• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 99 subscribers
  • Views 434 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.

TMS320C6678: C6678's CVDD pin seems to be grounded

Charles Chen
Prodigy 80 points
Part Number: TMS320C6678

Hi,all:

I have 13 TMS320C6678 boads with radiator.They run the same test at 85℃ envirment.

There is a temperture sensor near the DSP in each board,and will get the temperture once per minute.

During the test,I found that the temperture with 6 boards was about 100℃.

,another 7 boards was over 105℃ and the system was crashed,then I makrked them as abnarmal boards. 

Then I use Fluke 15B multimeter to mesure the resistance between CVDD pin to GND.

I found that the resistance with abnormal boards were less then 10 ohm.

Another boards of 13 were over 20ohm.

My question is:

1.What is the range of the normal resistance value between CVDD pin to GND ?

2.If the resistance between CVDD pin to GND is less then 10ohm,is it narmal?

  • 0
    TI__Guru**** 161600 points

    Hi,

    How did you measure the resistance? Between CVDD pin and the GND layer of the board or between CVDD and VSS?

    Do you have a reference TI EVM to perform the same measurement and compare the results? The TMDSEVM6678 should provide you with a normal resistance value between CVDD and VSS.

    Also see this thread:
      https://e2e.ti.com/support/processors/f/791/t/482325

    Best Regards,
    Yordan

  • 0 in reply to Yordan Kovachev
    Prodigy 80 points

    Hi,

    Thank you for your reply.

    I use the multimete to mesure that the red probe is on CVDD pin and the black probe is on GND pin.

    And I also read the post you provide.I also get the narmal values of the resistance between CVDD and GND base on another normal board.

    Now I want you to ask me the rang of TI`s reconmand about the first question.

    And I want you to judge the second question I ask is normal or abnormal.

  • 0 in reply to Yordan Kovachev
    Prodigy 80 points

    Hi,

    Thank you for your reply.

    I use the multimete to mesure that the red probe is on CVDD pin and the black probe is on GND pin.

    And I also read the post you provide.I also get the narmal values of the resistance between CVDD and GND base on another normal board.

    Now I want you to ask me the rang of TI`s reconmand about the first question.

    And I want you to judge the second question I ask is normal or abnormal.

  • 0 in reply to Charles Chen
    TI__Guru**** 161600 points

    Hi,

    This is not characterized as you imagine, because the manufacturing of the board also impacts this. That is why I referred you to the above e2e thread. Tom explains this issue in details there.

    For the second question, you could use the value of the resistance between cvdd and gnd on a TMDSEVM6678 and compare.

    Best Regards,
    Yordan

  • 0 in reply to Yordan Kovachev
    Prodigy 80 points

    Hi,

    I have already measured TMDSEVM6678,and found that the resistance value between CVDD an GND is about 31ohm.

    And the resistance value of  boards that failed in 85℃ enviroment are about 10ohm,but they run normally in 30℃ enviroment.

    The value of TMDSEVM6678 is larger than our boards.

    So can I jude that our boards have already damage?

  • 0 in reply to Charles Chen
    TI__Mastermind 46460 points

    Charles,

    I believe what you are seeing is not real.  You refer to it as a short when it is not. I recommend that you take your multimeter and measure the resistance between CVDD and VSS on a known good board such as an EVM.  I believe that you will see the same thing.

    You are probably using a common, hand-held, portable multi-meter.  This operates on a battery that is usually 1.5V.  When you are measuring the resistance, this 1.5V supply is applied to the probe tips in series with the meter movement / sensor.  The current through the meter probes is then used to calculate the resistance between the probe tips.  The CVDD voltage on the C6678 is less than 1.5V (nominally 1.1V or less).  Therefore, when you try to measure the resistance from CVDD to VSS after the C6678 is attached, you are applying this battery voltage to the board.  Since the CVDD supply is less than the voltage supplied, the C6678 CVDD circuitry begins to power up.  This draws a lot of current from the multi-meter which makes it display a very low resistance. This low resistance is greater than a true short.  The resistance displayed will be much lower when the 2 probe tips are touching.

    I hope this helps.

    Tom

  • 0 in reply to Tom Johnson 16214
    TI__Mastermind 46460 points

    Charles,

    In reply to your original question, you will often get measurements below 10 ohms.  This is normal.  It is common to see it as low as 2 ohms.

    You mention devices running much too hot.  That is why they are failing.  You must measure the case temp using a thermal grease with the temp probe firmly attached to the lid of the C6678.  Also note that almost all boards that use the C6678 will need a heatsink and may need forced air.  To measure the case temp with a heatsink attached, you must drill a hole in the heatsink or cut a notch in it.  The temp probe must be attached directly to the lid of the chip.

    Also note that due to manufacturing variations, some chips will run hotter than others.  This is normal.  Do you have an AVS-compliant power supply?  What is the CVDD voltage on the units that you are testing.

    Tom

  • 0 in reply to Tom Johnson 16214
    Prodigy 80 points

    Hi TOM,

    Thank you for your reply.

    It`s very kind of you!