Post Go back to editing

Possible issue with Adum3482

Hi,

We designed several Adum3482 into a product and found to our surprise that the VDDL1 pin can power our entire board -- with current flowing out of pin 1.

Schematic fragment here.

All IO is 3.3V, and I assumed that the best way of using this was to run VDD from 5V, and the logic level translators from 3.3V.  The problem is that I have approx 180mA flowing back out of pin 1, happening on multiple devices.   Raising pin 1  to 5V corrects this -- obviously.  

I did wonder if CTRL1 was part of the logic translator "domain" so I removed the 5V from there and drove it with 3.3V, no change.

There is a drop of approx 1.1V when supplying 180mA, so I suspect there is a diode involved.  The device does get hot when supplying current.

Has anyone ever tried these devices with VDDL lower than VDD ?

This is an urgent request, volumes in 6 digits involved.

  • Hello,

    The VDD1 and VDDL1 can work independently in any combination. 

    Everything about your issue is consistent except your experiment with the CTRL1 line, that should have stopped the current back flow.

    The CTRL line I/O is supplied by the VDDL1 supply, so its input should be referenced to the same voltage as VDDL1.  What you are seeing is that the CTRL1 line is higher than than its supply by 1.7V which forward biases the ESD protection diode, and tried to drag your 3.3V supply up through that diode, which accounts for the 1.1V drop that you observe.

    Please connect CTRL1 to the 3.3V supply and the excess current should stop. Using a pulllup resistor in the path would not hurt either..  You can also simply run all of the power supplies at 3.3V, there is no need to have a 5V supply on VDD1.  Pleas verify that all of the signals going into the I/O are 3.3V signals, you may have another sneak path.

    If you still have problems, please contact me again.

    Best regards,

    MSCantrell

  • Thanks,

    I'll check the CTRL1 line, may have had a bridge.

    Regards,

    Gavin.