There seems to be a voltage leak from HDMI connectors through ADV7623 and onto the 3v3 rail (TVDD_3V3).
ADV7623 evaluation board
NO power source connected.
HDMI source connected to HDMI in 1. (Source is Atlona AT-HD600)
Now i measure voltage from Testpoint TVDD_3V3 to GND = 3.3V
LEDs can be seen to power up faintly on the evaluation board.
(If i connect my source to HDMI out on evaluation board, all LEDs light up brightly).
Other iC's (uC, flash etc.) running of the 3v3 rail can (dependening on component tolerances) become powered up fully or partially by only connecting HDMI source to electronics designed as the recommended reference design for ADV7623 (and that used by evaluation board).
This can result in uC powering up in a undefined state when the actual powersource is connected, thus rendering the device not working until HDMI is unglugged and power is reset.
What are your recommendations to work around this?