Here's my build of the Smartfriends Smartrapcore CoreXY printer. (http://www.thingiverse.com/thing:651117) Quillford, on github, put together a fork using the Simple. It was a pretty easy build. My biggest hurdle was the firmware Quillford compiled from lincomatic. I couldn't get my homing to work, so I went through several flashes before I finally realized that the z probe logic was in reverse. Once that was corrected, it was smooth sailing. I haven't yet tested the top speed. I hope to get to do that tomorrow.
@mooselake, I actually didn't research any other builds. I saw this on /r/3Dprinting, and I figured it was a quick and easy way to try out CoreXY.
I still haven't tested top speed during print moves. I have my travel moves set to 150mm and I have had any issues with that. I was very pleased that my 5mm calibration steps came out nicely with out any sort of calibration. Using this after using a different printer with a heated bed really makes me appreciate that heated bed. I did a largish flat print that curled something fierce. On the side where it didn't curl the print was really smooth.
@JonS, the benefit that stood out for me is speed.
It does have a low amount of moving mass, but as you go larger you end up with a very long belt, have to wonder how calibration holds up over time as the belts stretch. For an XY plotter it really doesn't matter if the 1st and 500th prints are different by fractions of a millimeter, but kind of important for 500 layers of 3D printing...
JonS wrote:I'm curious about the stretching of the belt during operation (not just over time). Is this a potential issue?
I don't think so. New belts might need a little break-in time and subsequent adjusting, but after that they work just fine. CoreXY is becoming the gold standard, especially with the bigger bots. I read it on the internet
CoreXY gives the weight savings benefits of the H-bot config without any "racking" (the gantry binding between the outer rods). The downside is that the belt configuration requires more complex mathematics than a normal cartesian bot. The Sli3DR configuration addresses both of these issues and uses wire instead of belts.
"Metal shavings are my glitter" blog.mechanizedmedic.org
Wire! Yes, a lot of old plotters used wire instead of belts. I've been thinking about a way to use belts instead of threaded rod in a mendel type configuration but wire might be better, way easier to design the pulley pathways with wire, hmmmm.....