• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 34 subscribers
  • Views 1000 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.

LMT70: LMT70

Kailyn Chen
Genius 9560 points
Part Number: LMT70
Other Parts Discussed in Thread: TMP117

Hello，

Our customer use LMT70 to measure Skin temperature, but he found that the  response time  changes slower before after about 33 degree, the whole time need about half an hour to get stable, so he'd like to confirm if it need so long time?  Or if we need below 300ns to get stable, how to resolve it ?

Best regards

Kailyn

  • 0
    TI__Guru**** 244380 points
    Hello Kailyn,

    The typical 63% response time for the LMT70 for Stirred oil is 1.7 sec and in still air it is 73 sec. An important note here is that these parameters are dependent on the PCB layout.

    Q.1: Can you please let us know how the device is connecting to the skin?
    Q.2: The device layout on the board?
  • 0 in reply to Amit Ashara
    Intellectual 260 points

    Hi Amit,

    This is the sensor part of out product, the metal is Stainless steel.

  • 0 in reply to xinggang zheng
    TI__Guru**** 244380 points
    Hello Xinggang,

    How is the LMT70 connected to the stainless steel plate. Is the glue material thermally conductive? The path of thermal conductance is via the ball which seem to be on the flex PCB, creating a longer path for the heat to move from the steel plate to the device. This may explain why there is a large setting time.

    I am double checking on the layout guidelines to see if we can verify the thermal path. But in the meantime can you confirm the thermal conductance of glue material for the application?
  • 0 in reply to xinggang zheng
    Prodigy 20 points
    I am twofold keeping an eye on the design rules to check whether we can confirm the warm way. Be that as it may, meanwhile would you be able to affirm the warm conductance of paste material for the application?
  • 0 in reply to Ariel mazicko
    TI__Guru**** 244380 points
    Hello Ariel,

    What paste material are you referring to here? Is it the paste between the steel and the package itself?

    Instead of the steel path for conductance, if you temp touch on the flex board itself where the pads are soldered, is there a change in the response profile?
  • 0 in reply to Amit Ashara
    Expert 2690 points

    Amit Ashara said:
    What paste material are you referring to here? Is it the paste between the steel and the package itself

    Hi Amit,

      The paste is K-5205 from Kafuter. Here I paste some parameters of this thermal silica.

       Besides, LMT70 is mounted on soft FPC , here is the PCB :

      

  • 0 in reply to SeaFesse
    TI__Guru**** 244380 points
    Hello SeaFesse,

    Thanks for the data on both the paste and the PCB layout. They look to be fine here. As I mentioned, the primary path for the heat transfer is from the balls to the die (www.ti.com/.../snoa967.pdf Page-5; Section 1.2.1). The paste may not make a good contact with the pads of the device.

    Secondly, testing it on the human body directly and monitoring the temperature response is not the right setup. I would suggest a thermal chamber to test the device and its direct output on a scope; to ensure that there is no layout issues.
  • 0 in reply to Amit Ashara
    Expert 2690 points

    Amit Ashara said:
    I would suggest a thermal chamber to test the device and its direct output on a scope; to ensure that there is no layout issues.

    Hi Amit,

       In fact, we have done some similar experments.  The following 3 pictures is from 3  prototypes, and  it's tested under a constant-temperature water  tank.  Y-axial is the ADC12 sample value of LMT70 . X-axial is the sample

    counts value which  sample every 500ms. so we can see that the the tempturature response time is in the range between 50s-80s ,which is acceptable. However, if we place the prototypes on human body,

    the response time always last a very long time, sometimes almosy 1 hour.  That's our concern and issue.

    B.R

       Seafesse

     

  • 0 in reply to SeaFesse
    TI__Guru**** 244380 points
    Hello Sea

    In a constant temperature bath, the device is being heated well from the pads, i.e. there is a good thermal conductivity path. When connecting to the human body, the path of heat may not be reaching the pad of the device, which is why there is a such a slow response. I would have the following 2 recommendations

    1. Get the thermal paste to make good contact with the ball pads.
    2. Use a device like TMP117 which has a thermal pad for temperature conductivity, that can be made available on the PCB.