Hi, I'm currently using the ADV7611BSWZ on a new board and I'm having some issues with it when we do ESD testing on the unit that its in.
When we do contact discharge (as low as 3KV) to our unit, we sometimes see the HDMI video from the ADV7611 turn into what looks like TV static instead of the video we were expecting. The audio also turns into static noise. The IC continues to work though and we can query its registers over I2C. The HDMI input will recover if we continue to zap the unit or if we re-connect the HDMI plug.
The ADV7611 is currently ESD protected using a HDMI ESD protection IC from TI. The register settings on the ADV7611 are set almost exactly like in the ADV7611 eval board script for 1080p@60Hz (VIC 16 31 32). Just the HDCP disable is ignored and the input clock is 27MHz instead of 28.6363MHz.
For a software patch, we're querying the registers of the ADV7611 and resetting the chip if a value is not what we'd expect. The register that we're looking at to determine if something is wrong is I/O map register 0x65, which should be 0x8E when normally running. Specifically, we're looking at the MSB of 0x65 (the channel status valid bit) to determine a fault. During a fault condition, we seen register 0x65 assume any of the follow values: 4E/4F/5E/5F/6E/6F/7E/7F
We've also looked at I/O map register 0x6A and HDMI map registers 0x04 and 0x05, but they don't always give consistently different values from the norm during a fault.
Given the information above, are there better/other registers that we can be looking at so that software can determine if there is a fault from ESD? Are there any registers that we can try adjusting to see if it will improve the device's immunity to ESD events?
Also, to what voltage is this device ESD rated to? I see a generic ESD statement in the datasheets and user guide, but no number is given.
Edit: some additional info... simply manually toggling the HPA pin also restores the HDMI input, without having to reset the ADV7611.