TCA39306: Which Side Should Be Powered First, VREF1 or VREF2?

Part Number: TCA39306
Other Parts Discussed in Thread: SN74LVC2G66

I would like to use the TCA39306 as an I²C signal switch.

Both VCC1 and VCC2 are 3.3 V, but VCC1 powers up earlier and VCC2 powers up later. The delay between them can range from approximately 1 second to several seconds.

In this case, which side should the 3.3 V supply that powers up first be connected to, VREF1 or VREF2?

Are there any concerns regarding device operation, leakage current, or reliability during the power-up sequence when one side is powered significantly earlier than the other?

  • Hi User1008366,

    It depends on how the EN pin is being connected and power sequenced. The EN connects to the gate voltage of the passFET connecting SCL1 to SCL2 and SDA1 to SDA2. 

    If VCC1 then VCC2 powers up while EN = 0V, there should be minimal leakage from side 1 to side 2 and side 2 to side 1. 

    Same if VCC2 then VCC1 powers up as long as EN = 0V. 

    In the switch mode, when VCC1 = VCC2, the EN can be controlled externally with a GPIO to ensure high-z state when needed. 

    Regards,

    Tyler

  • Hi Tyler,
    Thank you for your reply.
    I understand that applying power to the VCC1 side first helps reduce the leakage current.
    I also noticed in another thread that, when the purpose is simply to isolate signals operating at the same voltage level, devices such as the SN74LVC2G66 were mentioned as being a lower-cost and simpler solution.
    Would the same recommendation apply in this case as well, where the signals operate at the same voltage level but belong to different power domains?
    Best regards,
  • Hi User1008366,

    Are you talking about power sequence differences between TCA39306 and SN74LVC2G66 device? 

    Regards

    Tyler