Wilson X carriage

I posted this on Thingiverse a while ago, but it has worked so well for me that I feel I should write it up properly.

The TL;DR version of my design

• Placed motor on rear of X carriage to balance weight front-back
• Heavy components in the center, light ones on the outside
• Geared BMG extruder, for increased filament grip, and less motor torque required
• Brought hotend closer towards the X-axis, minimizing effects on nozzle height from X-axis twisting
• Solid metal filament guide path, adding rigidity and reducing slop in the filament positioning.
• Extra mount holes for accessories like bed leveling sensors.
• Centrifugal fan, providing greater air flow at higher pressure vs axial fan.

Development

The existing problem

The Prusa i3 design has worked very well for me over the 3 years that I have been 3D printing stuff, but there's one area which I thought could use some improvement. That's the x carriage, the thing that actually holds the hotend and moves it around. On my Folgertech i3 clone as designed, the extruder motor is mounted onto the front of the x-carriage, and the hotend and extruder mounts on the front of the motor. This is not ideal, the main problem being that the all of these parts are reasonably heavy, and being cantilevered out like this all on one side puts a significant torque on the x-axis rails, enough to bend them slightly (they are rather springy). Because the hotend is placed all the way out at the end of this cantilever, it experiences the greatest effect of this deflection. Note that the rods bend more when the x-carriage is in the middle, and as it approaches the supports of the rods at either end, the bending goes away. The end result of this is that the hotend sags as it traverses the x-axis.

New, Standard, Better Hardware

Additionally, I upgraded my hotend to an e3dv6, the hobbiest standard for 3D printers. It's a tried and true design that has proven itself to work well on all sorts of plastics. However, it is a bit longer than the Folgertech extruder. When I found that some filaments were requiring too much force to push through the all-metal hotend, I upgraded to a BMG extruder clone. The BMG extruder design is also an excellently designed product, with properly shaped drive wheels that grip the filament better than a cylindrical knob and a passive roller, which is used in the cheap extruder that comes with many printers. The BMG drive wheels are even geared together, to provide a dual drive action pushing the filament on both sides. The BMG extruder is also geared down to the motor, requiring less torque at the motor shaft, and potentially more resolution as the motor is making more rotations.

Since I got the printer, I had also been experimenting with print cooling techniques. The Folgertech comes with no print cooling at all, and the first suggestion for cooling is usually to add in some axial computer fans. Usually, these will have a funnel on them to being the air down to a concentrated stream that can be directed at the plastic below the hotend. These, plain and simply, don't work. I have tried several funnel designs using axial fans, and even increased it to an axial server fan that sounds like a jet engine and moves a massive amount of air. As soon as you try to concentrate the air flow, a majority of the air is lost. I noticed that the official Prusa i3 uses a centrifugal fan for its print cooling. I picked up the cheapest one, and it was insanely better than any of the axial fans I tried. Also, it seems to be higher pressure air, meaning it can flow through longer ducts and around corners without getting blocked up. I was quite surprised during a print with this fan, when passing over a tubercular section of the print, it whistled like blowing across a bottle. That kind of air flow is much better than anything I had before, and if it's too much I can turn down the fan speed easily.

Attaching it all together

To address the problems with the X carriage described before, I decided to just make up a brand new X carriage. First, I placed the motor behind the X - carriage, which simultaneously distributes the weight on the X axis, and brings in the nozzle closer to the X axis. The BMG extruder mounts in front of the motor on the front of the X carriage, and the filament goes down through a short tube to reach the hotend mounted below. The fan then mounts on the outside of the hotend, and air is directed down a duct to get to the print. I also added some extra mounting holes on which to mount accessories like a bed leveling tool, or a pen or something. These can be designed later and just bolted on.

Problems with the filament path

The BMG extruder is strong, and the all metal hotend provides a lot of resistance to forcing the filament in. It's such a large force that the little plastic clips on the short section of tube between the extruder and hotend can't take the load, and I observed the entire X-carriage bowing inwards and outwards from the force. Obviously, this is not a long term success plan; I knew I needed some kind of positive coupling between the hotend and the extruder. The BMG extruder does provide an E3D style mount, which you would use if the hotend was mounted just under it rather than an inch or so away. I decided to make a metal extension for it, with a metal pipe and threading holding everything together. I modified a E3D clone heatsink so it had an M8 thread going out the top as well as the bottom, and then threaded a long steel shaft with M8 threads before boring a hole down the center. This hole perfectly fits a section of Teflon tubing, to cut down on friction as the filament runs down the tubing. Next, I cut out an E3D compatible mount on the lathe, and threaded it for M8. This screws on the top of the pipe tube, and the pipe tube is set with two jam nuts. The threading of the entirety of the pipe tube allows the distance to be adjusted, too.

Files

File:Wilson-3dprinterfiles-v1.zip - A whole bunch of misc files I have made (and some existing 3D printer parts) including my X-carriage design.