My spoilboard is getting pretty torn up so getting ready to cut and mount a new one, so figured I would double check calibration, square, tram, etc before cutting the new features so I have accurate locations. When doing the squaring I drilled four pins at 38" X and 48" Y which was a bit long in the back-left-to-front-right dimension. I also cut a smaller square 4" X and 4" Y but this one is longer in the back-right-to-front-left dimension. I can’t figure this out. It’s not a measurement issue as I’m only comparing which direction is longer, so even if the tape measure is off, it’s still relative. Any ideas?
Just one idea…
Off by how much? Your machine is not perfect and has some backlash, which means the location of a hole depends somewhat on which direction the machine moved to get there. Off by a few thou could be because of this, and your small square is more likely to be affected by this (relative to its size) than the large one.
When I need more precision, I modify (or write) the gcode so that the machine first moves to a spot “up right” (X+1,Y+1) of the target hole, so that it always approaches each hole from the same direction, so that the backlash acts on each the same way.
Related:
This is how I like to do it… Do it as big as you can so that anything you cut smaller will have a smaller error.
Thanks both. The method you proposed is pretty much what I follow. I don’t remember what prompted me to do the small square at the same time (I think so I could get a pair of calipers on it) but was really surprised when it showed out of square in the opposite direction. On the big square it was around 1/8" out, less on the small square of course but the backlash is a good idea to check.
Something to consider. You can have your 4 points out in the extreme corners perfectly square, even if you have a huge bow in your gantry, or if your two Y rails are bowed in the same direction. So that would give you different squaring results for a small square at different locations on your table.
To check for that, get a long straight edge and lay it on your X and Y axis’ one at a time and then run jog a Vbit or something along it to see how straight each axis travels from end to end.
Its pretty easy (for example) to pull a bow in the Y when you are attaching your spoilboard if you aren’t careful.
I created and printed two of these simple fixtures to assist holding the tape measure across the long distance and get an accurate measurement. The end of the tape goes in the slot, and the printed part sits on top of a .25" dowel I installed into sacrificial blocks like Eric’s video has. Made it easy to get an accurate measurement with a single person.
Step file and parametric fusion project attached.
Tape holder for diagnol measurements v1.f3d (102.8 KB)
Tape holder for diagnol measurements v1.step (65.7 KB)
Brilliant idea!