12V 1500mA should be fine (if it's really 12V, some cheaper ones may do interesting things, though usually to the lower direction, not the higher one).
I'm powering my board using Galileo Gen2's 12V 1.5A power supply for months without any trouble, so I'd rather suspect the LCD, especially given the fact it's only the I2C part that got impacted, in case of a faulty power supply it would first fry power distribution/regulation components I'd say.
That's was suspicion too.
However, the very same LCD was working fine with the Edison when powered through USB.
The problem now is that I am afraid of testing this setup again on another Edison... don't want to fry another I2C...
Any hints on how to debug this ?
When it was powered through the USB the board couldn't draw more than ~500mA, so the LCD problem might've not been surfacing and when it's got "enough" current, it fried itself (and the I2C outputs).
That's what I would suspect and other than that I'm not sure I could help any further, because any troubleshooting would require getting down with with some scope or logic analyzer and trying to understand what exactly got burned out and that's (a) quite hard to do via the forum and (b) I honestly just don't have such experience.
If anyone else could help - please chime in.
Thanks for the reply @AlexT_Intel! I understand it is hard to debug this from the forum
One question though: shouldn't the Edison have prevented this from happening ?
Well, that depends on what exactly has happened, I'd say. There's some reasonable amount of protection put onto those I/O lines I'm pretty sure, but at the same time there are certain limits those lines can handle, described in the docs and if the connected device gone wild and exceeded those, well, that's what going to happen.