Control Skipping Lines and Losing Position Mid-Cut

Yesterday in the 12th hour of a 16 hour 3-D finish carving, Control for some reason lost its XY position mid-cut and ruined the last part of the carving. I’m lucky I didn’t snap a bit or do any machine damage.

Here is the carving before the failure:


After the failure: you can see the left (X edge) is not in the middle of the workpiece in the yellow color wood:


Here are a few details about my setup – more or less a custom 10 x 10 CBeam with custom High Z.

Openbuilds high torque motors and OB wiring, Blackbox 16, water cooled 2.2KW spindle, and brand new Huanyang VFD. Spindle post 4 is grounded too VFD. I also have a copper ground wire going through the dust extraction hose out to a true earth.

I typically use an INTERFACE but on this carve I was using a Windows 10PC with the latest build of Control via USB. The PC had Wi-Fi and Bluetooth off, and windows update off.

I do have an Amazon Blink wireless webcam on the machine immediately adjacent to the BlackBox, and on this carve I had a GoPro shooting a Timelapse on a MagicArm on the crossbeams (with 110V power thru USB cable), and a Canon R5 SLR on a tripod nearby.

So I can begin troubleshooting, what is on the list of possible options that would cause the software or hardware to do this? Is it more likely to be EFI from nearby electronics, or computer software buffer/USB issues?

 

Are the other thousands of users running 100-500MB toolpaths for 16-36 hours that CONTROL can't load into it's editor or preview in 3D?

My ShopBot has never suddenly lost it's X/Y mid carve and skipped to another part of the workpiece, with cameras, timelapse dollies, LED light packs, etc. draped all over it for 36-48 hour carves.

Trying to keep it friendly here, it's hard to read "tone" on a forum....

 

I figured…you have never been anything but more than helpful.

Time to build a faraday cage around it I guess. I’ve got some experience being in those.

 

@David the swarfer thanks for the suggestions.

And yes Peter, my new HY VFD is wired per the specs on Josh's writeup page.

I have gone through all the mechanics and all screws, etc. are tight as far as I can see. The machine was off square from the stress of suddenly cutting in the wrong spot by plunging the bit deep into the wood but I fixed that.

So far I suspect it is either static electricity from the vacuum that somehow scrambled the X position of the cutterhead, or some type of EFI from a nearby electronic item.

Is there anything else you can think of that would cause just the X positions to suddenly be off? The Y and Z positions seemed to be correct, so the code kept feeding out the correct X/Y/Z positions from the lines in the GCODE file correctly, but the X motor moved the cutterhead to the wrong X location....

I am going to add (more) ferrite rings to all lines. And remove/distance all ancillary electronics. And the next time I run a long carve I may eliminate the vacuum in lieu of 10psi of air blowing chips away into the corner to cleanup after the carve is completed.

If I cannot solve this issue, then perhaps my machine would be better off in the hands of someone who will use it to make cutting boards. I am about $3,000 into this at this time as a machine purpose-built to run lengthy 3D carves at very small rest machining and stepover, after roughing on a ShopBot and if I can't do that with confidence then it's time to let it go.

 

• Hose Ground: Yes - 18ga copper wire running through the actual dust hose out to a foundation bolt in shop floor.
• Motors are Open Builds Motors.
• It is "custom" in that some of the 8020 is leftover from an Ebay machine, but I basically ordered every item in the Lead 1010 kit and retrofitted it. All wires, plug ends, motors, mech/drive components, carriages, etc are 100% OpenBuilds.
• Will check GRBL settings after work. I recall I followed your setup directions including speed rates and acceleration during initial build. But will be good to double check.