• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 83 subscribers
  • Views 336 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.

Compiler/F28M36P63C2: MCU always working in high temperature

Xiyao Zhao
Prodigy 50 points

Part Number: F28M36P63C2

Tool/software: TI C/C++ Compiler

Hallo, 

We are using F28M36 P63C2ZWTT YFF-82ADFOW on our PCBs to control the joints of the robots. The chip is also running at a very high temperature, about 67 degree measured on the chip surface. The chip stops working in less than 1 minute without heatsink and could work 2 minutes with a 10mm*14mm*6mm ( W*L*H) heat sink on it. We try 14 and the 14 are all more or less the same. But before we use F28M36 P63C2ZWTT YFF-82APZEW. (The schematic, the layout, the PCB material and the factory stays the same.)  The chips with this number is working well, but still with a high temperature 67 degree.

We use TMS320C28X Optimizing C/C++ complier V18.1.0.LTS.  and code composer studio as IDE. 

Can anyone help me why the MCU is always so hot ?

Why  F28M36 P63C2ZWTT YFF-82ADFOW could not work with the same high temperature, but F28M36 P63C2ZWTT YFF-82APZEW can.

If anything more is needed, please just let me know.

Thanks a lot!

Xiyao Zhao

  • 0
    TI__Guru* 95315 points

    Xiyao,

                  Is this design already in production or is it in the prototyping phase? 

    So even with a heat sink, chips with a LOT trace code of YFF-82ADFOW stops working in 2 minutes? Could you please elaborate on what you mean by "stops working"? Exactly what is the chip doing? Just frozen? Are you able to recover it by a watchdog reset or activating -XRS or power cycling?

  • 0 in reply to Hareesh Janakiraman
    TI__Guru* 95315 points

    Xiyao,

                  Could you please take a look at this erratum in SPRZ375L and assess whether it is applicable in your design? 

    GPIO: GPIO38 and GPIO46 Signal Latch-up to VSS Due to Fast Transient Sensitivity at High Temperature

    Could you try this experiment: Erase your application in flash memory. Program a simple GPIO pin toggling code and check whether it works at > 67'C. Do not use GPIO38 or GPIO46. Use some other GPIO, preferably one with an L.E.D connected, so it is easy to observe the status of the GPIO pin, without having to connect an oscilloscope.

     

  • 0 in reply to Hareesh Janakiraman
    Prodigy 50 points

    I would try this and measure the current into the chip. Now it is in prototyping phase.

  • 0 in reply to Hareesh Janakiraman
    Prodigy 50 points

    Thanks a lot for your advice.

    Regards！

  • 0 in reply to Hareesh Janakiraman
    TI__Guru* 95315 points

    Xiyao,

                  In your design, is there any provision to measure the operating current? If so, can you compare the currents between a board that works and one that doesn’t? 

    Also, is there some way you could cool the YFF-82ADFOW device            ? One way to do it would be to use something like (Chemtronics) Freeze-It spray. A better method would be direct very cold air on the heat sink. Not sure if you have any such equipment in your lab. I am interested in knowing if cooling the device helps it operate without shutting down. 

    In the GPIO experiment I suggested in my previous post, please repeat the experiment with GPIO38 and/or GPIO46. Check the level of the pin under the problem situation with an oscilloscope

  • 0 in reply to Hareesh Janakiraman
    Prodigy 50 points

    No, I have to sold a cable to measure the current.

    We tried to cold down the chip with a small heat sink on it. It does not work. Then we add a fan to cold it further. Then the chip work. 

    After measuring the current. I would give you a feedback then.

    Thanks for your help!

    Best regards!

  • 0 in reply to Hareesh Janakiraman
    Prodigy 50 points

    Hi, Hareesh,

    We are not using these pins. Thanks!

    Regards!

    Xiyao

  • 0 in reply to Xiyao Zhao
    TI__Genius 17381 points

    Hi Xiyao,

    F28M36 P63C2ZWTT devices are rated and tested well beyond 67 degrees and so should not have an issue with performance at this temperature, so long as the application is using the device within datasheet specifications.  It is also common see slight performance differences (within DS performance specifications) across different lots.  

    Can you describe what "Stops working" means?  For example does the device reset?  Does the code freeze in an ITRAP?  Is the device still running but just not working in the application as expected?

    Have you been able to narrow down what your application code is doing when it fails?  For example is it during a read or code execution from flash?  Is the watchdog timing out?  Is it a particular peripheral in action when the issue occurs? 

    If we are able to get more details about the failing condition, we may be able further suggest some specific things to investigate.

    Regards,

    Joe