Ok, at first I thought this was a simple issue of either small deviation with Bit Setter, or just one of my tools touching off a bit inconsistently. I can now tell that’s not the case. I just finished a test piece.
3D Adaptive, leave 1.5mm stock radial. 3 flute 1/4" upcut.
3D contour with a 2 flute 1/4" downcut.
I’ve done two tests, one at regular feedrate, and one at 70% feed rate thinking I was too aggressive. My regular feedrate aims for about 1.5 per tooth, the 70% run was hitting 8 tenths (0.0204mm) per flute for roughing and 5 tenths (0.0145mm) for downcut! So hopefully that can rule out tool deflection.
It’s not BitSetter, as then it would be a consistent offset at each depth.
It’s not my feedrate. Under 1 thou per tooth should be easy, all day long, right?
It’s not a tram issue as it’s consistent around the contour. I’ve also checked my tram with the Edge Tech Mini Tram tool.
I guess we could say “Hey, don’t treat your roughing pass like a finishing pass for the floor” - but the reality is, the floor of the cut should be at the same depth regardless of the type of toolpath used.
I’m stumped. I’ve had this issue for a while, but not forever.
Z-axis motor wiring all in good condition and all connectors secure? Send photos of the connectors in to support@carbide3d.com if you don’t find anything obvious to address.
Z-axis got a full lube a few months ago and has barely seen much use. With power off, a slight pressure on the spindle will cause it to drop on its own. Had to start putting blocks under sweepy when not in use
Control board is also quite new as the last one flaked out after a while, one of the homing sensors would no longer work. Support swapped out the board after diagnosing.
I’ve thought about losing steps… it’s scary to think about. A rabbit hole with no end to go down. Why now? Why not any other visible issues? This is stock Pro with C3D spindle. Also we should consistently lose steps on retract if it’s too heavy. If that were the case, the cleanup pass should be getting deeper and deeper as it fails to retract to the height it wants to, right? The final and deepest pass should show the worst.
Wear in the Z screw? This thing has had a pretty easy life.
Same thing. In the photo, there’s actually a 2nd clone to the left of that with axial stock to leave at 0mm, since initially I thought it was a simple offset issue. So on that one, the first and last depths are perfect, the middle two depths are cut too deep.
I just setup a test indicator, and hand coded 50 or so Z hops from 0 to 5mm. No deviations. Ran it once while pushing down firmly on the spindle, no issues. Ran it again while pulling upwards on the mount, no issues other then the gantry itself deflecting and then returning once I let go.
