Working on the Y Axis

Along with trying to finish my extruder I've been toying around with some different ideas for the Y axis.  I am defining the Y axis as the axis which carries the X axis back and forth on the top platform.  The X axis is the axis which moves the extruder head back and forth.  Originally, I had intended to drive both the Y and X axis using leadscrews.  This would allow for ridiculously high accuracy ((25.4mm/inch) / 18 TPI / (200 steps/rotation) = 0.00706mm/step) and the required torque from the stepper motors would decrease significantly due to the mechanical advantage of such an approach.  Originally I thought this would be a good thing, however I have since found that the cheap 5/16" rods that I have are prone to bending and deformations.  If you get one scratch in the thread the carriage will catch on it causing the machine to bind potentially tear itself apart.  After encountering some of these problems with the Z axis, I decided I didn't want to deal with it again.  

So I started moving forward with the traditional approach of using the chain to directly move the axes.  The main consideration I had with my design was trying to avoid binding at all costs.  If one side of the Y axis moves further than the other you can get a situation where the Y axis gets stuck or binds.  I like the idea that the current reprap Darwin design has where one end of the X axis has two holes that move along the Y axis rods, while the other end has only one hole that moves over it's guide rod.  I think this will greatly help with preventing binding and thus I intend to implement this in my design.  The other factor that I think have an impact on how smooth the carriers translate on their guide rods is where the chains are pulling from.  You can imagine that if the chains are pulling to the left at a height of 50mm above the guide rods, a significant moment is created in the carrier.  It will tend to drive the left most hole down into the guide rod, and pull the right most hole up into the rod.  Now if the chain is moved to pull only several millimeters above the guide rods, this adverse moment is greatly reduced.  For this reason I intend to try to mount the chain as close as possible to the guide rods so that the majority of the force goes into causing translation along the guide rods and not binding.  

So with these things in mind I spent a good deal of time playing with my design trying to find something that worked.  The problems I kept running into was that my motors are only single shaft motors.  They don't have a shaft coming out of both sides as some designs require.  Second, the guide rods are positioned towards the outside of the machine and would be difficult to get a chain all the way out there.  Then I realized that I can just switch my square corner brackets around so that the drive rod is now towards the inside of the machine.  This would make getting a chain positioned next to it easy to accomplish.  The only negative effect that I could think of is that the build area is slightly reduced, but as the recently released cupcake CNC machine points out, most parts are about the size of a cupcake anyways.  So I went ahead and switched everything around and designed, cut, drilled, and mounted the motor and bearing supports.  Here is a pic of the end result

You can see that a tiny bit of the build table is now covered, but the huge benefit is that I was able to align the chains directly overhead of the drive rods.  Here is a better shot illustrating that.

I think this will be more than adequate for driving the X axis back and forth along the drive rods.  I still need to get a die set to put threads on the ends of the 5/16" guide rods so that they can actually be anchored and help square up the machine, but that is for another day.  For now, it's 70˚F outside and I intend to spend the rest of the day outside.  Hooray for spring!