• Locked Locked
  • Replies 6 replies
  • Subscribers 30 subscribers
  • Views 523 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.

3V3 power rail failing on TMDXRM57LHDK development kit

Klaas Van Ammel
Prodigy 35 points
Other Parts Discussed in Thread: HALCOGEN

In our company we bought 2 pieces of this development kit and in short time both have failed with the same failure. After some time using it, the LDO which generates the 3V3 rail is heating up and the rail goes down (TPS73733QDRBRQ1). This also causes the 1V2 rail to go down. It may be that the the rail is down because of the internal thermal/current protection of the device. The failure is not reversible, it doesn't dissapear by shutting down and letting cool down the board. It looks like there is some hardware design error causing the LDO to fail after some time, perhaps it is dimensioned for too small current...

The board is used in a normal way. It is powered with the adapter delivered with the board (part of the kit). Code is generated with HALCoGen for the correct board, so drivers for peripherals are (should be) configured correctly. No modifications to the boards have been made and the interfaces are used in a normal way (Ethernet and USB for programming).

Is this problem familiar to someone?

  • TI__Guru**** 190806 points

    Hi Klass,

      I'm very sorry to hear that the board has somehow not functioned as expected. I have not seen a problem like this before.

      On the HDK, can you probe the pin at capacitor C168 and the capacitor C150? The C168 should measure close to 3.3V and the C150 should measure 1.2V.

      Currently, are you able to erase the flash? If erase is possible, what voltage do you measure at the two mentioned test points after the erase?

      I'm also interested to know what frequency you were running the device at. Could you give me the GCLK:HCLK:VCLK frequency at the time you started to notice the problem.

      Sorry for asking a few more questions. For how long you were able use the board? During this period what type of application were you running? Since two boards went bad, I'd like to know if they all happened to run the same application and failed or they failed with different applications.

     

  • in reply to Charles Tsai
    TI__Guru**** 190806 points
    Also at the upper right corner near the power input on the HDK you will normally see 4 LED lights that indicate the power status of each power rail. What do you see?
  • in reply to Charles Tsai
    Prodigy 35 points

    Hello Charles,

    The output pin of U39 (pin1) w.r.t. GND: 57mV. The output pin of U35 (pin1) w.r.t. GND: 21mV. The 5V rail is fine: 5,00V. The indication LEDs for both the 3V3 and 1V2 rail are off, those for 12V and 5V are on.

    It is not possible to erase the flash when we are trying to debug/load the program.

    Frequency settings: 

    • GCLK = 300 MHz.
    • HCLK = 150 MHz.
    • VCLK1, VCLK2, VCLK3, VCLKA1 = 75 MHz.
    • VCLKA2 = 0 MHz.
    • RTI1CLK = 75 MHz.
    • VCLKA4_DIV = 37,5 MHz.
    • VCLKA4_S = 75 MHz.

    The application has following functionality:

    • LED blinking
    • SPI communication with SD card
    • CAN communication
    • Ethernet, especially sending/receiving UDP packets

    The first board failed during CAN communication. With the second board we were testing Ethernet communication by sending UDP packets. But the board failed when it was doing nothing, it was just powered by the adapter and it was connected with a laptop via USB.

    We used the boards for a couple of weeks (1-3 weeks) before the failure arose.

    After removing the LDO U39 from the board, it becomes clear that there is somewhere a short on the 3V3 rail. The measured resistance is way to low. When connecting an external LDO via a breadboard, the 12V and 5V rail go in hick-up. The current limit in the used breadboarded LDO (TPS79633) is above 2A, making the polyfuse F1 to periodically blow and cool down again.

    Since we suspected one of the many tantalum capacitors to have failed in short, we checked every one of them. None of them were broken.

  • in reply to Klaas Van Ammel
    TI__Guru**** 190806 points
    Hi Klaas,
    Is the HDK plugged in another board or is it standing alone? Can you also restore any jumper settings to its original settings and remove/unplug any components (i.e. SD card, USB, Ethernet ) to the HDK. Please also check the J9, J10 and J11 connectors if anything that might have stuck to create any chance of shorts. With all this done do you still measure low resistance on the power rail?
  • in reply to Charles Tsai
    Prodigy 35 points

    Charles,

    The HDK is working standalone. No connectors have been connected to J9, J10 or J11, nor have any modifications been made to the boards. No jumpers have been connected to any pinheader. Without USB/power/Ethernet/SD card and U39 removed, we still measure a short on the 3V3 rail...

  • in reply to Klaas Van Ammel
    TI__Guru**** 190806 points
    Hi Klaas,
    I will get back to you.