Autodesk Fusion 360 Posts - 4th axis works, moving on to 5th axis

[CrossfireX]

Hi John,

Yes, I did Air moves above the parts before cutting the part and it looked OK to the naked eye, it wasn't until I lowered the Z and let her rip that I saw the problem, then I thought I fixed it, did some more air cuts on the other side, looked better, lowered the Z, same thing. I'm going to try Inverse time this morning once I get the current job off the machine, and see if it makes a difference.

I went to the trouble yesterday of writing some manual code to just do a single sweep of the line from front of stock to back with only X and A movement and held a sharpie against the spindle to see the line and it still didn't look right. and I think the issue is, if I program it to move from X0.0 Y0.0 A29.8 to X-140 Y0.0 A-29.8 F350 then the A rotates a lost faster than the x moves. or appears to. I need to convert the mm/min into degrees per minute and use a different feedrate on the A than on the X. Or plot points on long the line and give it specific point to target along the path, which will minimize the error margin.

[cncsnw]

if I program it to move from X0.0 Y0.0 A29.8 to X-140 Y0.0 A-29.8 F350 then the A rotates a lost faster than the x moves. or appears to.

That move should make X move -140mm, and A move -59.6 degrees. It should be done in linear interpolation: X and A movement in constant proportion to each other. The feedrate combines mm and degrees.

When you program that sequence, and it is done moving, does the DRO show the A axis at -29.8 degrees?

[CrossfireX]

HI Marc, I'll check, I've just got some jobs on at the moment, once they're done I'll get back onto this. I did try inverse time this morning, and it's still exactly the same. I notice that the slot comes out nearly 5mm wider than it's supposed to be. I'd really love to get this sorted, because If I can get this right, they want 180 units per month ongoing. this is a prototype that they want to put in a robot welder and weld a carbide "tooth" into the slot.

see the pics below showing the difference in the part. Ignore the blunt end mill, I want to run the trial and error with the old blunt one before I change it out.

[CrossfireX]

if I program it to move from X0.0 Y0.0 A29.8 to X-140 Y0.0 A-29.8 F350 then the A rotates a lost faster than the x moves. or appears to.

That move should make X move -140mm, and A move -59.6 degrees. It should be done in linear interpolation: X and A movement in constant proportion to each other. The feedrate combines mm and degrees.

When you program that sequence, and it is done moving, does the DRO show the A axis at -29.8 degrees?

Hi Marc, I just ran a quick trial again while I was changing out a job, and this is what I found:

I stopped the feed at the end of the slot and the angle (on the screen) was correct as commanded, but I noticed the Y axis was +30mm (and and change). I looked at the code and realized that's what it was calling for, so when it looks like it's over rotating, it only looks like that compared to the tool which is due to the Y axis movement (when I ran the manual test I had set Y to 0.00 so this is why the tool wasn't in the milled slot). But this is on the test where I used inverse time, i'll have to try the previous test to try it out and see what it does. I'm not sure if the I'm not sure if the CAM package is making the error, or the Post Processor. I know people use Fusion 360 for this stuff all the time with success, and I know people use the Centroid Post Processor all the time with success, and on the fusion forum, other people have simulated it without issue, and My simulation looks correct too, but it's not correct in real life.

If this is an issue with the Cam Package, I'll happily just go and buy a better CAM Package, if it's a problem with the Post Processor, I'll find someone to write me a Post processor that works. I believe it's not a problem with the machine, because the Machine has been working 100% perfectly. I mill a lot of weld preps for a customer, and sometimes they give me a weird angle they want me to machine, and I know I can dial it up in the A Axis then rotate it however many degrees and it's always perfect. I'm positive the issue is in the CAM or Post processor.