There are many that know things about calibration that I don’t. Since round had always been out for me and measurements never matched what was on CAM I just figured I would try to understand if it was a ratio (%) or specific distance out ( in both directions). My plan is to compensate with math with my own fine adjustment calibration. Hey, I understand it which makes a difference. Hope we both learn something here that prevents us from getting these results in the first place!
Where on the workspace you are cutting will also affect the errors you see. In particular, the errors tend to be worse at the top center and lower corners. The following links may or may not be useful:
After reading dlang’s post,“List of sources of errors”, I see some things that may be causing me some of my problems. It looks like it may be the chain sag. I am cutting these circles on the left side of the work space. I will try cutting squares and circles on all four corners and in the middle after I go over the re-calibration again.
I will look for flex it the motor mount beam too. (2 x4’s don’t bend, do they? Yes, they do! )
'dlang’s posted list had things I did not think up on my own.
Since I made the last post I thought I should check the length of the chain first. I mean the left length against the right chain.
Over the full length of the changes it is off a half a link.
I laid the two chain aside each other and it was very easy to see. I stretched both of them and it was true, one is longer than the other. After making sure they were tight I verified this.
Then I did the same from the other end. The first 3 feet or so they looked the to match, Then with each few inched you could see the miss match starting and at the end it was off about 1/2 a link.
I reversed one chain and looked again. I was thinking if the matched up the other way I could use them just be careful to install them that way. Well the two chains are not the same!
How do I get a match set of chain? Or is there a work around?
I will look into the Holey Triangular Calibration via the link you provided.
I was looking for an easier way. Well easier for me.
Is there an adjustment I cam make to the code or parameters where I could use the trial & error to get closer than 0.075 out of round. After re-calibrating this time I shaved 0.025" closer. to make round circles.
there is a thread I created, something line ‘sources of error’ that lists all
the possible error sources. There is also a simulator buried somewhere in GC
that lets you introduce error to many different items and see what the resulting
cuts would look like
but with this style of machine, there’s no X or Y value that you can tweak. If
you cut the circle in the center, it will be off differently then if you cut it
elsewhere on the sheet.
It looks to me at this point that I have to deal with chain slag,stretching, and the router must be in align with the ring the chain hooks to on the sled.
Cutting squares in the four corners measure correct. Meaning left to right and up and down in the center. Maybe I have to use the trail and error system more that I have been.
I just cut a new sled and at 18" left to right and 17-11/16 up and down, 1.7%. At 2-5/16 it was off 0.100"., 4.3%
I checking to old sled for being aligned to center.
I will look at your “List of sources of error” and see what I can do to improve things.