When starting out, we thought that the machine was pretty well dialed in in terms of the calibration.
We didn’t do the final calibration and didn’t realize that things were more than a “bit” out-of-kilter.
After running what was supposed to be a circular logo and not seeing it until it was cut from the test board, the circle was actually an ellipse…obviously it was rather important to do the final calibration! Once this was seen, decided to do some more checking of items that had been run as tests earlier.
We had run some additional pieces from a previous conference table design for further testing and found some major variation in the horizontal and vertical measurements. The calibration would need to be brought under control to be able to run items that required any kind of accuracy.
Long story short, did the final calibration and ran a quick test (9″ circle with an inscribed square). This appeared to be very accurate. But, decided to continue with some more tests.
Decided to build a small platform (16″ x 16″) for our 1/2-height drill press and also scribe increment lines to go into more depth for the checking of the accuracy (1/2″ and 1″ spacing). There were still some considerable inaccuracies; a fat 1/8″ over a 6″ length for the increment grid, along with a variation in height of the platform from top-to-bottom of a 1/16″ plus.
After consulting the “Great Google,” found some rather interesting things for improving the accuracy of the calibration process. One of the things that is really important, is the orientation of the gears on the motors. When initially done, this was eye-balled. After thinking about this, the variation here could be pretty broad when just “eye-balling” things. We found a template for creating a “level” that be used to zero in this orientation. This is a really outstanding idea and looks to be fairly simple to execute. You can go to Creating the 12 O’Clock Calibration Level in FreeCAD to get more in-depth information about this.