Other Parts Discussed in Thread: TCA9509
Tool/software:
Hello,
We are having some problems with the TCA9617B lC when trying to read the HDMI EDID over the I2C bus. It does not work reliably across multiple displays.
Our design operates at 1.8V on the A side and 5V on the B side, with standard mode speed (100kHz).
Here is a scope shot of a correct operation:
In the case of a failed operation, the scope capture shows the irregularities. It seems that the target doesn't drive the signal low when it's expected (ACK). There is also some strange CLK behaviour (not always the case) just before the problem, as it can be seen in the image. The voltage level also looks lower than expected in this case (~4.4V instead of 5V. If no operation is executed the signals are correctly at the level of 5V). These scope captures were taken while testing with the HDMI dummy connector from Adafruit.
Increasing the pull-up on the B side to 4.7k improved the situation for some displays, but not for all, and according to the HDMI DDC specification the pull-up on the DDC side should be between 1.5k and 2k, so we would not want to increase the pull-up value.
We also tried adding an RC filter to the DDC SCL and SDA lines. But this did not improve our situation. We also tried different level shifters but saw the same failure behaviour.
As an additional information, the design with this IC does not work reliably on the tested displays like ASUS PB248Q, LG 32UD59-B or the HDMU dummy connector from Adafruit. However, our design works fine on the tested displays such as Samsung SyncMaster F2380, Dell P2419 or Dell SD2721DS.
Do you have any idea what might be causing the problem and how we can improve the situation?
Kind regards
Franz
