This looks a classic case of machine flex, with the dead-giveaway being that the cut-side swaps when you switch climb/conventional cutting direction. EDIT: Or something loose (V-wheels) that’s causing excessive slack. At the end of the day the spindle is deviating from its path mechanically.
@radioteeth - I don’t doubt the flex. My cut preview always shows a full depth of cut throughout the job. I have and will continue to check the v-wheels. What still confuses me is how any of this causes one side of a shape to not cut all the way and the other to cut through beyond the set 100%. The bit spins the same directions and cuts against the shape just the same on both sides… no matter how large the object is. Looks like the bit digs in (down) when moving South toward the front, but pushes itself up against the material when moving North away from front. What would consistently cause this reaction?
UPDATE: I’ve run 4 jobs since my last update. The first two (the same file) had no cut/uncut issues. The next file, that I ran when I first shared this issue, DID have the issue as shown above. The fourth job (new file) had the issue, but it was 50% less severe.
For all four runs I reduced my depth of cut per a few of yall’s suggestions. I believe this has been the primary reason for any/all improvement in the cut/uncut problem. I may need to reduce further and/or reduce feed rate. If that continues to resolve the issue until perfect, I’ll conclude my feeds/speeds were too aggressive. That does not eliminate that existence of flexing/loose wheels/etc., but updating all of those elements had no immediate or noticeable change to my issue. Maybe due to less aggressive feeds/speeds those flexing/loose components are not being aggravated. Either way, I’ll keep my eye on those mechanical elements going forward.
I hope anyone in the future can see that reduced feeds/speeds did positively impact my efforts to resolve the cut/uncut issue. But, at this second, I’m not 100% sure that’s the only problem or if those changes are all someone else may need to implement.
Anything I’ve missed?
I had a similar problem and it took me a while to figure out my problem. Seemed like I had too much play or flex in my spindle, but when I would raise the z axis and try to move it it was pretty solid. Finally removed the spindle and Z axis plate (I guess that is what it is called…the plate that the spindle attaches to) and everything looked normal. I removed the wheels at the bottom and found that one was broken on one side.
As long as the wheel was engaged everything was fine, but when it got to certain levels where the broken piece was exposed, that wheel lost contact with the plate which led to crazy cutting patterns.
Here is a picture of my broken wheel. Replacement coming today, so hopefully I will be back up and running soon. Many thanks to the Shapeoko team for the quick replacement.
Hi @rvs0002, I just realized I never responded to your comment. Nice photo too. I need to go over my wheels more closely it seems. It may account for some of my cutting issues, though likely not issues that occur only when X & Y are engaged and Z is static. But who knows, I could be wrong.
UPDATE: I’m beginning to suspect my hold-down method. Or at least my individual use of the method. I drill holes through my stock and put bolts through that screw into threaded inserts that sit inside my wasteboard (made of MDF). That is not a bad method from all that I’ve read. However, I may be using too large of a drilled hole in my stock and thus it’s creating some play for the bolt. Plus, the MDF wasteboard is probably soft enough to allow for movement of the insert due to pressure from the bit against the wood mid cut.
Regarding z-axis, I’ve had several recent instances where the bit made a large jump off track. I’ve looked over the wood afterward each time and the most likely cause is the bit is not making it out of the cut (pass the top of the stock) before moving in the next X/Y direction. So, it gets hung up, then skips and continues cutting in the wrong place. Very annoying and costly. Oddly, a couple times the jump happened while on the second of 4 passes required to complete that shapes cut depth. A knot in the wood is what I’ve blamed in those cases.
I have more investigating to do. Thanks all