I’m attempting to machine a fairly simple design in a 12" x 12" x 1/4" sheet of PVC. My Shapeoko on the first attempt got about 40% into the design and then took off in a random pattern of its own. I’ve since made two more attempts with the same result, at some random point it looses its mind and goes random. I don’t see any problem with connections or the driving computer. I have a theory that maybe a static charge building on the PVC is the source of the problem. Has anyone experienced a similar issue? Is PVC something I should avoid?
I have no experience with the Shapeoko or with PVC but just a suggestion: try running the machine with the same G-Code without the PVC. Just let it cut through the air and see if it still misbehaves (maybe remove the spindle-enable line from the G-Code or just switch the router off, I don’t know how it works with Shapeoko).
Also, are you sure the movement is truly random? For example is the randomness on all of the axes or just X/Y? Does it respect machine limits or does it crash into the sides? Is it possible that it’s trying to run the original G-Code in a different place because of lost steps?
On the first and second runs it went pretty much random in the X-Y directions. On the last attempt it drove down through the work into the waste board before I shut it down. Not sure if it’s truly random but definitely not per the design. I don’t have the problem when working in wood. I’m using the Dewalt router and I notice the router bit is isolated from ground.
Some plastics can generate a lot of static charge when it’s cut. I’ve had them totally bung up my table saw with clogs in the 50mm suction hose and stick to absolutely everything, the saw, everywhere.
Are you using any dust extraction to get that charged dust away from the machine?
Is your spindle actually grounded?
No dust extraction but I do pause the project and vacuum the dust out then proceed. The spindle is not grounded. I know the PVC does build up a significant charge. I’m thinking a discharge may effect the electronics since the board is mounted on the gantry and thus close to the cutting.
If I remember correctly, Julien or Will would be a proper source for this, the Shapeoko isn’t grounded.
If you have a search, there’s a few threads discussing grounding the spindle such as this one.
I’d defnitely look into some dust extraction, especially in plastics you don’t want to be re-cutting chips and melting them onto the cutter, it’s best to get them out of the way quickly. Before I had my extraction boot I just stood over the machine with the shop vac whilst cutting really messy stuff like plastics.
Thanks Liam I’ll check it out.
The vac itself can impact the charge quite a bit as well. It’s not stuff sitting around that’s the problem, it’s stuff moving around.
There’s a big chunk in the wiki about static. Your problem is 100% static related.
Antistatic vac hoses help - but they can be expensive. “A piece of bare copper wire threaded through the hose” in my opinion better than nothing, but not by much. I wouldn’t start here.
humidity helps. Try misting above the board with a sprayer. If you can keep the air and material damp, it will help. This is your best, short term bet with things you likely have lying around (a spray bottle and water).
Power for vac, spindle, and electronics on different circuits can help
High quality USB cable to compute (with ferrites) can help.
A correct ferrite on the dewalt power cable (what’s on there is an anti-theft tag that LOOKS like a filter, but isn’t. If you don’t believe me open it.)
Computer running on battery while sending code/everything running.
BE SURE you run the spindle power not inside the cable track with the stepper cables. Preferably, run it at 90 degrees to a a hook behind the machine and a spring about mid-length to keep it from getting caught in the springenwerks at the router gets moved through it’s full extent. This keeps noise from the power out of the stepper wires and vice versa.
Good practice, ground the makita to the mount - this takes a number of steps, This is not as cheap or easy as it first appears to do right to be. You need to replace the power cable to the router, properly ground that to the mount, and bet sure you have a good ground all the way back to the controller. This takes time, some reasonable solder and modification guts, and the right tools to make the change an measure electrical connectivity and resistance between two points.
Easy, Good practice - open brushes aren’t helping matters any. Aa very fine mesh metal (easiest to find will be brass) screen clamped over the top of the router, then grounded to the mount. This will contain the rf noise from the brushes pretty well, but the mesh needs to be pretty fine, and can easily become an airflow problem if not kept clean of swarf. Almost makes a really convenient place to install a sock filter to help with brush wear due to debris. See other posts for directions on this filter/filter item. Dewalt brush wear A sock to keep metal from the filter from getting in the router, then a filter made from an unwound brass wool pad with a wire attached from it to ground, (better than a steel wool pad), then another sock as an outside filter for dirt.
The problem could be on the carbide controller, or it could be the connection to the computer - determining exactly which requires more detective work, but all of the above are targeting various things on both ends. There are more things you can do, but everything is a little more complexity.
And here’s the RTFF -
More on all this (50+ results from a search for "static control):
I’m purty sure that’s not a real good idea.
I can see a number of cases where it would be a particularly bad idea (ie. dead tree carcasses ), but not this one. I would like to understand why you say so in this case?
I’ve actually had good experience myself with this on a PVC project, it’s pretty primitive, but kept the pcv somewhat more contained, and I didn’t have any static issues once I started.
There’s a pretty big difference between flood cooling with a firehose, and misting a bit on PVC.
A little mist well above the board doesn’t seem to be something of particular concern - you’re trying to keep static down, not actually dampen things. A little moisture in the pvc isn’t going to harm a shop vac in any way, and might actually make it a little easier to keep at the machine instead of fluttering all over the place.
Hay guys thanks for the comments and consideration. You’ve given me a lot of things to consider. I was pretty sure that static was the root of my problem and it looks like the group has experienced enough similar issues. In the short term I like your suggestion of a fine mist over the work to dissipate the charge will work.
Static is so finicky, especially now that the weather is really warming up. Dryer days bring more static. I had all my static disconnect issues totally squashed all winter long, and they just started cropping up again a few days ago. But then I noticed that I had an unused vacuum hose on my dusty garage floor, right underneath the USB cable going to the machine. Once I moved the hose out of the way, the disconnects seemed to stop. So even something small like that can make a difference.
And I want to second the above tip about keeping the router power cable (and the vacuum hose) as far from the stepper wires as you possibly can. 9 times out of 10, my disconnect issues were coming from too much slack in the line, causing the power cable and/or vacuum hose to brush up against the rest of the wiring. As soon as that happens, I get a disconnect. This has a been the single most effective solution for my setup, but of course there are so many variables here it’s hard to say with confidence that it will do the same for you.
Actually I was able to solve the static issue at least for the current project. I was able to discharge the static buildup on the PVC by wiping it down with a damp cloth. I paused twice to vacuum the chips and re-wipe the PVC again. Worked great.
Having not (yet) had the dreaded static problems in my damp British shed…
Has anyone with the static problem tried shielding their existing Shapeoko wiring with some woven mesh?
Also, has anyone tried connecting a resistive static discharge to their X axis and Z axis to stop charge building up on these, neatly insulated by V wheels chunks of metal?
Testing my XXL there’s no connection between the HDZ and the X axis and none between the X axis and either of the Y axes or the main frame. This seems like an opportunity for charge buildup and interference to me.
The Carbide PSU is double insulated (according to the symbols on the label and measurement) so the utility ground is not connected to the local DC “ground” at the USB on the controller or the chassis.
My HDZ has a solid earth because it’s got a VFD spindle in the clamp and a solid earth to the shell of these is just not negotiable, turns out that’s also accidentally providing a resistive discharge route to my antistatic with copper wire dust hose (lucky accident there).
However, I’m now rather inclined to get some 100k or 1M resistors out and provide a ground discharge path for the X and Y rails. This should be enough resistance to not create any substantial ground loop but enough to discharge the static (like an antistatic wristband).
Using a separate (large value) resistive link for each leg in a star grounding system reduces the possible earth loop currents to negligible.
So, has anyone tried this before?
I think you’re on the right track. I’ve done this to electronics many times.
These double insulated tools are that way on purpose. I won’t interfere with that, if only for legal reasons.
Absolutely, I wouldn’t mess with the double insulated DeWalt / Makita type spindles or with the PSU to the Shapeoko controller, there’s a whole bunch of benefit to not having the Shapeoko 0V not being earthed, not least avoiding a nasty ground loop on the USB cable to the laptop / PC driving it, if both had a real earth you’d be having fun.
The Chinese spindles OTOH absolutely are not “double insulated”, not when you look at how they’re assembled and I would not trust one of those with 220+V three phase without a hard chassis earth to it.
I also don’t see any issue with putting a resistive ground onto the Z axis with one of those spindles, they’re insulated so they shouldn’t care, my Z is hard grounded via my spindle and that works fine.
I’m going to stick the resistive grounding of the X and Y rails on mine but I don’t expect to see any difference because it currently works anyway. It would be interesting to see whether this makes a difference for somebody in a drier climate that soggy England.
I recently went thru hell and back dealing with, what I assumed, was a totally demon possessed XL. Well after conducting a demon cleansing ritual, problem persisted, so I figured it must be static/emi that is the culprit. After completling the fixes listed below, I have had zero problem.
Here is what I did:
- Installed a grounded cable to my Mikita router. and routed the cable separately away from other cables.
2.Installed ground wires connecting router, to the holder, to Z plate, to X axis to Y axis (both sides.)
Got an anti-static hose for my dust collection system
Installed Ferrite EMI filters on router power cable, usb cable to laptop, probe cable, bit setter wires (maybe overkill but since i had some laying around…)
Performed a check on my electrical grounding and found that I actually had a weak/loose ground in the main outlet.
I also separated the power to my laptop to a different power plug from that of my cnc and router.
Hope this helps.
So, upon testing my shapeoko here with the Euro spec power adapter, it seems that the 230V power cable earth is directly connected to the 0V DC output and through to the ground plane in the controller board and from there to the mount plate and the Y axis and main frame of the Shapeoko. There is also the pin header for the 0V clip for the bitzero probe which is connected to this common earth.
Somewhat more intersting, the shield on the USB does not seem to share this common ground plane (on my v2.4e board), it would be interesting to know if Carbide is using an isolated USB chipset on the boards to reduce comms errors where there’s grounding problems relative to the host computer?
I’m going to be doing some more experiments on grounding the un-grounded parts of the machine and I’ll post updates on what worked and didn’t.