Post Go back to editing

ADV7182A very low current from all power rails

Category: Hardware
Product Number: ADV7182A


I need your suggestions to very strange case with ADV7182A.

In my design all power peripherals came from 1.8V (with ferrite beads) - the power sequence is simplified power-up sequence - all power rails and PWRDN and RST is 1.8V together.

I have 3 ADV7182A - two of them on the same I2C bus with different ALSB - the last one is in different I2C bus

The ADV7182 not responding to I2C - we get NACK - address 0x20/21/22/23 or 0x40/41/42/43 - other ICs on this I2C bus are responding with ACK

The current from 1.8V is very low - around 0.5mA - seems dead or not fake 

  • My first thought is to mention that ADI uses 8-bit I2C addressing instead of 7-bit addressing used by other devices on the I2C bus and the I2C driver expects 7-bit addressing.  So you'd have to right shift our script address by one before sending it to the I2C driver.  Give that a try.

    I'd expect currents to be low until you start pushing video through the part.  Current consumption is directly related to the clocking rate.

    I see no obvious problems with the schematic.

  • Thanks for your quick reply.

    Sorry, but I don't understand your suggestion.

  • Most I2C devices use 7-bit addressing and I2C software is expecting 7-bit addresses.  ADI video device scripts use 8-bit addressing.  Try taking the 8-bit address, shift it right by one (turns it into a 7-bit address) and use that as input to your software.  

    8-bit >>1 => 7-bit => i2C software which left shifts by one and appends the read/write bit making the 8 bits that show up on the bus.  Basically using 8 bit  addressing means all odd addresses are reads and all even addressing are writes.  A bit confusing but is works.

    0x20 => 0x10, so you send 0x10 to the I2C software.  The software will convert it back to an 8-bit packet with the read/write bit added.

  • Not answer to those address:

    0x10 to 0x14

    0x20 to 0x24

    0x40 to 0x44 

    Any additional suggestions ?

  • 1) schematic looks good.

    2) 8-bit addressing issue has been addressed

    3) This assumes RESET and PWRDWN are both high

    4) assumes EPad is really tied to ground.

    5) other I2C devices on the same 1.8V I2C works

    ? Exactly how are you implementing the 1.8V I2C bus?  

    Beyond this I am running out of items to check.