• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 164 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.

DS150DF1610: Internal device temperature.

madphysics
Prodigy 90 points
Part Number: DS150DF1610

Hi,

We're seeing hard lockups on several DS150DF1610's in a design. Where they stop responding to i2c transactions until hard reset and one of our concerns is possible temperature issues. Is there any way to get an internal die temperature out of the device via it's register set? I can't see anything in the data sheet but there are a few reserved/engineering registers. It would be nice to be able to see if we are close to the limit on things.

Other devices (the other 2 DS150DF1610s) still respond, so the i2c bus is not held in an error state it seems.

We're running 10Gbit/sec Ethernet or infinniband style traffic across 12 of the link almost constantly when we see this happen.

Best regards.

Marc.

  • 0
    TI__Guru* 93600 points
    Hi Marc,

    I checked internal register and on this part we cannot measure die temperature.

    One thought is to measure ambient temp, case, or board and device power consumption. From there, we can calculate junction temperature and make sure we are not over the limit(115DegC).

    Having said this, we have not come across a case where device is not responding to SMBus commands - as long as we are operating over the operating temperature.

    Regards,,nasser
  • 0 in reply to Nasser Mohammadi
    Prodigy 90 points

    Thanks for the quick reply, it confirms what I suspected.

    We've mounted small heatsinks on the chips and the surface temperature is not too high, they can be touched without discomfort. So I estimate 60-70 deg surface. We've got 6 of them on the I2C bus, along with about 4 other devices. 

    My current guess's are glitches on our 2.5V power supply or maybe even the i2c bus itself. However we've spent a lot of time testing that, as it's rather an interesting network and we've not seen errors on the bench, it's only when installed in our racks the users have reported occasional lock ups. They are going to try and document which chips are doing it and try to see if there is a pattern that matches data flow etc.

    If we get more info from them I'll do another posting.

    Regards

    Marc.