LTC4041 IGATE current

What is the IGATE source and sink current for both "ON" and "OFF"

  • +1
    •  Analog Employees 
    on Jul 14, 2020 10:29 PM

    Hello,

    We don’t have datasheet specs for both figures. By design, IGATE can source about 4uA and has a pull-down resistance of ~200 ohms in backup mode.

    However, the leakage on the IGATE pin must to be minimized to make sure the pass transistors are fully on (see pg 19 & 22 of the datasheet).

    Regards,

    Zack

  • Hi Zack,

    A bit of follow-up question:

    In the datasheet p. 22 it is mentioned that "the trace can be guarded on the PC board by surrounding it with VSYS connected metal."

    However we noticed that in the demo design DC2642A, the trace is instead surrounded by GND (highlighted red here), which would induce more leakage current.

    What is the right way to route this trace?

    How common is it to see any problems attributed to different routing strategy of this trace, and what kind of problems are they?

  • 0
    •  Analog Employees 
    on Mar 11, 2021 8:31 PM in reply to tuwuhs

    You are correct that the routing of this trace could be improved on the demo board. I do suggest you surround it with VSYS if you can.

    We have a pretty standard layout practices with our demo boards, so we ended up going with those here. That said, the IGATE trace does work well on this layout.

    The IGATE signal is, by design, very weak. Long or noisy routing can interfere with it. Most commonly, problems manifest in the form of IGATE failing to turn on the FET, sometimes repeatedly.

    Regards,
    Zack

  • Thanks for your explanation!

    So we are currently troubleshooting a design where the system voltage output seems to dip for a very brief moment causing the circuits downstream to reset. We've narrowed down the issue to around this FET driven by the LTC4041. The problem happens only occasionally though, and isn't easily reproducible

    We suspect the problem comes from noise in IGATE somehow causing the FET to turn off briefly (instead of failing to turn on in the first place), but could not find a good way to verify it.

    Is this something that you ever see happening?

    I can start a new question thread if you think there's anything worth discussing in more detail.

    Thank you again!