Note 1: I capitalize "X-axis", "Z-height" and the like for legibility when scanning/skimming; pedants can hate me later.
Note 2: This is where idealistic, that is mimal change for best outcome, modifications are. For measurements and construction details of the mechanical structure see Part 1.
Disclaimer: Due mostly to human error but some manufacturing tolerances, the following apply.
- Diameters/measurements under 50 mm have a +/-0.05 mm tolerance
- Lengths over 100 mm have a +/- 0.5 mm tolerance
Let's get crackin'! the simplest two axis to modify are the X-axis and then the Z-axis. Before we start, keep in mind that increasing the travel distance of any axis creates the ability to have the current wiring suddenly become too short. Effectively, you're going to have to re-work a good portion of the machine in order to have these modification 100% usable.
Either way, it is possible to stiffen the Z-axis a bit by boring out the carrier's rear bearing mount, the two mounts for the shaft, and dropping in an 8 mm shaft. Though the most flexion will take place on the front two shafts, the rear shaft can be upgraded to an 8 mm shaft and linear bearing to aid in rigidity.
Z-Axis
The graphite injection molded Z-axis-bearing-and-A-axis-NEMA-17-holder is 45 mm tall. It has the sheet steel cover that is a hair over 58 mm tall that bolts to it. If you remove this, you technically gain 17 mm of z-height travel. At this point, there is 25 mm of threaded rod left before you hit the coupler. Cut, file, or grind off another 8-9 mm of the back chassis's frame, re-position the z-height limit switch by 1" or ~25 mm, put a 1" spacer under the tower, and you just increased vertical build height by ~20%. In you want more and have a greater mechanical inclination, you can buy replacement Z linear rods and a threaded rod for the Z-axis drive for less than $25.
X-Axis
The X-axis is a different story. Both X-axis rods are friction fit into the graphite composite pieces and have a little bit of glue/sleeve lock/retention on two ends. With the help of some elbow grease and vice grips, I can matter-of-factly tell you you that the rods are 209 mm and 287 mm long. The top rod slides through 2 holes thus provides alignment for the bottom one to a degree. Their total is just under a 500 mm stock length.
McMaster's next size is 2x the length, 1.2 meters and though it's expensive compared to the printer at ~$32, it's not that bad. Now and 8 mm rod will be stiffer, but also heavier and the Z-axis carrier doesn't have enough meat to bore out the holes. Given the machine we have, I'd say that if you want to expand the X-axis too much beyond 200 mm, you'll need the rigidity of 8 mm rods for this gantry. The added weight will then necessitate at least 1 of the 3 Z-axis rods being switched out to 8 mm as well to counter the flexion.
Z-axis Carrier
There is one problem though. The X-axis carrier is only 17 mm OD for the bearing retention portions of the composite. On the Z-axis composite, the front two are ~17-17.75 mm OD, but the rear is 20 mm OD. You'd get a 1-1.375 mm wall on the front two if you wanted to put an LM8UU/LM8LUU in there but the rear bearing retention portion would give you about a 2.5 mm wall. BUT, the rear Z-axis linear bearing mounts are 11 mm thick, not 9 mm and the rear A-axis bearing retention portion of the composite is 20 mm. In other words, swapping out and drilling out the single Z-axis linear rod from 6 mm to 8 mm is feasible without facing integrity.
Belts
As soon as you start talking the extension of the X-axis or Y-axis though, you have to address the drive problem; the timing belts. I'm going to default to McMaster again on this. It appears that "MXL" is a designation for miniature extra light which is a designation for tooth pitch and that the number is supposed to reflect the number of teeth thus 236MXL means 236 teeth at 0.08" tooth spacing. McMaster's neoprene belts have more teeth that this, but their urethane belts are just fine. So as such the 235MXL...
Read more »
your equations need some work. The 1st term should be Pi*10.15 (circumference) rather than Pi*(10.15/2)^2 (cross sec area). The left side um values are correct but might be good to convert to volume )(cu mm) then to the width of the trace for a given layer thickness considering print speed.