Skipping Lines of Code?

[mrudisill]

We recently setup our machine and everything seems to be running smooth. Proper traveling, no binding, etc. 

Speeds and feeds I am confident with, but there's definitely problems with the pathing. There are some areas it is cutting where it shouldn't be, and some where it isn't cutting and should be. I don't think there is any skipping, otherwise all the geometry would be offset after a skip (like what happens regularly with 3D printing). 

I am curious if anyone has this issue, or potential solutions. My worry is it may be the board struggling to keep up or missing things, and this is just a 2D tool path, so I assume the problem would be much worse on 3D contouring. 

I would upload images but it is saying the forum itself has insufficient space :/

[MillRightWes]

What CAD/CAM are you using? What post processor are you using? What is the controller?

The boards should be fine with the streaming of the code, assuming no interference. We often cut 2D paths at high rates of speed (400 in/min) and do some 3D work at reduced speeds (about 150 - 200 in/min). The boards should keep up.

Some causes for errors in cutting are the lead screw or rack being misaligned, the v wheels being too tight, or the coupler/pinion gears being loose. Other causes are communication issues, meaning interference from somewhere. These can be mitigated by making sure the router and box power cords are plugged into different outlets, make sure the USB cord is not wrapped in the power cords, trying a different USB cord (maybe with a ferrite bead) or different port on the computer, make sure the electronics box is grounded to the machine frame, make sure and vacuum system is grounded, and maybe trying out a powered USB hub. Running the router power cable outside of the drag chain will help as well.

To check for mechanical issues, unplug the motors and push the machine around by hand with the MV. You should be able to feel any tight spots. With other machines, jog the machine around at high speeds over the complete length of travel. You should be able to see jamming. You can also run a program above the stock, not cutting, and see if you return to zero after the program is over. Any deviation from zero should tell you what axis to focus on, meaning if the machine returns to the right of zero, look at the x axis.

For communication issues, run the program with nothing else running above the stock. Gradually add in other elements for each consecutive run until something causes a problem. You should be able to narrow down where the issue is coming from. Check zero after each run. You can unplug the homing switches as part of the testing as well.

I'm not saying the boards couldn't be the problem, but there are areas we can check before the boards come under suspicion.