Hi all,
I'm experiencing an issue that for me it's more likely a hardware issue, camera sensor is just not visible on the I2C bus as if it wasn't connected when the AM5728 EVM board is cold, so communication can't be done and the device can't get probed and registered successfully. I'm using TI's image and the output log is the following:
#) mt9t11x 4-003c: input-clock-freq: 32000000 #) mt9t11x 4-003c: pixel-clock-freq: 96000000 #) vip 48990000.vip: loading firmware vpdma-1b8.bin #) vip 48990000.vip: VPDMA firmware loaded #) mt9t11x 4-003c: Product ID error ffffff87
After heating the board externally with air at 140°F for around 30 minutes or more if no luck the camera starts being visible on the I2C bus and the device driver gets probed and registered successfully as shown in the following output log:
#) mt9t11x 4-003c: input-clock-freq: 32000000 #) mt9t11x 4-003c: pixel-clock-freq: 96000000 #) mt9t11x 4-003c: mt9t111 chip ID 2680 #) vip2-s0: Port A: Using subdev mt9t11x for capture #) vip2-s0: device registered as video1 #) mt9t11x 4-003c: mt9t11x sensor driver registered !!
I have seen the same behavior with different combinations of three AM5728 EVM boards and two AM57xx EVM Camera boards. It seems that some strange physical phenomenon is occurring on the I2C lines that makes communication successful only when those lines are heated.
Has anyone experienced something similar? I would like to get this issue solved.
Regards,
Carlos Gomez