Not following the toolpath

[Ken Alexander`]

I set up another job tonight; to cut some holes in a piece of wood. I want to make sure this will work before moving on to carbon fiber. The holes were centered side by side on a piece of wood 120mm long (X), 50mm wide (Y) and 22mm thick (Z). The previewed toolpath looked fine. It imported fine into DrufelCNC.

0, 0, 0 was set to the lower left corner of the wood. I ran the toolpath and although it started making a nice cut it was making it about 1.5 cm to far up on the Y-axis. The design called for it to be centered on the wood piece. A few seconds later it lost its way and started cutting up and to the left of where it began. The cut also got very sloppy as you can see by the rough edges.

I aborted the job and then sent the spindle back to 0, 0, 0 to see where it would land. It looked it was heading in the right direction, then it took a turn and dragged the 1.5 mm endmill across the clamp and snapped it...which was when I noticed that the spindle wasn't raised sufficiently to clear the clamps.

I can't think of a reason why it would lose its way like that. At first, I thought it might have been some weird metric-to-imperial conversion that took place, but the hole it was starting to cut was the proper size. I guess it's time to learn how to read G-codes to see if I can find a problem...which makes me wonder if there's a visual way to step through (and back) the G-code line by line to follow the tool on the preview?

Once again, any help will be gratefully received!

Ken

[TReischl]

OK, from my past experience and that of pretty much everyone on this forum you will discover that if the piece previews correctly in the software it means you are having machine issues. We all have a tendency to point our fingers at that which we do not understand. It does not hurt to be able to read g code. In fact it helps to reinforce the idea that the software is producing the correct code.

My short list of what to look for:

Loose connections on the drive train, couplers slipping due to loose set screws, etc.
Static electricity generated by dust collector hose.
Goofed up thumb drive if you are using one.
Banging into hard stops on the top end of the Z axis.

I am sure others will wander by shortly and add to that list.

One piece of advice, if a piece is not copyrighted, post your .crv file or add it to something like OneDrive, Google Drive, DropBox, etc if it is too big to include in your post here. That allows all of us to actually see what is happening.

Edit: One other thing. . . IIRC the machine you are running uses Nema 17 motors. It is real easy to push those too hard and lose steps. I have been thinking about getting one to do tiny stuff and have been watching the vids. That machine runs really slow because of those motors.

[Ken Alexander`]

Hi,

Thank you for the suggestions! I had better results today after slowing the feed rate way down. Still learning about chip loads, etc. and see that the default settings were off the charts. I also set my origin as the center of the work instead of the lower left edge and the router seemed to like that a lot better. At least it returned to the same place when the job was done.

The machine is new and the connections are tight. No dust collector...haven't made much dust yet but I suspect my wife will insist on one soon! All work is done on the same computer, and so far the projects are small enough that they can be placed on the table in the middle of X and Y axis travel.

I tried a few simple projects this afternoon so that the G-code would be manageable if I wanted to look through it. No need to at the moment. The work done in the attached photos was cut with the remaining stub on my broken 1.5 mm end mill! Now that my confidence is coming back I might try a proper end mill again!

Thanks again!

Ken

OK, from my past experience and that of pretty much everyone on this forum you will discover that if the piece previews correctly in the software it means you are having machine issues. We all have a tendency to point our fingers at that which we do not understand. It does not hurt to be able to read g code. In fact it helps to reinforce the idea that the software is producing the correct code.

My short list of what to look for:

Loose connections on the drive train, couplers slipping due to loose set screws, etc.
Static electricity generated by dust collector hose.
Goofed up thumb drive if you are using one.
Banging into hard stops on the top end of the Z axis.

I am sure others will wander by shortly and add to that list.

One piece of advice, if a piece is not copyrighted, post your .crv file or add it to something like OneDrive, Google Drive, DropBox, etc if it is too big to include in your post here. That allows all of us to actually see what is happening.

Edit: One other thing. . . IIRC the machine you are running uses Nema 17 motors. It is real easy to push those too hard and lose steps. I have been thinking about getting one to do tiny stuff and have been watching the vids. That machine runs really slow because of those motors.