• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 63 subscribers
  • Views 912 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.

TPS2052B Glitch observed on hot plug

Neal Wilding
Prodigy 60 points
Other Parts Discussed in Thread: TPS2052B

We are developping a design which uses the TPS2052B to switch power to a couple downstream USB devices. In the low power state of the design, power is removed from the switch Vcc pin, but the enable lines are held active. Occaisionally when reapplying power to the switch, we see a near immediate overcurrent condition on the /OCx open-drain output(s). This occurs before the power supply has reached its full 5V (the glitch commences at ~ 900 mV), and the 5V supply rise time is ~ 50 us.

We consider this to be a glitch, as we expect the switch to have undervoltage lockout at 900mV. We also expect there to be a 10 ms delay between the first detection of an overcurrent event and the output on the /OCx pin, but we're observing no delay in the millisecond range between power up and the over current event. The only explination that we believe falls in line with the data sheet is if there is an over-temperature condition (in which case the /OCx pin is activated immediately), but we would be suprised if the switch is entering into its over-temperature state before having powered up beyond the undervoltage lockout. Is this a known issue, or may it be specific to our application?

Best regards,

Neal

  • 0
    TI__Mastermind 42795 points

    Hi Neal,

    Please provide detailed O-scope waveforms showing the details of the event (IN, ENb, OUT, and OCb and IN current if possible), and also the schematic including all pertinent surrounding circuitry.

  • 0 in reply to Eric Wright
    Prodigy 60 points

    Hi Eric,

    Please see the capture below, with the traces as follows: 1) IN (5V supply), 2) Current into the "IN supply" of the PCB, 3) Overcurrent Sense 2 (Active Low), 4) Out 2.

     

    And the following capture where trace 2 is no longer the current, but is the Enable 2 signal (active high):

     

    Although I cannot provide the complete schematic, I have attached the following sketch to help show the circuitry.

    When capturing these plots, I have removed any downstream devices and loads, I have also removed any downstream bulk capacitance. The current captured in the first plot is not the current into the device, but is the current for the 5V supply into the board, and hence it represents a few additional loads (mostly capacitive) as well as the TPS2052B and downstream.

    Things of note in the screen captures, especially the top one:
    The /OCx output trips before the inrush is very high, despite the inrush only occuring for a very short period of time.
    The /OCx output trips before the output voltage begins to increase, and the output voltage continues to increase for a period of time despite the overcurrent tripping.
    After ~2ms, the HUB IC disables the enable input of the TPS2052B, and the downstream output voltage begins to fall.

    Best regards,
    Neal Wilding

  • 0 in reply to Neal Wilding
    TI__Mastermind 42795 points

    Neil, thanks for providing the comprehensive detail on the problem. At this point, I don't have an explanation of what you are observing. I will duplicate your setup and try to re-create your issue.

  • 0 in reply to Eric Wright
    TI__Genius 9415 points

    I discussed the TPS2052b problem with Eric. I had a TPS2052b EVM and was able to duplicate what you are seeing on OC/. It seems that there is the issue when the enable is ACTIVE before or during power up. I could avoid the OC/ glitch when holding the ENABLE false until the power was stable.

  • 0 in reply to Bob Kando
    Prodigy 60 points

    Hi Bob,

    Thank you for validating the issue and following up. This issue seems to contradict the datasheet which states that the device "is designed to eliminate false overcurrent reporting", with one of the device main features being listed as "No OC\ glitch during power up". Having said that, I believe this is a mature product, and I would assume no change would be considered to fix this issue?

    Best regards,
    Neal Wilding

  • 0 in reply to Neal Wilding
    TI__Mastermind 42795 points

    Neil, which package version of TPS2052B are you currently using (SOIC, MSOP8, or SON)? Would you be willing to increase the switch operating current (trip current)?

  • 0 in reply to Eric Wright
    Prodigy 60 points

    Our design uses the TPS2052BDRB (SON package). We are able to slightly increase the trip current, however we would be hesitant to raise it too much higher. This is a USB application, and the power from each switch is supplying an individual downstream port where the expected draw is 500 mA, the additional overhead it to allow for transients.

    Best regards,
    Neal

  • 0 in reply to Neal Wilding
    TI__Mastermind 42795 points

    Thanks Neal,

    We will be in contact offline to discuss the options for you. Thank you for your patience.