• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 550 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.

LM95235 accuracy evaluation

Abe Chen
Prodigy 30 points
Other Parts Discussed in Thread: LM95235

Hi TI expert,

In my test, I found the temperature from LM95235 is about 12oC bigger than the CPU internal sensor. However the CPU vendor claim the value is at 4~50C deviation in Max.

The gap here may cause the thermal design failed.

 

So I need a way to evaluate the deviation of LM95235 value.

Do you know is there any other factor casuing the deviation(except  the intel mode, it bring in 2 C deviation in Max)? 

How to evaluate the track impedance influence? Is there any calculation model available?

  • 0
    TI__Genius 15760 points

    Hi Abe,

    Thanks for using TI E2E forums. I will ask the SVA team to reply to your question.

    Mayrim

  • 0
    TI__Expert 4905 points

    Hi Abe,

    Is this an Intel processor? If so you can use the TruTherm mode to bring the reading back to the proper value.

    If not then the standard 3904 mode should work.

    Series resistance indeed can cause an offset in the temperature reading. I suggest you check out the LM95235 datasheet applications section title "Calculating Total System Accuracy" (page 25). For every ohm of resistance you get about 0.62C of error. You can compensate several ohms with offset compensation but we do not recommend compensating 10's of ohms in this manner. I would be best to go to a device that include resistance cancelation.

    In addition you need to know the ideality factor for the target thermal diode to calculate and offset. Please refer to the "Compensating for Different Non-Ideality" section in the LM95235 datasheet (page 26).

    Noise coupling into the diode lines can also cause an error that looks like an offset (basically the diode rectifies the noise so you get an offset). Make sure you have a really good layout (treat the diode traces very carefully). See the "PCB Layout for Minimizing Noise" section in the datasheet.

    I hope this helps...