garageFab.cc

After I added the ENABLE control to both X and Z axes, it was time to calibrate the contraption. Besides centering the drill bit, there are basically 3 steps - a) determine which direction an axis goes when the axis direction pin is high or low, b) measure the distance between "all the way up" and "bit at the workpiece" positions, and c) determine the distance between the starting point on X axis (min switch) and the position of the first hole to be drilled.

Direction reference calibration

This was easy - an axis goes in certain direction until min or max switch is triggered; depending on which switch was triggered, the Arduino sketch prints whether Z_DOWN (or X_FORWARD) is high or low. These values depend on how the stepper motor coils were wired to the board. All other sketches then use these as reference values.

Distance between Z_MIN and bit_at_the_workpiece position

This was also easy - the Z axis is manually jogged to the position where the bit is in contact with the workpiece, and then the sketch moves it all the way up, counting the steps. The number of steps is printed as Z_OFFSET value.

Distance between X_MIN and first_hole position

Before getting into that, here is a short explanation of the drilling setup and sequence:

Take the angle stock that is at least 2" shorter than the length of the X axis threaded rod. Using templates, drill 2 pairs of holes at one end of the angle, and 1 pair of holes at the other end (the ends don't have to be a whole number of inches apart). Insert the angle into the long sliding element, single pair in first (two pairs oriented forward). The 4 edge holes are used to fix the installed angle with screws. Tighten up the forward nut combo on the threaded rod. Install 2" screw into the second forward hole so that it points sideways (hole labeled "SCREW"). The drilling routine would retract the angle backward until the X_MIN switch is triggered by this screw. From this position, the contraption would move the angle forward X_OFFSET steps to the first_hole position, drill, and would then keep moving it forward and drilling at 1" intervals until X_MAX switch is triggered. The X_OFFSET value is what we need to find out.

One accurate way to do it is to drill a hole (manually jogging Z axis) with the X axis at X_MIN position, and then measure the distance from this hole to the reference hole with a caliper. When two 3/16" hole centers are exactly 1" apart, the distance between the outermost points should be 1 3/16", and the distance between the innermost points should be 13/16". By averaging these 2 measurements, the actual distance between two 3/16" hole centers can be determined pretty accurately. The X_OFFSET (in inches) would be 1" minus that actual distance. Multiply that by 24 TPI * 400 steps/rev and you get X_OFFSET in steps.

I also tried to do it without drilling a hole, just by marking it on the area painted with Sharpie, and then measuring the distance between the mark and the reference line, but the measurement was not very accurate.

I thought of yet another way (but didn't actually try it) - to have a second reference hole on the template, 1" from the first one. Then the drill bit can be manually placed inside the second reference hole and the sketch can raise the bit and move the X axis to X_MIN, counting the steps. However there is a fair amount of play on the spindle of my drill (I bought really cheap one for experiments), and I think other (better) drills also have some play, so it would not be very accurate either.

Yet another (fairly accurate) method involves a reference hole 1" from the "SCREW" hole. The X axis needs to be positioned so that this reference hole is halfway inside the long sliding element - it seems we're able to perceive even minor imperfections of a small half-circle. From this (half-circle) position, the number of steps to X_MIN can be counted by a simple sketch. This distance between "half-circle" and X_MIN is equal to the sum of X_OFFSET plus the distance from the edge of the long sliding element to the drill bit centerline, which should be easy to measure with caliper - unless we moved the drill along the X axis (relative to the frame) after drilling the long sliding element through. The edge of the long sliding element should be pretty straight, too.

Testing

So, once the drilling contraption was all calibrated, I drilled lots of perfect perforated angle. Right? Well, pretty much, except the holes were not exactly 1" apart. And when I switched the angle around, the new holes (in the other angle leg) did not match up to the old ones. Oh, and the X axis motor stalled a couple of times. The results seemed bizarre. So I started doing some research on what could possibly go wrong with my setup and found the excellent resource, CNC Cookbook.

To be continued..