• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 1548 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.

TPS65094: PMIC resets after glitch on BUCK5 output

user4497266
Intellectual 300 points
Part Number: TPS65094


Hi
I'm using your TPS650940A0RSK for an Intel Apollo Lake design. We have produced 6 boards and on one of them I have a CPU reboot issue. It seems as a glitch on the 1.24 V output (BUCK5) causes the PMIC to reset.

Most often it resets within a few seconds after power is applied (sometimes the BIOS gets loaded and sometimes not. After I have powered on/off a few times it seems to reset less often. Sometimes it resets after a few minutes and sometimes it might never reset. To me this indicates a temperature dependence. All tests I currently do is at room temperature.

I read the e2e.ti.com/.../2270628 thread so I tried to increase the output capacitance (from 30 uF to 75 uF) but this did not help. I also tried to increase the input capacitance (from 5 uF to 10 uF) but this did not help either. The capacitors are X5R MLCC and the capacitance values I mention above is the actual capacitance at the DC bias point.

I have not yet tried to swap PMIC's between two boards. I want to try and find the problem elsewhere first.

Following screenshot shows BUCK5 output voltage and RSMRSTB:

Following screenshot shows BUCK5 output voltage and BUCK5 switch node. I have zoomed in on the abnormally high frequency "ringing":

Following image shows the schematic and layout:

I would be very happy if you could give me ideas on what to do to find the problem.

Regards, Jonas

  • 0
    TI__Expert 4155 points
    Jonas,

    This is something that I have seen before. This can be avoided by putting the BUCK converters (BUCKs 3, 4, and 5) into forced PWM mode through I2C writes to the BUCK4CTRL, BUCK5CTRL, and VR_CTRL1 registers (register addresses 25h, 26h, and 9Ch).

    We are also going to release a new spin of this device at the end of February, the TPS650945, which is the same as the TPS650940, but with BUCKs 3, 4, and 5 set to forced PWM mode, which completely avoids this erratic switching.
  • 0 in reply to Nick Jakse49
    Intellectual 300 points
    Hi Jack and thanks for your prompt reply.

    We will try to test this but we need to figure out how it is done in Linux first. We have added access permissions to the PMIC I2C registers in BIOS. I will update this forum thread as soon as we have tested it.
  • 0 in reply to user4497266
    Intellectual 300 points
    We have now tested to force PWM mode as soon as Linux boots. It seems to help, the PMIC has not yet rebooted while in Linux.

    However, it won't fix the issue fully since the PMIC will reboot many times before it gets far enough to start booting Linux. For a full fix we need the TPS650945.