But has anyone tried just swapping out the stock Mega with a DUE using the Maslow board? Is it 3v tolerent or does it require 5v?
From other research…
"The Due can do a lot of things much faster. It has more memory and far more processing power. It has more flexible and more powerful built-in peripherals.
It can’t easily interface with 5V devices though.
The Mega2560 has less memory and far less processing power, but interfacing with legacy 5V devices is far easier. It’s also more forgiving to the novice user than the more complex SAM3X chip."
I looked into it a long while back. I remember that the lack of EEPROM was one complication. The current location and settings are stored locally. If you’re checking into the DUE, the hardware needs to provide 3 hardware PWM pins and 6 pins with interrupt capability.
Unfortunately, no simple solution has been found to date.
In your foam cutting application, the drag of the bit in the stock shouldn’t be a factor, and you’ll probably be cutting all the way through in one pass around the profile. That alone will be the most significant speedup. Not sure how you’ll handle things like tabs for holding the pieces in place as they’re cut, but that shouldn’t be hard to solve.
Speeds: Currently the max feed is limited in the software to 800mm/min (~31 inch/min)
Reason 1: We don’t have acceleration / deceleration planing in the software yet. Imagine a sudden change in direction and the sled ‘swinging’, spoiling the cut.
Reason 2: We have tested 1000mm/min (~39 inch/min) and found problems that at the top of the sheet, where the motors are under the most load, one motor pulling, can not keep up the speed of the other motor feeding out the chain. The result on my Maslow looked like this (top should be a straight line):
Unfortunately, no simple solution has been found to date.