• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 929 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.

Original question:

LM35: High failure rate with hand-soldered LM35 TO-92s

LM35: Intermittent high output voltages with hand-soldered LM35 TO220

Barry McGinley
Prodigy 30 points
Part Number: LM35
Other Parts Discussed in Thread: LM339

Firstly, thanks for your time and attention.

We are using the LM35DT/NOPB attached to a heatsink and soldered onto a PCB. We have noticed intermittent failures similar to failure mode 1 in the original thread "High failure rate with hand-soldered LM35 TO-92".

Our unit performs a power on self test and sometimes an over temperature condition is detected. What is perplexing is that the failure happens very infrequently but it happens in clusters so that we may power cycle the unit 4 or 5 times in a row with it failing each time with an output that is too high to be a real temperature( 1.2V ) and then it will power up and pass on the next POST. We recently ran almost 5000 power on/off cycles without seeing any failures. Has TI ever observed a power up behavior like this? Could a transient voltage between Ground pin and the TO220 tab during power on cause this? 

Regards,

Barry

  • 0
    TI__Expert 7855 points
    Barry,

    I will need to check with our team to determine if this issue exists. I personally haven't seen this before.
    But first, please provide me with more details on to the test setup. A circuit schematic of the setup would be helpful.
    What is the power up time? Does this behavior occurs if you slow down the power up ramp?

    -Kelvin
  • 0 in reply to Kelvin Le THS
    Prodigy 30 points
    Hi Kelvin,
    Thanks for the quick response. In answer to your questions:

    1. Our circuit is pretty simple, the LM35 is powered by +5VDC and bypassed with a 0.1uF cap. The output connects to a series 1k resistor which connects to a 0.1uF cap shunted to ground. That RC node connects to a second 1k/0.1uF series/shunt and that second cap connects to the inverting input of an LM339 comparator.




    2. I haven’t measured power supply ramp time. I didn’t see a specification on VCC ramp time in the datasheet, does it affect this device?

    Regards,
    Barry
  • 0 in reply to Barry McGinley
    TI__Expert 7855 points

    Barry,

    You're using the LM35 as a switch? We have dedicated temperature switches in our portfolio FYI.
    So you have two RC low pass filters after the LM35 correct? During the failure events, you measured the output pin of the LM35 directly (before the RC filters) and got 1.2V? How far is the LM35 output from the comparator?

    Can you also tell me more about the heat sink? What is it connected to? Is this part is also power cycled during the test?
     
    Have you tried doing the same power cycling test but without the external RC loads and comparator?

    There are no specifications for the supply ramp time. I was curious to see if that has an effect on your failure rate.

    -Kelvin

  • 0 in reply to Kelvin Le THS
    Prodigy 30 points

    Kelvin,

    We did measure the 1.2V voltage directly at the LM35 output pin.

    The LM35 is a couple of inches from the LM339 comparator, they're both on the same PCB.

    The LM35 is mounted to a heatsink. There are also two power MOSFETs on that heatsink but they are electrically isolated from the heatsink.

    The entire system is power cycled.

    We tried shorting the first 1k resistor to see if we could induce the failure with a 0.1uF capacitive load on the LM35, we were unable to induce the failure.

    Barry

  • 0 in reply to Barry McGinley
    TI__Expert 7855 points
    Barry,

    Are you saying that the failure is dependent on the load?

    -Kelvin