Just want to share a tip and some comments regarding the support article for adjusting the angle of the laser arm (which should be exactly horizontal).
The procedure described (actually popping the wheels off the rail!) seems to be only suitable for coarse adjustments, e.g. for errors larger than 1mm. It won’t get you anywhere to correct a small error like 0.2mm.
For fine adjustments, I’m adopting the following approach, which has worked for me a couple of times.
Don’t pop the wheels off the rail. As described in the article, place a good square with one arm against the rail and the other against the laser arm. While continually keeping the square firmly against the rail and simultaneously holding the laser arm firmly against the square, move the square and laser arm together forward and backward several times. Verify that the arm doesn’t shift away from the square when you release it.
This doesn’t always work the first time, but it’s the best I’ve come up with. And from what I can tell, the fix is never permanent; eventually there’s always some drift.
For testing the result, I cut the same test pattern (file attached) on two differently-colored letter-sized sheets of card stock, oriented horizontally. Flip the second sheet top-to-bottom, and place it on top of the first sheet. Carefully adjust the two sheets so that all of the dots on the left hand side are aligned. (The edges of the paper won’t be aligned; that’s unimportant.) If you hold it up to the light, you should be able to see the light through the narrow slits on the left-hand side.
Then examine the dots and slits along the top and bottom edges to see if they are skewed relative to each other. If the square was perfect, they will be aligned.
By the way, a few comments about the support article:
“If you find that vertical cuts are shorter than horizontal cuts… measuring the height and width” … Surely this is an editing mistake. A non-horizontal laser arm would not cause such an effect. A correct description of the symptom would be that angles aren’t accurate, and squares become parallelograms. (But the lengths of the sides of the parallelogram would actually be correct.)
“This happens occasionally on some units when the power is off … Once performed, it doesn’t need to be done again until the machine is turned off” … It would be nice to understand why powering off might cause this problem!