I previously had a project on the Uno and Duemilanove. It required the use of the external 12V power supply to provide enough power to sufficiently energize some coils (otherwise I'd get lazy rotation.) Anyway, I recently ported this project to the Due board. The code went very nearly straight across and the shield was compatible since it takes 5V or 3.3V input and has no outputs back into the Arduino so no risk of exceeding the 3.3V The IC on the shield has a pin for additional 12V power supply to run it's internal circuits and boost the energy to the coils it's driving.
Here is the strange part. When I connect the same external 12V power supply to the Due board, the processor *seems* to be running normally but all motion at the coils immediately stops.
I have not yet pulled out a scope to see if the SPI pins are still working but that is the only difference (other than the 3.3V logic) between the Due and Uno that I can possibly imagine being a problem.
With the Due of course I had to reroute the pins for the SPI library which wasn't a problem...until I connected the external power supply. This is a real problem because in order to get the crisp clean response at the coils I need, that 12V supply must be connected.
Any ideas or similar observations?
Forget it. I was being a dope. The chip I'm trying to interface with was designed for CMOS 5V logic and has a minimum voltage spec on its pins of 4V. When powered by the pins themselves it can see the signaling. As soon as it as it's own 12V supply from Vin, game over with the 3.3V logic.
I'm having to integrate one of these:
TXB0108 to run the conversion. Sweet chip, does automatic direction detection. Breadboard prototyping unit available here:
In case anyone else finds themselves having to interface the Due with 5V modules and/or chips.