• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 98 subscribers
  • Views 356 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.

Anyone have the problem like me in TPS65001

Kevin_Chen
Intellectual 405 points
Other Parts Discussed in Thread: TPS65001

Hi there，I made a custom board which based on TI8168EVM, and using the Power management  chip TPS65001 for EVM_1.8V_A, EVM_1.8V_D and EVM_0.9V, the chip started correctly except EVM_1.8_A, because I put a wrong value divider resistance, the top one suppose to be a 261K ohm( the bottom one is 100K ohm ), I put a 22K ohm resistor instead, which only give me 0.6V on that LDO channel, so I changed it to a 261K ohm resistor, and everything seems fine, but after running couple minutes, the  chip was really hot, and the internal load between pin 14(output pin of LDO 1.8V_A) and ground dropped to around 8 ohm, then the output dropped back to 0.5V, but the weird thing was all the reset two power rails are still right!

Is there anyone know about this, thank you in advance.    

  • 0
    TI__Mastermind 43875 points

    Make sure that all the high current paths have sufficient copper. The original EVM had extremely small traces for some signals which should have been 50 times bigger, in particular the switching FETs did not have nearly enough connecting copper to the main supply and ground.

    Does the processor start up then fail or does it never start up now?

    What exactly do you mean by the internal load dropped to 8 ohms? Do you mean you are trying to measure the DC resistance of the processor power supply inputs? This is not really a valid measurement.

    Can you look at the power supply with an oscilloscope to make sure that the PMIC is not switching between on and off quickly, indicating that it might be detecting an over current condition.

    Are you able to measure the current taken by each of the power supplies?

    Do you have another board? If so, how does it behave? Is it possible you have a PCB error?

    When the system seemed to be running fine, what was it doing? Was it an intensive application or simply booting Linux?

    BR,

    Steve

  • 0 in reply to Steve Clynes
    Intellectual 405 points

    Hi Steve,

    we have couple boards can boot up, and can be connected via JTAG, the problem is lots of custom board failed after connecting the power, before applied the power, the Dc resistance of 1.8v_A and 1.8v_D are couple K ohm level, then after that, it drop to few ohm, so it that possible that my desgin cause the problem like breaking down the internel transisitor?

    Depends on the size of my constom board, I dont have so large room for tps65001(around 12mm*12mm), maybe the copper is not sufficient to carry the high current path, and I did not put shunt resistor or ferrite bead in the path, so it might be not so easy to test the current of that path, but the EVM board did not draw large current on that power regulator right?

    I uploaded the pcb of the power regulator, 10 mils traces are used between the chip's pad and large copper.

  • 0 in reply to Kevin_Chen
    TI__Mastermind 43875 points

    DC resistance is meaningless and cannot be measured when there is power anywhere in the system.

    All I can suggest is to try and measure the output current from each supply to see if there is once supply which is drawing excessive current. this might point you towards some other fault in the design. You can also try using an external power supply on the 1.8V. Removing the switcher inductor should allow you to get access to the 1.8V rail and isolate the switcher.

    Check the voltage levels on all supplies to make sure that they are correct. Try to measure as close as possible to the processor. Check the supplies with an oscilloscope too to make sure there isn't any very short term noise or drop in voltage.

    I don't think the 1.8V should draw a lot of current, likely less than 1A I would think.

    What type of memory are you using? If using DDR3 then make sure that there isn't some short between 1.5V and 1.8V somewhere.

    Bottom line... measure as many supply currents and voltages as possible. My guess is you have a short somewhere.

    BR,

    Steve

  • 0 in reply to Steve Clynes
    Intellectual 405 points

    Hi Steve,

    So the copper and trace size I used was correct?  Yes, I tested the DC resistence of each power rail on board in order to check if short somewhere or not before applying it to the power, the wired thing was couple of broads short only after apply the power, when we remove the processor on one of boards, the short problem is gone again, we are using ddr3, and 1.5V and 1.8V are isolated well and have no problems so far.

    And one of other thing might cause the problem is the via underneath the processor, the manufacture was not mask via very well, so if the processor not position well, it was very easy to cause short problem~