Sprockets

[bruce]

I am a newbie to Gearotic but not CNC or CAD and I have a few questions about generating a chain sprocket in Gearotic.
I have the program loaded and have selected a sprocket, it's for #25 chain which is selected and the number of teeth is 70. So far so good
I see the drawing of the sprocket and I unchecked spokes so it is a solid disk. The next thing (after saving it) was trying to create a 4th axis
g-code for it but all I see when i click on that button is a target, not the sprocket plus no g-code is created. So what am I doing
wrong here? I don't need the action part of this program, just be able to create the sprocket then generate the g-code using straight CNC
or through the use of a 4th axis.
My next question is the use of the 4th axis for cutting the teeth, I see the sprocket is set on edge vertically and the ball cutter is used to
cut the tooth profile. What about placing the 4th axis horizontal and cutting the profile of that tooth using a plain end mill then indexing to
the next tooth and do the same? Well this one might be a suggestion for later enhancements.
Thanks,
Bruce

[ArtF]

Hi Bruce:

  The best way to cut a sprocket is 2d, not 4th axis, though it can be done. 4th axis is a tangental shaving
technique and sprockets are not involutes so its not the best way to do them. However, to get Gcode,
tell the sprocket tool to "Add to Project" the sprocket so you have it on the screen.
  Now select it in the tree, then press the "Send to 4th axis" button. The sprocket should appear on the screen.

  There are now a few operation on top, rooting , for example, when you press that button, a path will appear. If you press it again, it turns off. You may mix and match the operations. Press Post to generate the Gcode.

  Its similar in 2d cutting, select "Send to Workbench" for that operation, select the sprocket on the new screen, press auto and a toolpath will be generated for the tool you selected. Press Post to send it out..

Yell if you get stuck, the link for videos on the screen can lead you to a video showing how its done..

Art

[bruce]

Ok so I got it to show the 4th axis now via your instructions.
But dang it now I don't remember how to change the cutter specs. I will look through the tutorials.

[ArtF]

Bruce:

Theres a toolbox button where you can specify the tool, and there are overide boxes on the screen to set feedrates and such. Experiment and turn on and off toolpaths to see the differences.. the simulation will show what that cut
would look like , and gives a good hint as to what to expect..

Art

[bruce]

Ok, got it and was able to change tool info.
Thanks for all your help.

[bruce]

So rather than use the A axis I went and used 3D for creating the g-code to mill this sprocket.
However I find that the G-code that was produced does not cut the hole in the center of the sprocket, is there something
in this setup that I need to do so it will do that? This brings up what if one needs to cut some other geometry's on the sprocket face
there doesn't seem any other way to do that other than generating spokes.
I decided to export the sprocket I created as DXF and then imported that to CamBam and from there was able to create G-code
that not only cuts the teeth but cuts the center hole too.
It took me awhile to get the DXF file to show on the CamBam screen but managed to figure that out and is not any issue with
Gearotic.

[ArtF]

Hi Bruce:

  Hitting the automatic button in the 3d workbench will generate profiles for all items including
the shaft hole, it should only skip it if the tool is too large to fit. Exporting DXF's for further processing
though is a valid way to add extra stuff, I do it myself using vectric.. While I had looked at adding drawing
tools and such, it just doesnt seem valid to duplicate so much work of others when there are great cad
programs like Vectric and CamBam out there..

Thx
Art

[bruce]

I managed to take the DXF from Gearotic on a sprocket with 60 teeth and imported it over to CamBam to generate
the G-code. That worked ok and produced the G-code and milled the sprocket, it looked pretty good until I tried
wrapping some #25 chain around it. The chain fit ok on about 20 or less teeth but as the chain would wrap an error
kept accumulating until I was seeing a huge mismatch of about 1/2 the pitch.
I had posted the problem I had with this sprocket on CamBam (supplied the DXF file) since that is where I imported the DXF file and
someone there measured the pitch around different spots. They found the pitch to be .257" not the .250" it should have been, there's
the accumulated error I was seeing.
So what did I do wrong in generating the DXF file for this sprocket? I have attached the DXF file.
Oh I did a 10 tooth sprocket too but since it is smaller the error was smaller and the chain fits ok, or least it seemed it does.
Thanks,
Bruce

[ArtF]

Hi Bruce,

  It looks good to me. The number on the screen show the pitch radius to be
2.3884 inches, if we calculate the circumference, thats 2*PI*2.3884 = 15.0067
for the total, and with 60 teeth, that .250111 pitch. so if the cut was done to
that spec, the sprocket would be out by .0001.

    If I put this sprocket in CAD and draw a circle to the root., it has a radius
of 2.3217 , while an outside radius measures.. 2.4563, ( At least in Vectric
thats what I get. ). Now if I subtract those I get .1346 and divided by 2 is .0673
, which if we add to the minimum radius, we get 2.389, or .0005 away from
Gearotics calculation, which may be my drawing of the radial circles being
inaccurate.  This shows the center of the roller is on pitch at a radius to give
.250111 pitch on the sprockets, I think the .000111 is a rounding error
but would require a gear with many more than 60 teeth to accumulate;

  If you measure the root radius, can you confirm two opposing roots are
at a diameter of 4.6434 or so? Any increase in that number will make teeth go tight,
and lower and they get bound short..

  The large the tooth count the more critical the zeroing gets, but if the sprocket
agrees with the dxf, the root should match I think..

Thanks
Art

[bruce]

I measure root to root diameter of 4.655 inches.

[ArtF]

Bruce:

  Sorry for the trouble. Im trying to figure out where the error is.
With that root, it sounds correct to the numbers generated and to a pitch of .25011
with 60 teeth..

  Ill look deeper into the spec for the #25 to see if i can figure out where the
error is.. Are the rollers sitting flat in the radius of the root?

Art

[bruce]

Thanks for checking into this Art.

The rollers that seat right are sitting flat but as you move away from the good ones the rollers soon start binding on the back side of the tooth.
In this case the rollers start binding against the tooth like the pitch is increasing and not decreasing on
the sprocket. It's hard to describe so maybe a picture would be better way to explain.

I am not convinced it isn't something I am doing with setup or some other parameter.

[ArtF]

Bruce:

Looking at the code, Im thinking it may be me. Im triple checking.. Ill let you know soon..

Art

[ArtF]

Hi Bruce:

  Ive run through the math, and I did find the numbers were reporting a bit wrong  on the screen for root depth,
Ive fixed that, but otherwise the numbers appear to be correct. I also checked them against known formulations
from the table at this link
http://www.gearseds.com/files/design_dr ... cket_5.pdf

  If you use the sample in this pfd for example, you'll find Gearotics drawing identical.

  All I can think of is that the tool  may have been just a touch smaller than entered in CamBams table,
it would take only a very small wear to add up over 60 teeth..

    Take a read through the link, let me know if you see anything you figure I may have missed,
but the formulas on that page are the ones I use. They dont really conform to normal pitch
calculations for spurs , but they do produce the example results to within .001 on all my checks.

  I may be missing something Bruce, Ive been doing this long enough to know that sometimes
the obvious can fool me, but I do have several reports of good sprockets cut too, so Im stumped
as to what, other than a tool size error could cause it.

Art

[bruce]

So I tried to make this sprocket again.
I had ordered a commercial sprocket from Amazon for ANSI 25 chain and 60 teeth. I received it today and
measured it's thickness which was about .090 inches and fit the chain perfectly (It better!), then I placed this
sprocket over the one I had made a couple of days ago. Wow it was from eyeball measuring almost a perfect match
with no change in pitch I could discern.
So I surmised (incorrectly) that it was the thickness being the issue so I found some stock a little thinner and machined
another sprocket. Well same issue, it was like the pitch changed but comparing it to the commercial one looked good.

I then measured the root to root diameter and on the one I just machined was 4.655 but the commercial one measured
4.635! My end mill measures .125 inches so I am not sure why the root to root is off but what i plan to do now is
tell CamBam that my cutter diameter is .115 inches but use the .125 inch cutter. Should that not reduce the root to root diameter?
Maybe there is a better way to make this change?