• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 277 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.

TPS2120: MUX not passing IN1 to OUT after a power cycle

Jonathan Lui
Prodigy 40 points
Part Number: TPS2120
Other Parts Discussed in Thread: TPS2121

Hi,

We are using TPS2120 (2:1 Power MUX) with two 5V power sources at IN1 and IN2. PR1 pin is shorted to IN1 and we are not using OV1/OV2 (both grounded).

When the system is first turned on, IN1 is live with 5.26V @ 2A and IN2 is turned off. The OUT voltage is fine as it shows 5.11V @ 2A. After the system is running for a few minutes, if IN1 is turned OFF, then back ON, the MUX no longer passes the 5V from IN1. OUT is showing 0V. However, if we immediately turn on IN2 (5V), the MUX "wakes up" and passes IN1 to OUT again. Is there a reason for this behavior? We didn't expect to need IN2 to be turned on to wake up the MUX after a power cycle at IN1. Could this be a BGA assembly or thermal issue?

Thanks,
Jonathan 

  • 0
    TI__Mastermind 32650 points

    Hi Jonathan,

    If PR1 is directly connected to VIN1, this means that VIN1 is prioritized as soon as it becomes valid, i.e, at 2.8V.

    When you turn off VIN1, the device will prioritize it until it hits UVLO. If VIN1 is sent high again, the device recognizes VIN1 and prioritizes it. Since the device selects a source that is coming from an invalid to a valid state, the device undergoes a settling time followed by a soft start time. Is this what you are seeing? Could you post waveforms that describe VIN1, VIN2, VOUT and IOUT? WE would also benefit from a schematic that shows how TPS2121 is set up.

    Thanks,

    Shreyas

  • 0 in reply to Shreyas Dmello
    Prodigy 40 points

    Hi Shreyas,

    Thanks for the reply. Schematic to the TPS2120 is attached here. I don't think it's a settling time/SS issue (Css=1uF). After a power cycle on IN1, I waited at least 15 minutes for OUT to get IN1 again, but OUT is still 0V despite IN1 being valid again (~5V). It seems the MUX only "wakes up" if I turn on IN2, then OUT will get the power from IN1 again (confirmed by dual bench supply). I can turn off IN2 after it "wakes up" the MUX and OUT will still be fine (getting sourced by IN1). Can you think of situations where the MUX gets "stuck" when IN2 is off even though IN1 is back to a valid voltage?

    I don't have a scope waveform captured yet, but I can tell you that IN1 was turned back on after a second that it was off. IN2 is off during the whole time (unless I turn it on manually to wake up the MUX).

    Side notes:
    -SEL (net MUX_MODE is always 0).
    -The big red cross is only designating a DNP (R31), please ignore it.
    -Both SW_PWR (IN1) and PROTECTED_5V (IN2) are 5V sources although IN2 is off in our test conditions.
    -IN1 is intended to have priority.

    Thanks,
    Jonathan 

  • 0 in reply to Jonathan Lui
    TI__Mastermind 32650 points

    Hi Jonathan,

    I think we need a waveform as requested to see exactly what is happening.

    Have you been able to reproduce this event on multiple TPS2120 devices?

    Have you tested this on an TPS2120 EVM?

    The only way I could see this occurring is if for some reason, the current limit is being triggered and the device cannot cannot start back up.

    Thanks,

    Shreyas