Hey Gary,
Your issue might be related, but I was asking @sputnik who posted the origin shifting issue here in this thread. I am getting his issue, but haven’t experienced your symptoms myself.
Might not hurt to reach out again to the team via a support ticket. From what I gather they were quite shorthanded back when you started this thread.
I have a solution for the layer shifting that I worked with the team to work out. (Thanks Alice at Snapmaker!).
@gbeeton Although I don’t think it would resolve your problem, it is plausible. And it is a 2 minute job, so couldn’t hurt to try!
The Issue
The Rotary sticks and slips origin on layers (regardless of 3D Printing, CNCing, Lasering):
The Goal
To loosen screws that may have been overtightened, leading to the rotary sticking and slipping.
Step #01
Remove the 5x Screws on the back of the Rotary Unit
Step #02
Loosen the 4x Plate Screws slightly and evenly. I probably only did mine 10-15 degrees.
Step #03
Put it all back together and get creating!
I’ve done 19 burn test, 7 of which were larger runs from between 45min and 4 hours. All have been flawless now. Prior to this, I was having an issue every 2 or max 3 runs.,
Hope this helps some others as well.
Interesting. It’s worth a try, but it may be a few days before I have an opportunity to try this out. I confess that it has been some time since I revisited this issue. So I don’t know if any subsequent firmware updates have addressed the issue. But I have not had any further feedback from tech support on the problem.
Gary Beeton
I finally had a chance to do some testing. To see if maybe a firmware change had fixed the problem, I tried running the same g-code file that originally caused me problems without making any changes. That failed just as the original run had. Then I tried backing off the screws and running again. That failed as well.
Over two years and still the 4-axis mode is not functional.
Agreed: The math behind this is truly complex. Rotation and the x and y Axis resulting in hyperbolic curves, that is hefty and definitely nonlinear math.
The dimensions of the drillbit including shank length and diameter as form and length (or better rotary silhouette) of the working part need to be taken into account not to push in too deep to non cutting section of the bit or worse to the collet or drilling through overhanging parts from already milled slices.
Luban does not handle neither of those challenges well.
As a consequence it just cut the snout and arse (including the tail) off an in all other aspects finely carved lion!
Well Dudes: I will print it with a gel-printer.
If you can tell me how to reach their tech department. please tell me.
Unfortunately your analysis is correct, and they admitted so - no idea if and when they plan to fix this. It would be very cool if they’d fix it - it would make Luban a very powerful tool!
You can contact tech support here:
https://snapmaker.formcrafts.com/support-ticket
I have a solution to the groove problem! Alice from tech support looked at my logs and determined that “the B-axis value is too large”. Don’t ask me what that means precisely, but she recommended that I use the Luban slicer (I had previously used Fusion 360) and re-run the project. And that that did the trick!
I also discovered that there is an updated luban-fusion 360 configuration file available on snapmaker’s support site than what I had used:
Perhaps that fixes the B-axis value issue when using Fusion 360? Sadly I no longer have a Fusion 360 license capable of supporting 4-axis machining, so I cannot test that.
On a side note, while she was helping me debug my problem Alice had me test one possible issue regarding the rotary module. It looks like a slight variation of (or perhaps a more accurate description of ?) the solution provided to @CandleFX:
That turned out not to be the problem in my case.
There is indeed a limit to the angular speed of the rotary module at 45°/s - beyond that it skips steps: Bug: Luban allows faster speeds than the rotary module can accept - and Fusion will not know about this limit (neither does Luban as it seems), so you need to do a bit of math yourself - see the linked thread for details.
