EX Rotary / Fusion 360 Post Processor Possible Bug?

, ,
1

First of all I am not sure what is causing this, the post processor, fusion, or a few brain cells shorted out. I am trying to cut some table legs for an angled table, yes I could do this on my table saw, but figure this would be learning exercise for the rotary. There are two issues I am experiencing with cutting this, both come down to the X axis coordinates created in the post seem to be wrong. I was able to machine both front legs and they look fine, however when I went to machine the Right Rear Leg, ( 4th manufacture stepup ), on the first Corner Cut (5th operation), what happens is the bit does not cut all the way through. Instead it ends short of the full width of the corner. If I extend the pass by .7-1 inch or so I am fine, but on the simulation it works perfectly, just in real cutting it does not. And looking at the generated gcode and moving the spindle to where it says, it is doing exactly what was processed. (this issue was present in both front legs as well, but I just put the in the compensation and proceeded.

Second issue is when I get to the 2D Pockets, these are pockets for dominos, these pockets on the first side in the Rear Right Leg setup was totally off. Instead of putting the pockets where the should it seems to have moved to the wrong side, where the X coordinates are off. And really looking at it, things are even what I would expect backward. My rotary is mounted along the Y axis with the motor at the max Y limit and max X. My setup is shown in the image below, along with what it actually cut, like it rotated it. And since I am not in the shop now, I am going to have to go look at the two front legs closer to make sure things are where they should be.

I have attached the entire fusion 360 file of the entire design, the resultant gcode for this setup. Please don’t be too harsh on the design :wink: But if any one can see why this is not cutting right, please feel free to smack some monkey knowledge into me. Any help is greatly appreciated, especially before I give up and use the table saw and then end up cutting it 23 times to get it right there as well…

image
image1084×675 220 KB

angled_end_table_v2 v39.f3d (1.0 MB)

RLL-try3.txt (41.2 KB)

Thanks,

Tim

2

I took a look at this and I don’t immediately see a “smoking gun” that would be your problem, but I do see a few head scratching things in your setup…

image
image1920×1138 136 KB

For some reason your stock on this setup is far away from your source model… I’ve never seen that before, that could potentially be an issue

image
image588×1150 60.7 KB

Another issue that’s gotten me a few times is using “relative stock”. I like to be explicit about what my stock is if it’s a cube or a cylinder. If it’s anything else I draw a solid for it and use that as the stock.

To really suss this out I’d make a VERY simple model of something like a simple L shape or an arrow.. something you could test out to see if it’s cutting mirrored.

Rotary stuff can be hard, but once you get it figured out there’s a LOT you can do in Fusion. This type of model is perfect for Fusion as it really takes advantage of the true 4th axis capability.

3

Well I did say Right rear but really meant I was cutting the rear left. Hadn’t really inspected the right rear for errors yet since I haven’t got there. I was about to do as you suggested and take that leg, and clone it to a separate project and cut it down to a few inches to be able to cut multiple test. But I am out of ideas on what to change on the rear left one, I’ll try the change to the stock.