Still pretty new to CNC. Done some successful cutting with Openbuilds CAM, so decided to move over to Fusion 360 as that's what I use for CAD / modeling for 3D printing. Want to start to experiment with more 3D cutting. I'm using the OpenbuildsGRBL post processor by swarfer to generate the g-code. I homed my machine, then moved the bit to the starting point of my material (matching what I set in Fusion), and set zeroes. Started up the job. The spindle spun up, but then the tool plunged into the material about 1 mm (guessing here, but it appears to be about the depth of the first cut), then moved across the material to where it was going to do the first cuts. (Which was at the far end of the material.) The rest of the cuts went well and as expected, other then having this slash across the top of the material. (Well, the rest of the cuts up until the point when my PC decided to reboot about 1/2 through the job.) So what am I missing here that it would move Z down just a bit before moving X and Y? Like I said, all other cuts seems to go as planed, it was just that initial move.
swarfer/GRBL-Post-Processor relates @David the swarfer (; explains it better than I can but I think if yours plunges down your homing is setup the wrong way round, G53Z0 should be at the top (retracted)
OK. Trying to wrap my head around this. I see at the towards the start of my gcode, there's a G53 G0 Z-10. I do recall, after homing, I had to move down to about -9 or so to reach the top of the material, before I set my working 0. Since the G53 uses machine coordinates, then that would make sense that it plunged in about 1mm at start, since my working zero would have been about -9 in machine coordinates, if I understand this correctly. My endstop is at the bottom of my Z c-beam. (FYI, this is a pretty stock LEAD 1010). So it triggers when the Z is full retracted. So, if I'm understanding correctly, my Z0 machine coordinate should be fully retracted, which should be correct, I think. I guess the question is, why it does the Z-10 move at the start, vs a slight positive move to clear the material. Probably something I've screwed up, just not sure what yet.
Or you have a Z-10 move, with only Z-9 of free space? And it goes up, jams hard against the end, skips steps, which knocks the position out? Something like that, watch it closely and look at the gcode at that point, sending it line by line from the serial console can be helpful too to see which moves break things for you
I was watching when it happened. It was the very first move when starting the job. Most definitely didn't hit an end. The move itself was only a 1mm move. After your first comment above, I'm confident it was this: G90 G53 G0 Z-10 Since I was sitting at machine -9, it ended up being a -1 move. Taking a look at the post processor now, I think I may have figured it out. Going to test a couple of things and will followup shortly.
So, it was a setting in the post process. Didn't pay that much attention to it at first. There is an "End of job" position for each axis. Z is an absolute position, in machine coordinates. It was set to -10. Apparently the processor also uses this value as a safe Z height at the start of the job as well. So that's where the initial G53 move is coming from. (Verified by changing this value and post processing again.) So, even if this had not happened in the beginning, it very well would have happened at the end of my job, if any material was in the way of this final move. So I guess I'm glad it happened when it did, rather then messing it up at the very end! Lol.
Awesome! I understand now! I dislike the G53 in there personally, i like hiding machine coordinates and just have users focus on a single coordinate system (so z safe is G54 at Z0+5mm up for example) I know i know G53 Z0 is industry standard safety , etc etc but its just sooo confusing for new users.
