Ran a rough and finish path using the above method, and using linkage, to carve this mouse from the game Root. Tool path wasn’t perfect - cut the top of one ear, don’t know what it was doing on the feet - but I thought the detail was very good for something so small. And very little fracturing. I wasn’t sure the sword was going to make it through intact. The much larger FDM version was printed on the Snapmaker using the dissed stock black filament.
1 Like
That’s not too bad, maybe a bit finer of a stepover would help the fuzz, or slowing it down just a tad. (I don’t know your settings) but very surprising the sword and ears survived at all, I would have expected them to pop along the wood grain lines, especially the ear. The sword has the benefit of going with the grain.
As far as the weird path, yeah… I don’t know what Luban randomly does, I wish there was a better, free 4-axis option. Kinda like the ring I did above, just… cuts right through the model for no reason.
I have problem, after couple succesful attempts Luban is now suddenly incapable of generating second toolpath, it always starts from the absolute beggining, could somebody do quick video tutorial for the process? Thanks
I can’t believe how little support there is from Snapmaker on this topic. I appreciate the tutorial above but this should be a build in feature not something we have to hack together. After spending well over $3K I really expected to be able to swap a bit.
Very good tutorial. I’m wondering if you have any suggestion for multiple finishing passes, without carving too much air? I’m looking to do carving in aluminum. Roughing pass works well (except for the aforementioned random gouges - which have sometimes broken a bit (as they don’t seem to adhere to work speeds either - or the combo or rotation/plunge doesn’t), but then, due to being harder material I still need to do multiple finishing passes in the order of 0.1mm stepdown. A single finish pass would certainly want to carve too much aluminium in one pass. Do I need to create multiple toolpaths, one leaving 1.5mm of material, then 1.4mm, 1.3mm, etc.? or is there a way to simulate/fake “stepdown to the roughing height, then 0.1mm”? Carving air is apparently 900 hours of wasted time, so cheats or ideas from others would be greatly appreciated.
Multiple toolpaths would be the easiest, just changing the remaining like you mentioned. However, you SHOULD be able to merge them all together (if you’re not changing bits between finish passes.) I have a project running in Luban currently, but once it’s finished, I’ll do a mockup run for some examples. I haven’t messed much with the rotary module recently, it’s been hectic, but I want to do more with it.
Cheers. Yeah, I’m doing just that as a test run right now. That B=90 has caused me issue with tool changes where it “magically” decides to mill 90* out from previous. Not sure if you’ve seen that?
Also - multiple toolpaths seem to just “shrink” the model on each pass, meaning that narrow parts still become narrow before wide parts are machined, meaning that tailstock still isn’t useful, etc, whereas what I’m hoping to eventually achieve is to take (for example) 1mm off the OD where needed on the first pass, then 2mm off OD, etc, so that if the original OD was 40mm, the thinnest will be 40, 39, 38, 37, etc., meaning stock is as this as possible, and when it gets thin, it’s only going to be machining those thin sections, meaning more of the stock is held rigid for longer. Let me know if that doesn’t make sense and I’ll try and create an animation to explain it.
many thanks for your guidung steps, Skreelink.
But one thing is unclear to me. How did you calculate the stepdown of 17.5mm in the finishing part
You need your finishing stepdown to be half the diameter of your stock so it’ll make one pass over the surface instead of doing a ton of air cutting and just wasting time.
Thanks @Skreelink for your guidance! Helped me a lot to understand the Luban-Approach. I received my Rotary module today and jumped right in. My first attempts were failures, because I tried to mill too brittle features with too aggressive strategies. So while I was surprised that Luban is indeed able to calculate the relatively complicated “Linkage” paths and does a pretty good job there, I was disappointed that there is not really much to be done about the milling strategies. The title of this thread would better be: Two-Pass rotary guide, because it is either milling tons of air, or have one roughing pass and one single-pass surface pass…
So long story short, I try to understand what other options are there, and learned by now that there are basically three rotary strategies:
- Indexed milling: You mill from one side using 3-axis strategies, then turn a bit around the 4th axis, and do the next side until you have your result
- XZA: Basically what a lathe would do, i.e. work piece rotates and milling bit operates in X and Z (Y and Z for Snapmaker) , so it moves up and down and forward and backward → Problem: you do not reach all parts in complicated geometries
- XYZA: This is what Luban calls “Linkage”. The workpiece rotates, and the milling bit operates in X, Y and Z, which allows it tor each anything that 4 axis can do.
I looked a bit around for software that can do this and is affordable, but with limited success. I tried the DeskProto trial - this can do XZA and indexed, but it is still somewhat limited. All other stuff I ran across is expensive like hell, or 4th axis is still in an experimental state…
Would appreciate some hints!
1 Like
Good trick but have in mind that after the 1st job at some points maybe more than 1 mm material is left to cut. And this (depending on the material) maybe will be too much for the VBit in one step.
Anyway you did a great job with your description 
@Skreelink, you state V: GCode Optimization (Optional, but recommended), is optional. If i choose not to install this, is there anything else I need to do to speed up processing time? I’m new to 4th axis milling and can’t stand the long carve times.
I actually purchases Vcarve desktop to be able to do roughing passes but I cant get it to work correctly with my SM. It wants to plunge deep into the workpiece and then start turning the spindle. I found a SM post processor for Vcarve but that does the exact same as the original PP for Vcarve. So I would love to be able to decrease milling time. Thank you in advance for your help.
Tony
It’s completely optional, mostly to reduce noise (although it does mirror the output, so you’ll have to compensate accordingly).
I do not have any experience with vcarve, however, check the generated gcode file and see where it puts the M3 line to turn on the spindle. It may be placing it after the initial movement instead of before it.
@Skreelink, I must admit - I know nothing about editing PP or G-code generating. Do you know of a source that will teach me how to develop my own PP? I’ve attached a snapshot of my PP and the M3 code but it’s all alien to me. #clueless.
In the header you sent it moves to Z10 first, which will be 10mm above your set origin. If your origin is beginning in the center of the object, and it’s larger than 20mm diameter, then it’ll try driving into the material before starting the spindle. I don’t see why that movement is there, except maybe it’s expecting you to have set the origin at a flat surface, then it’ll move a safety 10mm up, but that only works on a 3 axis setup.
Make a backup of your post processor and then remove the "G0 Z10.00 F600" line and try again.
1 Like
@Skreelink thanks for your help, unfortunately, I’m so new to G-code that I don’t even know how to remove what you just showed me. I guess I will research YouTube and try to figure it out. I really appreciate your help. Have a great day…
Sorry I couldn’t be more help.
@Skreelink, your help was priceless. Your recommendation was correct and I was able to delete the recommended G-Code. Now Vcarve works as it should. I don’t feel like I wasted $350. Again, thanks for you’re help.
Glad to hear it worked out.
GCODE is finally a long list of commands sent to the printer doing something (move, print, extrude etc.).
GCODE files can be edited by a text editor like notepad.
Make sure that you are not using any word processing software (like Wordpad or Microsoft Word) because they will add additional formatting to the file which the device doesn’t understand.
I can recommend notepad++ on Windows, it offers additional comfort while searching and replacing text, and it can open large files as well.
For the particular case you need to use the “search” functionality to find the part in the GCODE file.
Instead of removing it you can also comment it out giving you a backup opportunity in case you made a mistake