Prototype #2: as seen on...

Prototype #2: Improvements and Specs

Prototype #2 has many improvements over typical CNC setups.  First being collapsible.  Now people (like me) who can't afford a permanent space to be used for a CNC can store the machine in a closet when not in use.  It is easier to carry it from point A to point B as well, carrying lighter parts than heavier wholes.  Though being collapsible creates issues as well.  It can take longer for setup and has to be aligned every time to make sure the dual X axis tracks are parallel.

A second improvement is having an open table, though some CNC machines have this as well so it's not as a break-through.  This does allows for oversized materials to be placed through the CNC, like a tree log or long lumber.

A third improvement in prototype #3 is it allows materials/stock to be worked on from above the surface, like any other CNC machines) but also below the surface.  I can do engraving work up to 6" below the surface if I wanted to.  This means I can prop my machine on top of an oversized picnic table and engrave on the table top.  I can technically mount the machine to a living tree and do some work on it if I wanted to.

As a typical setup, I clamp an MDF wasteboard onto the CNC then mount the stock to be cut on the wasteboard (as pictured).  Cutting area is about 17" x 17" x 8" depth.  Depth can be easily adjusted by bolting the router adapter plate up or down, or by making a new adapter plate.

Prototype #2: pictures

Prototype #2: Final Model

This is the end result for build #2.  Because of my limited amount of space, I designed it to be collapsible for ease of storage and transportation.  But because of this setup it requires a 4 controllers and a 4 axis controller setup; 2 motors to power each end of the X axis.  The bottom part of the legs can be removed by removing the two bolts on both ends.  Overall it is about 3 feet cubic area by size.

It is not truly portable as I envisioned, more collapsible.  It is slower than I hoped but precise (roughly 0.001" accurate).  I had some alignment problems originally, was making skewed ovals instead of circles, but I was able to correct the problem.  As my first CNC I completely built and designed, I was happy especially as I never had a game plan to go by.  And because only 3 months prior I thought it was impossible to build your own CNC.

Prototype #2 took about 1.5 months to build and complete and about $1200 total in parts and materials (most of which salvaged from build #1).  I fried a control board by accidentally crossing wires, so had to buy a new $200 board to replace it.

Learning the CAM (computer aided machining) programming can be the hard part especially if you are on your own.  CAM can easily cost as much or more than the CNC as they are not cheap, but cheap and free/open-source ones can be found, though they may not do as much as retail CAM programs.  Luckily I already knew basic CAD programming and am experienced in the computer field (my 1st career of 7 years was in the IT support field).  Go to the CNC Zone forums for more info on the open source program alternatives if you can't find any retail software to play around with.  I personally use RhinoCAD to do modelling, MasterCam X4 and sometimes RhinoCAM plugin for my CAM, and Mach4 to control the CNC.

Prototype #2: Building

(My main hard drive crashed in Summer 2009, so most of my process pictures are gone).

So after the incomplete 1st attempt, I decided to make my own CNC design with improvements I find fit.  Because I am working in my backyard with minimal tools and no welder, I decided to choose wood as the main material again.  Wood is a common material for DIY CNCs, it can create highly accurate CNCs when properly built.  I do not have any plans for prototype #2, sorry.  I designed and fabricated it as I went along.

For prototype #2, I decided to use Baltic Birch instead of MDF as per prototype #1 plans.  I did not like MDF as it does not handle screws that well.  I discarded the homebuilt bearings and slides (using skateboard bearings and aluminum angle bars) as it was not up to par for my expectations.  I went on eBay and found used high precision bearings and slides.  The 1st picture shows the back of the Z axis using fully supported slides and Thompson bearings, really nice stuff!

The X and Y axis (shown in picture #2) I used precision slides and bearings (forgot which brand) but not fully supported.  They work great but found that aligning them during assembly is much harder than fully supported bearings.

I used the Acme 1/2" 20tpi lead screw from prototype #1, as the anti-backlash lead nut.  TPI is teeth per inch, so 20tpi means it has to spin 20x to move 1 inch.  The result of this combo is dirt slow!!  It crawls about 1" per second movement.  I recommend anyone in using a multi-start Acme screw (as I have done on my newest CNC) as it will speed up movement tremendously.  More TPI means more precision, but who really needs to be accurate 0.0001" increments?  Your backlash tolerance will be more than that so in reality your precision is only as good as your backlash movement.