi Sirs,

    Thanks for your reply.

    Could you talk more specific for those description? What is concern based on your understanding if pop internal thermal diode(TMP75) to monitor PCB hot spot, not remote device you recommend here? I don’t think there is much deviation between internal and remote diode.

  • Hi Shu-Cheng,

    When talking about "remote access to device temperature" it is typically referring to reading an external diode/BJT temperature that is NOT the one in the local temperature sensor (in this case the TMP75).

    In this case it sounds like you are measuring a single PCB hotspot which a local temperature sensor such as the TMP1075 or TMP75 can be used.  

    Thanks,

  • Hi Sirs,

    Thanks for your reply.

    According to temperature result that we measure on the TMP75 in our GPU card, it shows the around 15C temperature gap between the sensor reading and thermocouple. And I also refer to TMP75 datasheet shows the thermal resistance is 76.1( ̊C/W), it means that once “1W” through the TMP75 and will cause the 76.1C temperature gap, please let me know my understanding is correct or not, if my understanding is correct, how can we prevent temperature gap in our case.

  • Hi Shu-Cheng,

    Please look at section 2 and 3 of the below is an app note which explains the parameter you have in question.

    These parameters are used for devices with self heating. 

    Which is reading the hotter temperature the thermocouple or TMP75?   Where is the thermocouple located compared to the device?  

    http://www.ti.com/lit/an/spra953c/spra953c.pdf

    Thanks,

  • Hi Sirs,

    Thanks for your reply, update as below.

    Thermocouple is on the top surface of TMP75, and TMP75 sensor result is always higher than thermocouple result.

  • Hi Shu-Cheng,

    What are you using to keep the thermocouple connected to to the TMP75?

    Thanks,

  • Hi Sirs,

    We use the glue for thermocouple on the TMP75.

    The glue we used is Satlon High Temperature Resistant Binders D3 glue and Primer 606  Catalysts.

    And below link is detail info.

    Please feel free to let me know if you have any.

     

    D3 glue : http://satlon.com.tw/eng/products/p_5.htm

    Primer 606  Catalysts : http://satlon.com.tw/eng/products/p_6.htm

  • Hi Sirs,

    Sorry for pushed, any update on here?

    Thanks!!

  • Hi,

    Sorry for the delay,  the top of case does have a high thermal resistance which may take some time to be in thermal equilibrium with the die temperature.  I would suggest placing the thermocouple on the PCB hot spot as close as possible to the TMP device and to get a faster temperature correlation. 

    Thanks,