We can agree it is rare? And also would you rather have it pop off the rail or rip itself apart when that happens?
I like my machines to be low stress. If you screw up, worst case it print a new part of replace an endmill. I never want people to be scared it will self-destruct when they make the inevitable mistake. It pops off the rail you might lose the project you were working on and you will learn to never ever set anything on the work surface again. I pushed a screw gun off the table, I have never done it since. I ran over a cord once, never again, I tried to put the dust shoe on while cutting, nope won’t do that again either.
Yes, we can agree it’s very rare ) Not really a big deal, just something I kept in the back of my head for the next machine I am inevitably going to build someday.
The only real big issue I am having is the bearings not being able to handle the foam dust. On the other hand I understand that LR3 wasn’t designed to mill EPS and it is something I am going to have to deal with on my own.
I did find that I needed to bumpbup the X motor current, but once I did that, everything has been fine for machine power.
The resistance of the tool moving through the work seems to be greater than the resistance offered by the machine kinematics.
For the captured rail, I’m playing with that for the LR3, if I can ever take enough of a break from using it. I’ve had a rail in the machine room since Christmas, lol.
The worst thing to happen to my LR3 stemmed from a disconnected Y endstop switch, and the home sequence took off on only the one side. The machine ripped the belt holders apart. I’ve thought this before, but I’ll be working the build to have normally open endstop switches in order to lower that machine stress. My thinking is this: in a failure mode, normally closed switches read as triggered. This is a good thing if you don’t want your 3D printer to sink a 250° nozzle into your PEI bed, and for a single switch axis like X or Y it just messes up the home location. Generally a betterbfailure mode. But we use many two-switch axes, and some of them don’t deal with a lot of skew very gracefully. X and Y on the Primo getting skewed puts a lot of undesireable force on the trucks, and the tube clips, and exerts bending force on the gantry tubes. On the LR, homing just one Y side does bad things to the belt holders, and isn’t great for the belts. By contrast, a failure mode where an endstop never triggers runs the machine, still parallel into an endstop where one motor stops, and the other makes scary but harmless noises. Things are much less likely to break on the Primo, and the belt holders don’t take nearly the strain on the LR.
For a while I was shipping the Jackpot with the Current super low. I have changed it since.
No that we can set a homing current separate from the usual that seems like an okay idea but This also seems like it should not be needed very often. At least not often enough to abuse the endstops like that every home?
I’ve never checked… Does FluidNC have skew prevention capabilities?
E.g. a deadband where if you detect a homing trigger on one side of an axis you have to trigger or stop within some time/number of steps on the other side?
If not, you halt/throw an error?
I concurr, it shouldn’t matter when things are working properly. However, just like data backups that shouldn’t need to be used, its the failure mode that matters.
Low current homing sequences (which I also use on the Duet boards) help limit the abuse and destructive potential, but I still ended up snapping the belt holders on my LR when a NC switch wire broke. I believe that I would have had no such issue had the failure mode just run the endstop into the stop. The switch and the endstop take the exact same wear and tear while things are working correctly.
One other random thought; when running high feed speeds, there can be correspondingly high accelerations at direction changes.
I worry about acceleration limits on my 3d printer but haven’t given it any thought on my LR3, so I don’t have any idea what the defaults are, but it could
be where missed steps are most likely to occur.
A softer ramp up to 25cm/s might help and wouldn’t add much time to the overall job.
Accel tuning sort of has to happen for the hardest material you cut. Softer material higher accels, harder, slower. You have to listen to how is sounds while cutting a 90 degree corner on a pocket. At least that is the way I kinda do it. The defaults are set really slow, from there you should only go up.
Waiting on steel, for me. It’s a holiday here so I can’t pick it up today. Then I need to go buy some lumber to make a table, which I also can’t do today unless I want to pay Agent Orange prices… and I don’t.
Aside from that, I have a bunch of stuff ready for assembly, and a .DXF to cut struts with. I am going to cut them on the beta build, and not on the LR3, even though I could.
.