• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 293 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.

Original question:

TMP116: Thermal pad isolation issue

TMP116: TMP116AIDRVR

Jason Ehsan
Prodigy 20 points
Part Number: TMP116

Hello -

I have an application that inadvertently connected the 3.3V supply rail to the thermal pad of the TMP116 during system integration.  The temperature readings from the sensor were noticeably off (several degrees) after this happened and through some searching on the e2e forum, discovered this related thread that offered a likely explanation.  During our failure analysis of the boards that had the inadvertent shorts, I checked resistance between the thermal pad and the 3.3V and 0V nets on the boards.  I also took the same measurements on some new boards that had not been installed in our system.  Using a Fluke 117, the resistance values on the bad boards were as low as 200 Ohms and generally in the 10k - 100kOhm range.  On the new/good boards, the readings were initially outside the range of the meter (>40MOhms), but then I started seeing some lower readings, e.g. 10k - 100k, and 1M - 30MOhms.

For the TMP116, is there an expected resistance we should see between the thermal pad and the supply rails?  I am concerned that the readings are not always in the high MOhms range.  Is this much variability expected in the sensor itself?  Can I use the resistance readings as an indication of a potential problem?

On our board, the thermal pad is electrically isolated from other nets and is only connected to a copper pad on the back side of the board.  The TMP116 thermal pad is soldered to the board in our application.

Thank you,

J

  • 0
    TI__Expert 6295 points

    Hi Jason,

    The variation you are seeing on the good devices is not a concern as long as the thermal pad is floating or grounded. This is because the thermal pad is not electrically connected to the die so what you are effectively measuring is the resistance/conductivity of the mold compound between the pad/leadframe and the die itself. From a manufacturing perspective, the conductivity of the mold compound can vary as you have found by testing different units. And even though this is not a characterized parameter we do test it in production. I hope this clarifies your concern.

    Best regards,

    Simon Rojas

  • 0 in reply to Simon Rojas Avila
    Prodigy 20 points

    Hi Simon,

    Thank you for sharing this information, it is helpful.  What kind of variation are you seeing in the resistance of the mold compound when you test in production?

    -J

  • 0 in reply to Jason Ehsan
    TI__Expert 6295 points

    Hi Jason,

    Unfortunately I cannot share production result distributions or any kind of test data as this is proprietary information. But I can tell you that the variation that you have measured is within what we see in production.

    Best regards,

    Simon Rojas

  • 0 in reply to Simon Rojas Avila
    Prodigy 20 points

    Thank you Simon.

    Can you share the resistance value threshold of the mold compound that determines whether a part passes or fails?

    -J

  • 0 in reply to Jason Ehsan
    TI__Expert 6295 points

    Hi Jason,

    This test is performed a bit different from what could be useful for you. The overall idea is to measure current flowing out of the thermal pad which should not be more than a few uA and that is all I am able to tell at this time.

    What I would recommend is performing some kind of leakage test on the thermal pad by bringing it slightly below ground using an SMU and measuring the current. You should be able to see a difference when comparing the known good devices vs. the bad ones.

    Best regards,

    Simon Rojas

  • +1 in reply to Simon Rojas Avila
    Prodigy 20 points

    Hi Simon,

    Just writing to let you know that I will be performing these tests next week.  Thank you for the recommendation.

    -J

  • 0 in reply to Jason Ehsan
    Prodigy 20 points

    Hi Simon,

    We performed some measurements today using an Agilent B2902A.  Channel 1 was set to supply 3.3V and 0V to the sensor.  Channel 2 was used to bring the thermal pad below 0V.  On sensors that passed our functional test, < 1uA was observed on Channel 2 down to -25mV.  On a sensor that failed our functional test after being damaged in our system, I immediately hit 100uA (limit set on the SMU) when connecting the thermal pad to 0mV.  Idle current consumption of the bad board is also higher.

    Thank you,

    -J

  • 0 in reply to Jason Ehsan
    TI__Expert 6295 points

    Glad to hear Jason, thank you for following up. I will close the thread now.

    Best regards,

    Simon Rojas