I’m new to CNC, but I have a lot of experience with laser cutting.
I’m having an issue with my Shapeoko 5 Pro where cuts are slightly out of square. Originally it was more than 1/8" out of square, so I unscrewed all the slats, loosened all the screws on the table and the frame screws, did the diagonal measurements using a tape measure and made them as close to each other as I could, and screwed everything back together.
The amount it’s off is very small now, but it’s still there and it’s enough to ruin flipped parts. I’m just not sure what else I can do… Does anybody have any strategies to get a better diagonal measurement while everything is unscrewed? There aren’t really any reference points so I was using the opposite corners of the outside slats with a mm measuring tape.
Do I just have to keep doing this hour long task over and over and making test cuts until I am lucky? I feel like someone out there has a better way!
I’m not totally sure how to know which axis is the issue. I have a part that is a 13" X and 9" Y rectangle, and the left side of it seems slightly lower than the right side, so I guess the Y axis would be the issue? It’s like the points are:
0.0,0.1 13.0,0.0
0.0,9.1 13.0,9.0
That’s scary that you have to fight with it all the time! I was hoping this would be a little more accurate.
I don’t own a S5 Pro so I’m not sure if this will help. If you move Y all the way back, is the distance between the Y limit switch and what ever triggers it the same on the right and the left? If not, is there an adjustment for that?
You said the left side is slightly lower than the right. That implies a Z axis issue.
Have you faced the table surface ?
What are you using to hold the material ?
I have to face both sides of the stock before I can be confident I have coplanar surfaces.
I have to check to be sure the bottom of the stock is actually on the table.
I have moved the stock up when using clamps the butt up against the stock.
If you’re 0.100 out over 13", that indicates about 0.400 out over the ~52" between Y rail bolts.
So secure your right rail to the table somehow & move the left rail 0.400" forward.
I used a master square on the cross braces under the table & was very close. (within 0.005" over 8"). Then I adjusted using an indicator to move one rail.
As Tod1D mentions, I used a very large machined square with high accuracy bought only to align machinery. This was done with the cross braces and Y rails snugged up and then checked on all for corners with the square. It’s as close as I can ever get it to square. I was surprised how close it was just using the old corner to corner cross measurement. I made several parallel sharpy lines down the Y axis on the spoil board and when running the surfacing tool paths they were square the full length.
I have one compound miter saw that never would true up on its fence for a 90 deg cut. It has a slide assembly and I finally determined the saw head fixture was cocked on the sliders slightly. It now only rough cuts lengths of material vs precise miters for picture frames etc……
Thanks everybody for the help! For future people finding this thread, this is how I got it much closer than the diagonal measurement technique:
I ended up mounting a 90 degree V Bit and put a large poster board on the table.
Then I used the 3 4 5 triangle technique.
First I jogged close to the top left of the poster board, and jogged down in the Z until the tip just barely put a hole in the board. Then I jogged the Z so it was floating above the board.
Then I went to MDI in carbide motion and typed G91 G0 X400, which moves it 400mm to the right. I jogged Z back down to put another hole, jogged Z back up, and typed G91 G0 X-400 to put the cutter back over the first hole.
Then I typed G91 G0 Y-300 to move the cutter down 300mm, and made another hole.
Once you have this triangle, you can measure the longest side and it should equal 500mm exactly. Mine was off a tiny bit, so I pushed the side of the Shapeoko a little bit and repeated the steps starting in a different spot on the board. I did that a few more times until I had it dialed in.
It’s much better now, though strangely it’s still not perfect. If I cut two 5" squares next to each other then put them together by the same face, they’re visibly not 100% aligned by ~.5mm at the farthest. My diagonal measurements are identical to each other, and the triangle is a perfect 500mm, so I don’t know what else that could be.
Just to confirm you machine is seeing the numbers you expect, it might be worth reviewing the gcode in another program like ncviewer.
That way you get an interactive graphic as well as the text.
Drill a 1/4" hole or bigger on each corner (furthest), put at slip/tight fit dowel pin, then measure the diagonals. Another way to do this is to attach dial indicator to the spindle and run it against known square that is square, bigger the better, like a granite square. Zero the indicator along the Y, then read the X, vise versa, and see how much delta measurements on X.
Adjusting the Y limit switch trigger points could bring it in even closure.
Make sure your fixture is also square with the spindle.
I wish SO5 has an adjustable hardstop (plunger) that we can adjust the Y1 and Y2 individually to square things up then we can adjust the trigger point also.
Check in with support@carbide3d.com — the switches are positioned on a fixture which is as precise as we can make it, and the entire gantry is then tested using those switches and if need be adjusted based on that testing — that said, they have slots, and at a minimum one could mark the current position of the one you are working on using blue painter’s tape or something, then loosen and adjust.