EDID replicator only responds to the selected input. Is there a way to cause the EDID replicator to respond to all DDC input requests, whether or not that input is selected?
The ADV3002 EDID replicator should respond on all inputs regardless of the selected input. Is +5V present on the unselected inputs?
We now have a test setup to work with.
It appears that the EDID requests from any input are only responded to IF the channel is selected.
Also, the response comes from the output channel (IE the monitor), not from the EDID Replicator.
There is 5volts from each of the 4 inputs, and an on-board 5V supply. (AMUXVCC).
The on-board EDID EEPROM is being read on power up, verified with a scope.
The pin #63, EDID_ENABLE is high.
We expect the EDID replicator to respond to all inputs, and we do not see it responding to any input?
How can we determine if the EDID Replicator is working?
You can read out the contents of the EDID replicator SRAM by reading from i2c device address 0xA0 from the ADV3002 I2C interface (I2C_SCL and I2C_SDA pins). Have you modifed the vendor specific data block (VSDB) as described the "Source Physical Address Assignament" section of the ADV3002 datasheet? Can you confirm the state of HPD_[A,B,C,D] after the EDID EEPROM load is complete?
If the VSDB tag is missing, the EDID EEPROM load as shown in Figure 25 will not complete and HPD pins will be low. If the EDID EEPROM load is complete all HPD pins should be high.
More importantly, I'm concerned that the ADV3002 is not applicable in your application. In your application is the ADV3002 output connected to and HDMI output connector that is then connected to a HDMI display with it's own EDID? The EDID replication feature was intended for applications where the ADV3002 is integrated in an HDMI sink or as a port expander in front of an HDMI Rx device where there is no other EDID on the DDC bus.
Hi Dave, Thanks again, as this is a challenge.
Tried to read out the SRAM via 0xA0, but will only read one byte, (returns 0xff) and then I2C does not ACK the next read. Still working that issue. We did not setup the VSDB because there is no CEC available in the system...
Is there a generic 'NO CEC' somewhere? We have the CEC buffer off.(Register 04).
At this time all of the HPD are high, and they do pulse during startup 10mS.
Our application does use the system as a sink. A few notes, this is a DVI (not HDMI), no CEC, and we have been trying different ideas to determine why the DDC does not respond. In the final configuration, we would like to use the EDID replicator.
I think that's the issue. The ADV3002 needs to find the VSDB (specifically the vendor specific tag code) in order to complete and make the EDID accessible to the HDMI input ports.
Try using the example EDID attached.
Let me know if this works.
Also, once the ADV3002 EDID is working, you are going to have to make sure there is not a separate EDID on the ADV3002 output, otherwise there will be potential DDC bus conflicts.
I have a problem with the ADV3002 does not respond to physical address 0xA0. The I2C link never gets ACKed. The registers at 0x90 work fine. Thanks, Bill
Let me try to reproduce your issue in my lab. ADV3002 should respond to address 0xA0.
I am able to read from address 0xA0 on an ADV3002 evaluation board in the lab. The ADV3002 responds with 0xFF if the SRAM is not initially loaded, but it still ACKs address 0xA0. With EDID_ENABLE=0, can you force a device reset with the external RESET pin and then initiate an i2c read from address 0xA0?
Can you tell me the part number of the EDID EEPROM on your board?
Retrieving data ...