After a month and a half of reading, thinking, reading and designing, I'm ready to start on this journey.
It occured to me one day, that a lot of my 'play' in electronics, and mechanical things were working me towards some sort of robot. I have been designing and building things for years, and after a long break, was thinking of starting something biggish.
I stumbled across an engadget post about RepRap one day, which made me aware that rapid prototyped parts are not just for gently handling and looking at, but can be quite structural as well... (the RepRap's key feature is that much of the structural components are printed on other RepRaps) I looked at my pile of stuff, and decided, "I can do that"... So I started to read....
I decided I wanted to design my own cartesian bot, based on things I've read, rather than try to follow someone elses 'recipe'.
A month later and using all sorts of ideas from all sorts of people (if it looks like something you did, and posted about, I likely got my ideas from it), I came up with this initial design:
The bearing design was presented to me one day by Gilles on the RepRap IRC channel, and I really liked it... it was a pivotal moment for me as it caused the design, which was constantly changing around in my head to suddenly 'gell'...
I have, and like to use, a tig welder to make things out of aluminum, so I altered things to make use of it, and save some space.
In the end, I didn't like the idea of using MDF to build my machine out of. I expect to move this printer between my upstairs electronics room and my outside shop. The thought of lugging a big MDF box up and down my steep Netherlands stairs didn't appeal.. I have a decent metal working shop, with a chop-saw for cutting aluminum, belt sander, the welder, and various fabrication tools, unfortunately not including a milling machine, or lathe.
I liked the idea of using t-slot beams. They keep things looking clean, give me adjustability for lining things up as I assemble, and can be used for something else later if I want.
I came up with this:
It is based on FlexLink t-slot from RS Components, the XC size (44mm square)... It is a bit beefy, but I figure I may be wanting to push a palm router with this, and want it as stiff as possible. Plus I like the idea of using a double-wide vertical beam for fore/aft stiffness, and the smaller size wasn't available from RS in the double width. I also may decide to make it all larger if I get more space. By starting heavy, I will be able to keep all my connection hardware and just change some extrusion (and the drive system)...
I decided to use the FlexLink linear bearing for my Z-axis...
Bearings are mostly as per Gilles design...
I'm using 12mm bearing bar, and over constraining the bearings for added stiffness.. I think I'll be able to keep my tolerances close enough by drilling all the holes for the mounts in a stack.... We'll see in a couple weeks... =)
As you can see, it follows quite closely to the Mendel layout, but doesn't use RP parts.
I feel that I can scale it up to a larger size if needed at a very reasonable cost in time and money.
I also gave thought to replacing the belts (which are full 10mm thick and not cut down as per Mendel) with leadscrew drive for milling. I would have gone directly with leadscrews right away, but it seems that no one is doing a great job of printing with them at this point, and the added speed of belts seemed prudent. With the t-slot I can change things around relatively easily to suit.
I'm using Nema 23 motors for strength, Pololu 1/16 stepper drivers for added precision, Arduino Mega mother board. It all seems a lot heavier duty than a Mendel, but hopefully it will give me the added strength to do 'beefier' things with it.
I have the motors and electronics already, 350 euros worth of t-slot hardware and 2 to 3 hundred more in other bits are sitting on my RS account ready to be ordered in a couple days...
I'll report back on how things go.