GearHeads Corner

Hi

I am trying to design a planetary (inside) gear and GCode output as follows:

Circular Gear Module
Round Gear (Planetary)
Spur
Module 1.0286
Pressure Angle 20
Tooth Count 21
Involute

GCode Processing
2.5D Profiling
Tool Dia 1mm
2.5D Depth 2.8mm
Depth per pass 1
Spindle Speed 7000
Feedrate 1

A simulation of the GCode shows one tooth is deformed, the GCode appears incorrect in three places (for each pass) as follows:

          X 1.6293  Y -9.1360 Z -1.0000
          X 1.6430  Y -9.1325 Z -1.0000
          X 1.7666  Y -9.1035 Z -1.0000
          X 1.8810  Y -9.5902 Z -1.0000
          X 1.7666  Y -9.1035 Z -1.0000
          X 1.8901  Y -9.0745 Z -1.0000
          X 1.8901  Y -9.0744 Z -1.0000

          X 1.8901  Y -9.0745 Z -2.0000
          X 1.8901  Y -9.0744 Z -2.0000
          X 2.0137  Y -9.0453 Z -2.0000
          X 2.1282  Y -9.5321 Z -2.0000
          X 2.0137  Y -9.0454 Z -2.0000
          X 2.1373  Y -9.0164 Z -2.0000
          X 2.1373  Y -9.0163 Z -2.0000

          X 2.1373  Y -9.0164 Z -2.0000
          X 2.1373  Y -9.0163 Z -2.0000
          X 2.2609  Y -8.9872 Z -2.0000
          X 2.3753  Y -9.4740 Z -2.0000
          X 2.2609  Y -8.9873 Z -2.0000
          X 2.3844  Y -8.9583 Z -2.0000
          X 2.4142  Y -8.9521 Z -2.0000

The Y values 9.5902, 9.5321 and 9.4740 appear to be incorrect.

Pat

Hi Pat:

Can you drop the project here with the toolpath in it. Ill run a check to see why those two values seem high...

Thx
Art

Art

Project file attached (hopefully).

Pat

thx pat.

Ill check it out..

Art

Hi Pat:

I think the GCode is fine ( looks that way here..) but the simulation was wrong. Your gear was fairly small so its resolution was so tight that the simulator was moving too far between interpolated points. This made the odd tooth look wrong and showed an asymetric gouge at the tips. Redownload and try it again, the simulator will now run to the
actual resolution of the internal models. It no longer shows the aberations. I think the numbers were fine, but the sim was just wrong .. :)

Thks
Art

Hi Art

I've loaded the file I posted here and re-generated the G-Code from that.

The tap file has the same numbers as reported earlier eg

          X 1.6293  Y -9.1360 Z -1.0000
          X 1.6430  Y -9.1325 Z -1.0000
          X 1.7666  Y -9.1035 Z -1.0000
          X 1.8810  Y -9.5902 Z -1.0000
          X 1.7666  Y -9.1035 Z -1.0000
          X 1.8901  Y -9.0745 Z -1.0000
          X 1.8901  Y -9.0744 Z -1.0000

The value Y-9.5902 is clearly wrong and this is before loading to a simulator.

I have single stepped the simulator and the mill bit jumps when tracing through the Y-9.5902 value. There are three of these jumps which cause the gouge at one tip.

So, I don't really understand your point that the simulator is causing the problem.

What should I redownload?

Pat

Pat:

  If I simulate a 1mm tool cutting that planetary,. it sims properly with no gouging. Before the fix the sim clearly showed a gouge on
several teeth. Th fix was a fix in the sim that was making mistakes, not the toolpath code..
I cant see any bad numbers in mine, I even ran the file though Mach4 and couldnt see any gouges or illegal numbers. I cant tell
any other way that the numbers are wrong. The numbers such as the -9.507, ( which is .5mm jump in one spot )  may just be the tip of a
path ( that 1mm makes a sharp point in the top of the tooth root as a matter of fitting to the space, so its not clearly wrong to me,
but its a matter of context, I need to see the whole file and how it all fits, that Y-9.5902 could be proper.. )

Can you post the entire Gcode file for me, Ill run it and see if I see a gouge.. but I cant see one from my tests, but your code may be different
from mine...

Art

Hi Art

Thanks for taking the time to look into this.

My code is attached, note that my zero is in the centre.

Pat

Hi Pat:

  Ahhh, better. I now see the artifacts you speak off. Weird.. I seem them only one 1 tooth, but one all three passes so 3 aberations..
all on the same tooth. I will look into this to see why. This is a bug for sure. Thx, Ill notify you when its found ..

Art

Hi Pat:

  I cannot figure out how your getting that artifact. Heres a photo of what I get when I post GCode for that project gear..

Im thinking somethingnis different between our settings..

Can you

1) Create your toolpaths in workbench , then save the project. The workbench should save as well so that when I open it I see
your toolpath. Post the project file here as you did earlier..

2) Tell me any settings liek depth per pass, and # of tabs, theres obviously a bug, but I cant yet reproduce it so I need to figure out what causes
that aberation. Are you in Metric or inch for example?

Thx
Art

Hi Art

Attached is the project file after creating the toolpath. On reloading I can't see the toolpath.

The steps I take are:

start gearotic 2
open Planetary.gth
click Output Options/CNC
click Project Tree/Shaft/Planetary--Driver
click Gen Paths

Artifacts should now be visible. there are three on one tooth and repeated for each pass
(5 here because I haven't changed the thickness to 2.8)

I've also attached a jpg of the window showing the artifacts.

Hope this is enough to reproduce.

Pat

Hi Pat:

  Ahhhhh,  a light bulb goes off. Sorry for the confusion.

Can you upgrade to the DEV version of the software. Its CNC functions will not have that problem.
I will look to updating the release version to get rid of that bug. The DEV version has a much improved
CNC section with a built in simulator for checking such things. (Its why you didnt understand my
comments on the sim. or the workbench being saved.)

Update to DEV version, then , after opening the program, click the large question mark on top menu, then
watch the video on the CNC Workbench for how to use it. Its toolpaths will not have that bug in them.
(this is why I couldnt reproduce it.) Yell if you run into difficulty, Ill help you as you go..

Thx
Art

Hi Art

Have upgraded and have used the workbench to produce the gcode and simulated it - no sign of the jumps.

Many thanks

Pat

Pat:

  Thx, sorry for the confusion,,, my fault, Ill update the Release version shortly to the DEV version..

Thx
Art