A 1" square pocket tool path in VCarve Pro v11.504 yields .985" all the way around
A 2"x1.5" rectangle yields the same .015" smaller actual cut all around
Bit is Amana 1/4" downcut endmill measuring .247" (best I can measure on the sharp edges)
Water cooled spindle has no runout that I can detect using .001" dial indicator with bit installed
Calibration of machine (X50 WW) tested along 30" X and Y path is dead on with Woodpecker rule
Machine is 1 year old and “I thought” I had measured it when I first got it and 1" was 1"
I being to picky and asking too much?
When I did my first inlay I had to make many adjustments in order to get the pieces to fit. I assumed it was my technique. Now making a simple socket set holder out of 3/4" Birch and that’s when I discovered the discrepancy. As an example I would try to create a circle that is .81 inches and it would come out roughly .795"
What say you, oh smart and friendly folks of the Onefinity CNC world?
Well, I’m probably in no position to give advice as I await my very first CNC machine, but I seem to remember reading somewhere that you can set your toolpath/bit to run “inside the line”, “outside the line”, or “on the line”. Could this possibly be contributing to the error? Just a thought…
I am curious if the issue scales or is a constant?
Could you try a few tests squares of increasing random dimension (1"x1", 9"x9",…, 15"x15",…) and see the results - I believe you mention you have a precision ruler?
I wondered that as well, but couldn’t get my head around how it would show up/impact the milling of squares. Doing a quick backlash test with an indicator would be a good trouble-shooting step (good routine in general from time to time I would think).
On my new Journeyman, I typically have to adjust feature dimensions by about 0.003" (via a negative “stock to leave” value in Fusion 360). That’s when doing conventional (non-climb) cutting with a spiral bit.
After triple-checking all the obvious dimensions and machine calibrations, it dawned on me that a bit will almost certainly experience some side forces as it cuts the workpiece material. How much depends on a lot of things, including whether you’re using a climb or conventional cut, and the sharpness of your cutter. But regardless, no machine is perfectly rigid, and most have at least a little backlash. So any side force will necessarily deflect the bit slightly away from it’s ideal path.
Machine rigidity isn’t likely to change over time, but backlash certainly can.
I agree. But I was only cutting .15" deep in Birch to test things and lowered my feed rate to 50 ipm and my 1/4" downcut bit is short so that wont deflect either. At least not enough to detect.
I’m gonna be 61 next month and I can’t remember what I ate for dinner last night, but I’m pretty sure I tried the precision of the machine in the beginning (a year ago) and finding it to be dead on. The error isn’t cumulative. In other words it’s the same whether I make a 2" square or 6" circle, the error of .015 to .018" is always there. To me that points to the bit, but all of my bits??
I still really love this machine and I think I’ve only scratched the surface for it’s use, so I am not complaining. Just trying to learn. And I’m VERY GRATEFUL to those who have been offering suggestions!
Hi Bill - do you get the same variance in size when you do a profile and pocket? I remember my shapeoko would make perfect profiles, but pockets would always be the wrong size (I don’t recall if they were too big or too small).
Assuming the gcode is correct (perhaps you can post it), I can’t think of anything off the top of my head other than a mechanical issue as (the other Tom) pointed out.
And I agree that 0.015" is too much error to attribute to bit deflection. But I still think you need to investigate backlash, although it’s admittedly suspicious that you’re getting a similar error on both the X-axis and the Y-axis. Even with zero force on the bit, backlash can cause positioning errors when the destination positions are approached from opposite directions.
Here’s a simple test you could try. Using the keypad on the controller, set it to move 0.005" per jog, and jog each axis back and forth by that amount. Measure how much the spindle actually moves. If it’s not moving at all, increase to 0.010" (by jogging twice in the same direction), and try again. Etc. – until the spindle does move when you jog. That should pretty quickly establish if you have backlash, and roughly how much. Even if the problem is way back at the motor coupler.
you’ve tried a lot of different things. Have you tried re-setting your controller? Are you on 1.09? 1.1 was buggy… and I just noticed 1.2 is out (saw it in a screenshot) Maybe just reset your controller to factory defaults and then run a new test.
