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.

Skin temperature measurement with LM94022 or TMP112 or TMP102

Other Parts Discussed in Thread: TMP102, LM94022, TMP112, LMT70, TIDA-00824

Hello,

I would like to make a skin temperature measure, on the wrist. So I selected 3 temp sensor. The LM94022, the TMP112 and the TMP102, which are different about resolution, accuraccy and price, so I will do a choice later.

-Are these parts suitable for this ?

- I imagine a thermoconductor material between the top of the package and the skin, is a good idea to make this measure ?

  • Hi Samuel,

    Are you open to other temperature sensors? If so I recommend using the LMT70 which has 0.13C accuracy at the human body temperature range of 20C to 42C. In fact there is a TI Design about this very application: www.ti.com/.../TIDA-00824. Please see the TIDA-00824 Test Results document for more details of the implementation.

    -Michael Wong
  • Hi,

    I saw this part but I would have wanted to avoid it because of the package is very very small and it's BGA layout so it's not possible to wired it myself or it appears difficult. In more it's an expensive part too...

    But I saw the advantages, it's a little part, therefore good thermal response on small PCB and very precise charateristics. I understand well the utility of this sensor in this application but I have all the same some questions with regard to the other sensors and the application in general.

    - In the application note (which is interesting), the aim is to create a little PCB for better thermal conduction, so can I consider to use the same principle for the part I mentionned especially for the LM94022 ?

    - Is there a minimal thermal conductivity value recommanded for the thermal epoxy ?

    - I'm not sure, can you confirm that the small LMT70 PCB of the application note is completely potted in the thermal epoxy ?
  • Hi Samuel,

    Do you have an accuracy requirement for your skin temperature sensing? This will determine which IC will be best for your application.

    But yes, you can use the tiny PCB design or mount the temperature sensor on a flex PCB. I cannot recommend an exact thermal conductivity value but it should be high. The tiny PCB was placed on a bed of thermal epoxy to provide a good thermal transfer between the case and the PCB/sensor. The thermal epoxy can also be used to fully encapsulate the temperature sensor.

    -Michael Wong
  • Hi,

    Unfortunately, I don't know which accuracy I need, but I think to test the 2 sensors LM94022 and LMT70 to compare the performance.

    It's OK for me, thank you for support.

    Sam
  • Hi Samuel,

    You can evaluate the LM94022 and the LMT70 by ordering their respective EVMs:

    www.ti.com/.../lm9402xevm
    www.ti.com/.../LMT70EVM

    -Michael Wong
  • Hello,

    I have a question about the accuracy of LMT70/LM94022. For exemple on the LMT70 datasheet we find  +/- 0.13°C max from 20 to 42°C.


    My question is : if I have a temperature of 37°C to measure, if my sensor measures 37.1°C, I have an error of 0.1°C, OK. With this same sensor, is this temperature error of 0.1°C the same in the complete range from 20 to 42°C ? in fact ? does this accuracy error behave as an offset ?

    Same question for the LM94022.

  • Hi Samuel,

    It may or may not. When I tested this sensor, some parts showed a stable offset while others shifted throughout the temperature range while still maintaining the specs in the datasheet (+/-0.13C at 20-42C).

    -Michael Wong
  • Hi,

    Thank you for the fast reply.

    Regards,

    Sam