• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 144 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LMT84-Q1: temperature sensor failure analysis

Arvind Munde
Prodigy 100 points
Part Number: LMT84-Q1


Tool/software:

Hi Team,
We have been using this sensor in large volumes for over six years, and it has consistently performed well without any failures. However, in our new design, we recently observed two instances where the sensor failed after testing. In both cases, the sensor output was 0V, whereas it should have been approximately 0.9V under normal operation.

It's important to note that during start of test, both the supply voltage and output voltage were found to be within the specified limits. Also we have crosschecked by replacing sensor with healthy one on same ECU and checked faulty sensor on healthy ECU. This sensor is powered by 5V LDO. 

In light of this, we request your support in identifying the possible root causes of failure in this new design and guidance on how to systematically verify what might have gone wrong.

  • 0
    TI__Genius 12985 points

    Arvind,

    I want to clarify a few things. When you say that the sensor output voltage was within specification during the start of the test, does that mean the LMT84s in question were good initially and then found later to have a 0V output? When you swapped the healthy sensor and the faulty sensor, did the failure follow the sensor or the PCB?

    What is the failure rate of LMT84s in the new design? Are the two instances different in any way? If you have performed an A-B-A swap and found that the issues follow the 2 parts, and there many other examples of the new design working just fine with other LMT84s, you may need to submit this as a customer return.

    Thanks

    -Alex Thompson

  • 0 in reply to Alex Thompson
    Prodigy 100 points

    Yes, at the start of the test, we verified that the sensor data was within the expected range. However, after some time, the ECU reported an over-temperature fault and was unable to continue operation due to this condition.

    We replaced the sensor in the faulty ECU, after which it successfully passed all tests. Additionally, when we assembled the same faulty sensor onto a known-good ECU, that ECU also raised an over-temperature fault.

    It is worth noting that both failures occurred during the same type of test. However, another ECUs subjected to the same test, equipped with the same type of sensor, is operating normally and has accumulated significantly more runtime (several times that of the failed units without any issues).

  • 0 in reply to Arvind Munde
    TI__Genius 12985 points

    Arvind,

    Thanks for all the additional details, this is very helpful for me to understand the scope of the issue. Since the issue follows the parts, and your ECU is shown to work OK with known good sensors, I think you should submit these parts as a customer return.

    How long does the failure take to appear after it starts up OK? What are the test conditions that are triggering this failure? If TI accepts this as a customer return, we would like to be able to replicate it. 

    As for why this is happening, I'm not sure. If the PCB is passing with a new unit I don't think anything on that PCB has failed to bring the output to 0V. I think this may be something we want to do FA on.

    Regards

    -Alex Thompson