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.

UCD9244: over-current fault

Part Number: UCD9244
Other Parts Discussed in Thread: UCD74120, UCD7242, TMS320C6678,

Team, 

My customer is seeing this intermittent issue:

I have a board with two TMS320C6678 DSPs.  They are being powered by a UCD9244 along with UCD74120 for CVDD and a UCD7242 to supply the 1.0V and 1.5V.

When the board is up and running I observe that there is a current draw of about 1.6A on the 1.5V supply.

What I am seeing is that at intermittent times the board will power off with the only failure indication being that the FLT line for the 1.5V rail has tripped.  I have done some investigating and determined that the failure is not due to temperature, but it appears to be from an over-current fault.  I connected a scope to the board monitoring FLT, IMON, and PWM of the chip.  What I am seeing is that the PWM coming from the UCD9244 seems to change drastically causing the IMON to spike and a fault to be thrown.

I am seeing this on multiple boards, and it is very intermittent.  I have seen boards run anywhere from 1 minute to over 24 hrs.  Any help would be appreciated.

Thanks

Viktorija

  • It could be a number of things but I would start by looking at the compensation on the 1.5V rail, this should be a simple configuration fix if that's the issue.
    Also, is it possible the feedback on this rail runs by any of the switch nodes of the power stages or any other noisy circuits on the board.
    Can you forward the schematic for the power stage portion of your design and the configuration file for the UCD9244?
  • Thank you. Waiting on customer's inputs.

    Viktorija

  • Brad, schematics were sent to your email. Please confirm you have received it.
    Thanks
    Viktorija
  • Got it, Thanks, I will look into today.

  • Couple questions based on first pass over documentation.
    Should the 4th rail (1.5V) be running at 1396kHz, the other rails are set to 751kHz.
    The 100uF generic output capacitor chosen from the design library of the Fusion GUI has a 100 mohm ESR, I suspect that this probably should be much lower. Is it a ceramic capacitor?
    Can you just verify the output inductor listed is the HM72A-06R68LFTR, no issues with that just want to make sure I'm working off of the correct components.
  • Brad,

    from customer:

    1.  I originally had the 4th rail at 751KHz as stated in a previous email.  The issue occurs when I have the switching frequency set to 751KHz like the other rails.  As an experiment last week I upped the switching frequency to 1396k.  Once I made the switch I have not had a board power off.  That is why I suppled two xml files.
    2. The bulk capacitors attached directly at the UCD7242 are as follows:
      1. 1500UF Kemet T530X158M2R5ATE005
      2. 100UF AVX TLCR107M004XTA  Actual ESR of 5ohms (Tant)
      3. 47UF Kemet C0805C476M9PACTU (MLCC)
    3. Yes that is the exact part number for the Inductor.

    There could be some value discrepancies entered in the schematic for the GUI.  I would enter values, quit and return to the GUI and values would be removed.  I was having difficulties with the GUI, and had to try things with different versions.  The current version seems to have a lot of bugs.

  • Brad,
    Have you had a chance to look into this?
    Thanks
    Viktorija
  • Viktorija,

    I bumped their schematic up against the project file and it did not seem to far off.  I did add some of the additional capacitance from the processor pages into the rail#4 compensation.  I've attached two files they can try, file with TI appended closes out around 25K, the other file with TI_10K appended closes out around 10KHz.  I included this if they are still seeing significant jitter on the PWM with the other file.  These both run at 751KHz.

    Question: Are they seeing the same level of jitter on all the rails?  Do they see any ripple/noise on the BPCAP pin, if they do they might look at adding a 2.2uF to that pin.  Is the 3V3 supplying the controller clean, I see its running from an LDO in the schematic so I wouldn't suspect that but good to check.

    Brad

    VoIP-III_081518_TI.xml

    VoIP-III_081518_TI_10K.xml