I have cut a quite tall piece which worked fine. But at the end of the job, where CM asks you to turn off the router, the Z-axis did not go up to the max travel and as a result the brand-new bit snapped at the edge of the pocket, I had cut. That is very annoying, but I think it is also a potential safety issue.
Is there a particular reason why the Z-axis is not kept at its highest position? I did a few other moves and the head always drops down a bit. (if it is started from above the pre-programmed height, if it is below it moves up obviously.)
If you were working with very tall stock there is a possibility that your retract distance was too large causing the axis to lose steps when trying to retract. This will shift the working range of the axis down and can lead to cutting too deep and the issue you described.
Thank you for the reply. I did plan ahead for that by reducing the safe Z on top of my stock height. It was close to max Z, but below max Z. I did not anticipate that weird drop of Z You can observe the issue quite easily, for pre-programmed moves. For example after the homing sequence when it moves over to request a tool for measuring (I have a Shapeoko Pro with a Bitsetter installed and enabled). The first thing that happens after the homing sequence is done - as it is preparing to move to the front - is the Z dropping by some 2 - 3 mm.
So at the end of my tool path the Z was in that region between max Z and “let’s drop it a few mm” Z. As the bit was inside a pocket at that point, the unnecessary drop of Z caused the bit to snap, as It first dropped Z and then moved to the back.
Sorry for fuming, but it is just not necessary in my opinion and I do not see a reason why the machine would need to do that ever.
Start with that, and figure out what you have done to make “the machine” perform in that way. As long as you keep thinking that someone else is doing it, you’ll not figure out what the problem is.
I guess my explanations are poor. My G-code has no effect on the pre-programmed (macro?) moves. I can take a video of the homing cycle and how it drops as the X rail moves forward, if that helps?.
Not my video, but here you can see the Z drop after homing, just before it moves forward to do the initial tool measure cycle. Homing the Shapeoko Pro - YouTube
That motion is the home switch pull off move at the end of the homing sequence. The axis finds the switch backs off approaches a second time more slowly, finds the switch again, sets zero, and backs off ~3mm. That height is likely what is defined by CM as the maximum of the envelope and where it would position itself for the move to the back at the end of a program. I’ll confirm the next chance I have to spend some time in my shop. The behavior of CM is fixed. You may be able to use an alternate G code sender, there are several discussed in this forum. Can you change your fixturing (off the front edge or recessed into the spoil board, or remove a slat depending on model) or reduce tool stick out to ensure the retract plane has clearance? Deep pockets are difficult with the small Z axis travel.
I have made a few taller items with deep pockets (~2") and ended up making them in slices with alignment features (dowel/screw pilot holes) for easy glue-up. 4 slices make a lamp base for edge lit acrylic. Screws from the bottom hold the assembly together for gluing the first 3 layers, the top is only held on by the screws for access to install the LED tape.
Thank you for the insight. At the moment, I am not looking for a workaround. That project is done now and I am just trying to get more information.
What is linked is around second 34 of the video and I do not think that is the pull off move. At that point Z had been homed already. The pull off move for Z in that video is around the 14 second mark, IMHO.
From my observation that “dropped / pulloff Z” is not the upwards limit of the envelope. IIRC, it is possible to manually jog the Z up again until you hit the position max. Z that was found in the homing cycle. I am looking forward to hear what you find at your machine.
It is software. If there is no particular reason for the move (like shown at the 34 seconds mark), then the programmers could set that distance to 0 instead and keep Z at its highest position during all pre-programmed moves.
That all happens before any XY movement. Looks like there is another small drop in that video.
I think Motion uses a safe height of machine position Z-5?
You’re cutting it close if that 2mm is what broke the endmill.
I was able to run a quick check on my system (3XXL from April 2019 somewhat modified; HDZ - one of the last pretty blue ones, 80/20 extruded aluminum bed and subframe, HDPE spoil board, 1.5KW water cooled spindle, I just installed Neil’s 3D printed belt tensioners this weekend with new belts and V wheels , etc.).
Homing, either the first part of initialization sequence or $h from MDI leave the Z axis at a machine coordinate of -3. At the beginning of the tool measurement cycle, when moving to front center to have you insert a tool the Z axis is lowered to Z-5mm and returned there at the end of the Bit Setter cycle. The retract motion at the end of program brings the Z back to -5mm before moving to the back of the machine.
If your program had retracted higher than -5mm the final motion would be a plunge to that level prior to running the machine to the back. I would suggest creating a simple part with similar dimensional requirements and testing with alternate G code senders to see if these behaviors can be avoided. If you want to stay with CM you will need to account for the “missing” 5mm of stroke or petition @robgrz for a software change to keep everything at -3mm (machine coordinates). You can also shift the router/spindle a bit higher in the mount to cheat the top of the working envelope if you don’t end up with the opposite problem of not being able to go deep enough.
, etc.).