Proper selection and mounting of the bed is absolutely crucial to building a 3d printer that is reliable and easy to use. I am using aluminum tooling plate for my bed which is a very common choice as it is rigid, flat, and thermally conductive.
Traditionally, 3d printers use a screw and spring system to level the bed, however I decided that adjustment mechanisms will not be necessary for my design. I am confident that I can build it sufficiently flat and level to begin with, and that my frame and linear motion will not drift over time.
I will be mounting the bed on four stainless steel spacers. Stainless steel was chosen for its low thermal conductivity, which should allow the servo to remain cool even when the aluminum bed is hot.
I purchased four 12mm spacers from Misumi. Ordinarily I would expect this kind of thing to arrive +-0.2mm or worse, but the individual spacers were incredibly close to 12mm which was very surprising.
That being said, there is still room for improvement. I'm more concerned with having the four spacers match than their overall length, so I will be shortening the two longer spacers. The easiest way to make sub 0.01mm adjustments is by lapping with wet-dry sandpaper.
I used 1000 grit paper and made a small number of strokes in varying directions to ensure I didn't knock them out of square. Every two strokes seemed to remove about 0.001mm of material so this went quite quickly.
I know with absolute certainty that the MX80L linear stage is flat, and now my four spacers are perfectly matched as well. As long as the aluminum plate is also flat, the axis should require no alignment at all.
I shortened the bed wiring and terminated them with connectors for convenience, and prepared to assemble the bed.
Unfortunately I never modeled the heater in CAD and simply assumed it would be 10mm thick over its entire surface. In reality, there is a bulge where the wires connect to the heater which interferes with my cable chain mount.
Luckily the part was 3d printed so it was easy to fix. Here is the revised mount freshly printed with support still attached.
And here is the cleaned up mount next to the one it will be replacing. Note the Velcro attachment point to keep wires in place.
With the bed finally attached to the stage I can finally run an indicator along the surface to see how it turned out.
I was surprised to discover that the bed is 0.05mm (0.002") out of level, with all of the error coming from the aluminum plate. Technically the aluminum plate is still within spec, but I was expecting such a small section to be nearly perfectly flat.
As with the Z axis, I chose to shim with tape. I could have alternatively shortened two of the spacers by 0.002" but then the bed and spacers would have become a matched set and I would have to track which spacer matches each mounting location. Tape is much more convenient.
Sweeping the indicator across the table one more time...
The table is still has about 12 microns (0.0005") of total error, and is approximately 5 microns concave. This is solvable with a few swipes of 1000 grit sandpaper if necessary, but I don't expect such a small error to have an impact on print quality.
No comments:
Post a Comment