Fusion 360 No Rapids - Cannot run drilling paths in Carbide Motion now

So,

It seems that the Autodesk plan to make Fusion unusable without a license is going splendidly well, at least for me. Can anyone spot something I’m doing wrong here?

Since the “update” I now get this when I try to start a job in Carbide Motion generated in Fusion 360 using a drilling toolpath;

Here’s the intro to the GCode, with the helpful “we broke it” note from the Fusion team;

%
(Wide End Holes)
(Finishing)
(T9994 D=4 CR=0 TAPER=118deg - ZMIN=-10 - spot drill)
G90
G17
G21
(When using Fusion 360 for Personal Use, the feedrate of rapid moves is reduced to match the feedrate of cutting moves, which can increase machining time. Unrestricted rapid moves are available with a Fusion 360 Subscription.)
G28 G91 Z0
G90

(Drill5 3)
T9994 M6
S2500 M3
G54
M7
G0 X5 Y-86
Z15
G1 F0
G0 Z5
Z2
G1 Z-0.5 F125
G0 Z5
Z1.5
G1 Z-1 F125
G0 Z5
Z1

Whereas a toolpath generated before the “upgrade” looks like this;

%
(Wide End Holes - 4mm Spot Drill)
(Finishing)
(T9994 D=4 CR=0 TAPER=118deg - ZMIN=-10 - spot drill)
G90
G17
G21
G28 G91 Z0
G90

(Drill5 3)
T9994 M6
S2500 M3
G54
M7
G0 X5 Y-86
Z15
Z5
Z2
G1 Z-0.5 F125
G0 Z5
Z1.5
G1 Z-1 F125
G0 Z5
Z1

The Carbide Motion log says this;

So, it seems that Autodesk has decided to sabotage users by setting the feedrate to Zero to ensure that the job takes infinite time.

Editing the bold lines to remove the F0 removes the Autodesk F-You and the job seems to run.

Thoughts from folks who know more about GCode?

1 Like

Here’s an easier to read diff;

???
That doesn’t seem right.
That G1 F0 only applies to that line. It doesn’t make sense to have a zero feedrate, hence the error, but I still see rapids.

3 Likes

Would it be useful if I posted the full files?

Yes. Rocking a baby now, but I’ll check later tonight if no one beats me to it.

Do you have GCode for that task? :wink:

Should I go post on the Autodesk forums and suggest that some GRBL regression testing might be in order for these changes?

Test Drilling.nc (1.5 KB)

Wide End Holes 1 - 4mm Spot Drill.nc (1.3 KB)

1 Like

Interestingly there’s another error generating F0 at the end of the file too (having manually cleared the fault at the start);

G1 Z-10 F100
G0 Z5
G1 Z15 F0
M9
G28 G91 Z0
G90
G28 G91 X0 Y0
G90
M30
%

Other toolpaths seem to work, a contour path just ran fine, if fine includes time to grow a beard whilst it crawled between profiles.

Looks to me like these toolpaths have the rapids as you would want. The retracts on the drill steps and the XY reposition for the operations are rapid. That F0 is weird…what PP are you using?

The key is the G18 arcs, lower acceleration, low feed, and constant engagement.

4 Likes

Same PP as before the Carbide 3D option;

Which suggests to me that Autodesk broke the drilling output in some way that the PP for GRBL didn’t like?

Can you, or anyone, set up a simple file with simple profile cut and an adaptive pocket? Share the Fusion File and resulting CAM operations here. I’m curious…

I was trying to test that for you, but realized the new Fusion release must not have been rolled out to France yet because my rapids are still there.
Version says 2.0.9009, can anyone confirm what version they have, the one that took the rapids away on Oct 1st ?

1 Like

Setting one up at the moment

@Julien as of 10:05 AM PDT; 10/2/2020

2.0.9009
Active Plan: Fusion 360, Personal
Windows 10 (18362

@LiamN To be honest, I think that this is more of “honest mistake plus inadequate QC” than “intentional sabatoge”. Some combination of @fenrus and I are working on it. Do me a favor and reply with an *.f3d file as well as a *.nc with the current post-processed code. We can ask someone else to produce the *.nc with the desired post-processed code later (I have plenty to work off of right now so this isn’t limiting a first draft)

1 Like

I tried that and rapids plus toolchange looked fine in camotics

my play time is often limited to weekends; I hope to have a tool operational some time tomorrow…

(I hope to also have it fix up the weird “F0” thing but I suspect that is short term in nature)

2 Likes

Why’s it always take you so long to complete these complex programming tasks?
:wink:

1 Like

Yep, I’m sure it’s not deliberate, normal developers don’t deliberately break code. On the other hand, if you’re removing or restricting functionality in an update which has already caused a fair amount of negative PR, doing your regression testing properly is expected. It doesn’t seem to be affecting any other toolpath either so I’d assume that there’s a poor interaction between GRBL and the additional “no rapids” new code added by Autodesk which there is inadequate test coverage for.

Thanks Neil for the hints on what’s going on in that code.

There’s no rush from my perspective, other than to document the issue here so that other users aren’t confused if they run into it before Autodesk fixes it, which I’m sure they will, once they’re aware. I’m editing the g-code as necessary to run jobs, it was beyond time for me to learn to read and edit g-code anyway, I’ve put it off for too long.

I’m also not suggesting that we should confuse this bug with the possible rapids post-processor as this is a bug in Fusion or the Carbide post processor in Fusion, not something we should try to bodge over in the output code. Fix bugs in the layer they occur. It’s easy enough to just search for the “F0” instances in the g-code and just put a non-zero number in until it’s fixed, for the small number of people who even use the drilling toolpaths anyway.

I also posted a few workaround approaches for people to use whilst there’s no rapids fixer or for those who don’t want to post-edit their g-code with the attendant complexity.

The “test drilling.nc” file posted above is a current output file with the F0 statements which trigger the GRBL error.

Here’s a simple file as requested by Neil, is it sensible?

Four ops;
A drilling op (with the F0 problem):

An adaptive clear of the internal pocket;

A simple contour of the internal pocket;

A slightly more complex contour of the external sides;

And here’s a zip with the f3d and the four nc files as generated;
Simple Test Paths.zip (140.3 KB)

As built using this Fusion;
Screenshot 2020-10-02 at 18.30.24

So, do y’all think it’s time for me to go post the issue report on the Autodesk forums or are there things I’m missing still?

1 Like

Probably be good to watch (and report) https://forums.autodesk.com/t5/fusion-360-support/bug-drill-operation/td-p/9779575

4 Likes

Thanks Dan,

Subscribed to that and posted that the Carbide processor suffers the same fate.

2 Likes

There’s a ticket with the dev team now.

5 Likes