Troubleshooting a small job.
I’m trying to figure out what is causing this discrepancy between what I see in the gcode (as simulated on Fusion360 and chilipeppr) and the actual cut. If you look at the pictures you’ll see that some teeth are being cut incorrectly (end mill goes straight through the tooth tip). I’ve repeated this job three times and it runs the same path each time. Does this signify a gCode issue or an error between chilipeppr and the tinyg?
Thanks,
Darren
One of Two things 1 the feed rate is too aggressive and the mill is pulling it’s self into the work try and have slow down in corners… 2 As you cut around the gear you remove material and buy the time you reach the other side it’s too weak to cut. I would cut down on 3 passes is apposed to on single pass. The first 3 passes cut all the material and leave and on the third pass some tabs that you cut off in the last pass
I would recommend that you check your machine for square and everything is tight., check to make sure that each axis moves smoothly without binding. Slow the machine down by 20-percent to see if that solves the issue. Check for racking in the axis that causes the issue. Check your drivers to see if you might have a driver that is unresponsive as an intermittent issue. Hope this helps
If you don’t mind sending the file to me I can run it in a simulator to see if I see anything in the g-code
I would need to see the actual file to help resolve a g-code issue
I’m starting to see a number of small issues popping up with Fusion360 cam users and the Gcode produced. Offer up your Gcode here for someone to run through or try a different CAM. Not saying this is the issue, but there is something with the arcs being noted here and there.
I agree with Brandon. I sometimes had my TinyG not like hsmworks (the back end for fusion cam) arcs. So it would cut the corner. Perhaps the post processor from fusion needs some adjustments.
I’ll upload my gcode this afternoon. This sounds very much like what Brandon and Eric described - gcode is definitely telling the machine to cut straight through when there should be an arc. I wonder if it could be due to using smoothing or feed optimization in Fusion360.
@Darren_Lee are you correctly telling fusion it is an inside cut not an outside cut? And do you have you clear heights set correctly for linking moves in fusion?
It’s set as an inside closed contour (it’s a ring gear profile). Three passes at 2mm each and a final finishing pass. I have a separate gcode file for the outer cut to remove the gear from the plywood. It also uses 3-pass @ 2mm depth, but with tabs to maintain hold on the material. No issues with the outer cut, but it’s just a circle.
Looking at it again, there is a regular pattern to the odd cuts. Evenly spaced in four sets. Also of note is the inner-most toolpath cut correctly (remaining wood core has no shorn teeth), so it is deliberately removing those tooth profiles.
Here’s the Gcode for the teeth cut
I have ran several simulations with different simulators and controllers and I can say that it is not your g-code that is causing the problem. If it were the g-code I would have the same issue that you are having no matter what machine I used. In this case I believe it one or more of the following. Post Processor, interpreter, mechanical and/or electronic.
Thanks for checking the gcode. I’m guessing there are some are some moves that the tinyg does not like in the code, I’ll try to redo the toolpath with different settings on Fusion and see if things change.
Download Camotics here http://camotics.org/download.html and run a simulation. Camotics is free and I have used it a long time. You can also run a simulation with Mach3 and I believe you can run a simulation with Easel if you upload a compatible file. Your g-code is missing a compatible post processor so I would start there.
@Darren_Lee shoot over your DXF, tool, and depth, dive and cut speed and ill create a Gcode from CAMBAM and isolate this fairly quickly.
Just imported into CamBam and verified the design, it appears that the file is clean.
@Tracy_Ranson can you dump that file here and see if Darrin can run this?
The toolpath simulations in both fusion and chilipeppr look fine. Based on some of the history of tinyg that I’ve googled I’m wondering if there’s some arc commands that it doesn’t like. I did use the Tinyg post processing option in Fusion but I am not sure how good it works. I wonder if I use the grbl-adapted selection in Fusion it might avoid the issue
FYI I was only cutting this for practice on the new machine. Quite new to diy CNC. Interesting issues like this are why I’m practicing in cheap plywood 
