ADuM5402, state of logic output when power is not well supplied

Hi,

We are assuming what would be occured when power supply from primaly side is stopped by any reason during operation, for example, malfunction at internal or external circuitry, at ADuM5402.

In this case, how is the state of primary/secondary side of output pins supposed to be ?

 

Subject to page.24 of datasheet, POWER CONSIDERATIONS, we are assuming output become to be high-impedance, but it can allow to understood that it may be default low, or remain state just before power down.

Especially following last phrase are not clear to see...<The outputs on the secondary side hold the last state that they received from the primary side. Either the UVLO level is reached and the outputs are placed in their high impedance state, or the outputs detect a lack of activity from the primary side inputs and the outputs are set to their default low value before the secondary power reaches UVLO.>

If you provide another explain by focusing a case power supply suddenly stop during normal operation, it would be very much appreciated.

Best Regards,

  • 0
    •  Analog Employees 
    on Apr 25, 2013 12:15 AM

    Hello Usaghi

    Keep in mind that the two sides of any isolator can work somewhat independently.

    The secondary side normally gets its power from the isoPower transformer.  It regulates the voltage by sending a digital control signal to the primary.  This provides power to the digital channels on the secondary.  If power from the primary is interrupted, the output voltage will fall until it reaches the UVLO threshold (about 2.4V) below that the output buffers are off making the outputs look like a high impedance.

    One exception to that is if there is another source of power on the secondary side, either on the Viso line or at an input.  If Viso is being held up by large caps, or another power supply, then it can act like a power input and provide power to the secondary above the UVLO level, and the secondary data channels will remain active.  They will likely change to the default low state after a few uS since the are presumably not getting any data from the primary side.  A similar thing can happen if digital signals from other chips are present when the isoPower is not.  The digital signals will push current into Viso through the ESD protection diodes and again the secondary can be on in a default low output state.

    So it is not simple behavior if parasitic power is present.

    I hope this is clear.

    Best regards,

    MSCantrell

  • Hi MSCantrell,

    Thank you very much for your clarification. I think I can move one step forward to understand the part.

    I will let you know when I have additional query.

    Best Regards,

  • Hi MSCantrell,

    In order to have clear overview, I tried to make matrix of cases which is described in previous discussion, as follows.

    Please could you confirm them or remedy when it is not correct.

    Power supply
    on
    Primary side
    Power supply
    on
    Secondary side
    Primary side input: data at Vix
    output: Result at Vox
    Di
    rec
    tion
    Secondary side input: data at Vix
    output: Result at Vox
    No No input no data => output Z
    input data exist => output Z
    output Z <= input no data
    output Z <= input data exist
    No Yes input no data => output L
    input data exist => output L
    output Z <= input no data
    output Z <= input data exist
    Yes No input no data => output Z
    input data exist => output Z
    output L <= input no data
    output L <= input data exist

    Z : High-impedance ,

    L : logic Low

    Best Regards,

  • 0
    •  Analog Employees 
    on May 15, 2013 5:54 PM

    Hello Usaghi

    The table is largely correct, except where there is data present on a side that is supposed to be unpowered.  A digital input can provide enough current to allow the part to try to operate so some of the cases where data is present on an input and the power supply is off may try to send data across the barrier and you will get indeterminate operation.  That side of the part could be turning on and off in an uncontrolled way.  The only way to get completely determined behavior is to be sure that the data signals from other devices are off as well as the loacl power.

    on on output: Result at Vox rec output: Result at Vox Primary side Secondary side tion
    Power supply Power supply Primary side input: data at Vix Di Secondary side input: data at Vix
    No No input no data => output Z
    input data exist => output Indeterminate
    output Z <= input no data
    output Z <= input data exist
    No Yes input no data => output L
    input data exist => output Indeterminate
    output Z <= input no data
    output Z <= input data exist
    Assume isopower disabled to do this, Yes No input no data => output Z
    input data exist => output Z
    output L <= input no data
    output Indeterminate <= input data exist
    Z : High-impedance ,
    L : logic Low

    Best regards,

    MSCantrell

  • Hi MSCantrell,

    Appreciating for your always quick and clear answers.

    Best Regards,