Surface Milling help needed

[Jcadwell]

I'm trying to mill a mold to make two halves of a R/C glider body. I built the surface model using Rhino3D, saved it as an IGES, and used Mastercam 9 to write the CAM. I'm using a 1/4 flat end mill for roughing, and a 1/8 ball for finish. The roughing pass is pocketed at a step down of .10", and the finish pass is scalloped.

The code (linked to below) was posted for Turbocnc using the TCN324B post available for using Turbo. I'm using a Taig Minimill with steppers, which normally works well.

Now for the problem... Notice in the code I posted there is a comment that says "(problem starts here)". At that point the cutter deviates from the expected path and cuts through the mold wall and to the minimum Y limit (less than 0 on the readout), where the motor then steps out. Then the cutter moves along the X axis (in the negative X direction, past X0.00) until the X axis leadscrew unscrews all the way. It is almost like the mill is trying to cut a giant arc, and never understands to stop. The mill then continues making X and Y movements, but since the X axis is no longer attached the machine just cuts air.

I've tried verifying this code with both Mastercam and Deskam, and the problem didn't show up.

Any thoughts would be most appreciated. The files are available for download at

ftp://cnczone:setfree@www.ppcadcam.c.../New%20Folder/

Please reply here or to jcadwell (@) gonzaga.edu

Ignore the parantheses around @. It helps keep me from getting junkmail if I don't type the entire address as one block.

[Jcadwell]

There is a picture up as well. Thanks.

[ger21]

Bad Link

[HuFlungDung]

Hi JCadwell,

I took a look at your file by importing the backplot into OneCNC. I actually do not see anything like that movement in your code. So I think you most likely have suffered a DNC signal glitch. That is what it looks like to me.

Check that your DNC cable (if applicable) is properly shielded with only one end of the shield grounded. The other end should float. This acts as an antenna to drain radio interference off.

The shorter your DNC cable the better. I had glitches using a 50 foot cable once. I changed to a 6 foot cable and have never had a problem since.

[HuFlungDung]

Here is a simulation shot of the machined surface. No gouge here

If your problem is reproducible, then I suspect maybe your controller software is having a problem. You might try regenerating the program, using I and J coordinates with your arc settings, instead of R. There are times when R values can be ambiguous.

[Jcadwell]

Definately reproduceable. I let it mess up three times at the exact same spot. I don't thinks it is the cable. It's short, and I haven't ever had any problems before. Especially not reproduceable ones.

How would I go about using I and J values to replace R?

I may try running with a different control software to see what happens. Thanks.

[HuFlungDung]

You need to reconfigure your Mastercam post to output arcs with I and J instead of R. Sorry, I don't know how you make that adjustment, but there is lots of help here who would.

[Jcadwell]

Ok. I'll read up about that. I rewrote the code to use a different roughing method so hopefully it will work.

I was also running the code off a floppy. Copying to the hard drive made things speed up dramatically, and may make a difference. Thanks for the help, hope it works, John Cadwell.

[cadcam]

Ok so you are using the MPfan post am I correct?

And here is a picture of what your Gcode is outputing not prior to gcode.
Picture showing some of the code in backplot

review this.
Now send me your MC9 file and I will check it and repost it with IJK instead if you like.

[Jcadwell]

I am using the TCN324B to output to Turbocnc as the machine controller.

I got the code to run by rewriting the roughing pass and running from the hard drive rather than the floppy, so the molds are now made. I just don't know why I got weird outputs. Every time I ran the verfiy routine in Mastercam or the Deskam simulator I got outputs similar to those run by you guys... Just makes me scratch my head a little.

[cadcam]

Every time I ran the verfiy routine in Mastercam or the Deskam simulator I got outputs similar to those run by you guys

As you make this statment, Mastercam does not show you what you are getting from the post but prior to the gcode that your machine will see.

So This happens with all cad-cam systems. If you have a old or bad post and it looks greate in say MC Varify does not mean you will get the exact same info to the machine.

So what I was showing you was your Code you were sending to the machine.

So I say the trouble is on the machine side or as was stated earler the machine might have a problem persay with "R" call out instead of IJK.

I just programmed a project for a customer that had an older style Horzontal CNC were the control would not keep the G02 modal and would gouge the part unless the next line hade it actully coded line per line.

So hope this helps with this long ramble.

[Jcadwell]

That does make sense. Not sure how to edit the post to make it work any better, but at least I know to run the G Code through another verifier, which I can do. I did that with Deskam and it didn't show any obvious errors. I'm stumped, but I have the part I want... Wish I knew why.