Differences between gearotic cutting and Thornton

[ArtF]

Chris:

    The thornton gears look like epicycloidal tooth gears. ( which are fine for a clock , they dont look involute though..),
it may just be the photo that makes it look that way to me. But the spoked gears dont look good. Do they look that way
on the screen? Im thinking one of two things is going on..or both..

1) Make sure you poin distance ( in the options dialog) is set low..to around .001 if you in inches mode.
2) It looks like you have backlas in your cnc table. I could be wrong but it looks like your teeth are hooking to
the left or right depending on the quadrant fo the cut.

  Can you show us a close up of 1 spokes gear so I can get a better idea of what it looks like, I can probably
advise you better if I can see it closer up... :)

Thx
Art
 

[JustinO]

Chris,

Can you tell us about, and post a picture of this CNC router?

Thanks,
Justin

[John T]

Hi Art
These teeth remind me of the rounding I was getting from having the acceleration set too low in Mach 3 - is that a possibility?

John

[BobL]

Hi guy's;

I use to generate similar looking gears awhile back too on my big old cnc table, was backlash in the table itself, as Art mentioned maybe a good idea to check linearity of the table first.

Cheers
Bob
;)

[Chris Barfoot]

Thanks everyone your help is appreciated.

I have attached a close up of the spoked gear - btw it was cut with a 0.8 end mill.

The router is a chinese 6040.

[Chris Barfoot]

The point distance is set to 0.25 in metric.

Oh and yes the thornton and the gearotic are epicycloidal (D'oh)

Thanks
Chris

[JustinO]

The distortion does seem to be related to one axis more than the other. Check for backlash. It could also be flexing or something loose in the structure of the gantry or spindle. You could put your indicator against the bit (with the machine off and unplugged) and see how much movement you get when lightly pushing in the directions of each axis.

--Justin

[ArtF]

Hi:

  Yes, this does look like backlash issues. One side is different from the other. You can see it in the spokes as well.
Id set the point distance to .01 or .005 , it will maxamize your resolution. In Mach3 ( if your using Mach3, ) make sure the acceleration is as high as the table will allow.  That will minimize distortion from CV issues.

  If none of that works, youll have to seek out a way to reduce the backlash in your axis. ( looks like one ( Y?) is much worse than the other..)

Art


 

[John S]

Those 6040's whilst being quite well spec'd out do have issues.
One that probably relates to this problem is the bearings on the ball screws.

Check to see that they haven't just crammed a bearing into a hole in the alloy profile instead of using a bearing block.

[Mooselake]

There's a long topic on 6040s on CNCzone.  Perhaps there's something there that will help you tweak your machine.

Kirk

[Chris Barfoot]

Thanks for the advice.

I trolled through all the posts on the 6040 and it seems to be common that it will lose steps which explains so many faults I have had. Became very evident when trying to cut a 1.6 mm gear with a 0.8mm cutter - had the bepth per pass set to 0.2mm and by about the 4th pass it was starting to lose place and cut away the teeth. Anyway decision is made to give it a brain transplant and have orderd a gecko 540 to replace the yoocnc controllers.

In the meantime I have included a photo of where I am up to so far.

Thanks again everyone.
Chris

[ArtF]

Chris:

  Thats a beautiful looking start. Yes, Id also do a brain transplant if I were doing something so delicate
looking on that machine. Ive been very tempted to try my hand on plate brass as well, Ive never really used
metal before for a clock or device, Im usually wood or plastic, seems more forgiving, but I do think brass is
the way to go.... After seeing yours Im very tempted myself.. good work!

Art

[BobL]

Chris;

Awesome detail, keep it up.



Cheers
Bob
:)