Fior some time now I have been toying with the idea of building a 3D printer. I looked into the various versions of the Rep Rap extrusion printers and while they are pretty cool, they just didn't meet my resolution needs for jewelry and highly detailed model parts. A friend told me about the B9 Creator which he owns and upon checking it out I was blown away by the resolution of the parts as well as Mike's design and build quality of the printer itself. I seriously thought about just purchasing a kit (the price is certainly reasonable) but in the end decided to build my own DIY version based on Mike's open source files. One reason was the cost savings (the total bill of materials will run around $1,500 to $1,700 of which $750 is for the projector) but also because I just love to build machines and have a full machine shop in which to do the fabrication.
The first step was to take the Parasolid files for all the parts and import them into SolidWorks, then create an assembly of the parts which allowed me to very closely inspect (virtually at least) how the machine is put together and functions. Again, my hat's off to Mike for working out a very clever design!. Here are a couple of images of the SolidWorks Model:
Then I basically took all the key components and designed a new housing around them. Some things, like the Vat and X Axis mechanism, will remain largely unchanged from Mike's design, while other systems, like the Z Axis and Projector mount/adjustment will get some fairly major revisions. This isn't a criticism of existing design, I just figured that since I had the chance, and based on some suggestions for mods that I've read in the forum, I'd take a shot at some modifications. If the work, great, if not, not big lose!
The overall framework for the machine is composed of horizontal 8020 aluminum extrusions sandwiched between two CNC precision milled MDF sides. The precision rods for the Z Axis and the projector up down adustment are held by turned collars attached to the 8020 pieces. This is a technology that has been used by John over at Microcarve.com to create his desktop CNC routers and it works extremely well, is very precise and cheap.
Here are some views of the preliminary SolidWorks model for the machine. Still a lot to work to out and add. In the end the top will be enclosed by panels with amber plexiglas for dust and light protection.
The slide for the Z Axis is composed of bronze bushings pressed in place into the ends of PVC pipes which have been reamed out to the correct size. The PVC pipes are in turn glued to a block of MDF which has two precision Vee slots milled into it. This is a system developed by Microcarve which has proven to be precise. I've increased the over length of the bearing slide to 4 inches to reduce the possibility of microscopic racking and the influence of the lead screw, in order to reduce the possibility of Z banding. The nut for the Z axis lead screw has not been worked out yet.
I'm beefing up the lead screw from 1/4-20 to a precision Acme 3/8-12 pitch screw. For a couple of reasons. 3/8-12 will be less likely to flex and transfer side to side force to the slide, again to reduce the chance of Z banding. Also, I have a lot of 3/8-12 acme lead screw materials left over from other projects. Waste not want not! Because the software is designed for a 20 pitch lead screw and not a 12 pitch one, instead of direct driving the screw from the end of the stepper I am adding reduction timing pulleys and belt, a 24 tooth pulley on the stepper, as 40 tooth pulley on the lead screw. This will create the proper drive ratio. An added benefit of the pulley drive is that it reduces t he transfer of vibration and deflection from the motor to the lead screw, which can sometime make it through the spider coupler.
Another feature I'm planning to incorporate into my design is a full 3 axis slide adjustment for the projector, using the same bushing/PVC/MDF technology of the Z Axis. This should make adjusting the projector up and down for resolution a breeze, as well as moving the projected area around the build window at 50 microns a snap, prolonging the life of the PDMS coating. I don't have any views of that mechanism yet but will post some as soon as I do.