There is always a lot of confusion around here between us and Uk power.
I have been fighting a disconnect as soon as the spindle is turned on for months. It will only do it if I have a power cord hooked to the computer/tablet. I have tried 5 different computer setups.
All I can say is let me know what you find that fixes it and good luck.
Just an update on my situation, I had to go and get a replacement dewalt router to get my machine at least cutting again. In the meantime, I ordered a shielded cable to connect the spindle to the vfd, ferrite rings for that cable, an emi filter for the outlet to the vfd and usb isolator. One question I have is how you go about mounting the usb isolator or do yall just have it free floating? I have some plastic that I could make a holder with stand offs if that would be ideal?
Always good to have redundancy, especially when it is a critical piece to making revenue for the business! 
Random thought to test, if your carbide motion controller is mounted on the Shapeokoās frame (typical OEM location). Try testing if you get disconnect issues with the spindle unmounted, and not touching the frame (pretty sure the motion controller pcb grounds to the frame too - donāt know for certain anymore). Could help determine the parameters that you get your disconnect issues with out making your machine inoperable again and verify if any of the additional components bought help.
I was thinking the same thing! I wont be able to have any down time until the weekend of the 12th at the earliest unfortunately but i will definitely keep yall posted as I hope all this helps someone else!
You guys donāt isolate your controllers? With aluminum fixture plates and steel core belts its pretty much necessary.
How do you go about āisolatingā your controllers for your machines?
I put on a piece of 0.125 closed cell foam in-between. Controller heat management isnāt effected even after leaving it on for days.
Glad the Dewalt is working for you, no emi issues?
Do you put the foam between the PCB and the enclosure or the enclosure and rail?
I have the aluminum bed and have not had any disconnects but I have not put steel core belts on yet.
Do the bolts not connect the board to the rail, grounding it to the machine? Im sure this is not the case, so pardon my ignorance.
I have had disconnects with the dewalt router since Ive owned this machine, before any upgrades were done. I did add a grounded power cord to the router which attaches to a bolt on the housing but I havent done this to the new router. Other than running the new grounded cord outside of the drag chain, i havent tried any other methods to minimize my disconnects as they, most of the time, were not that big of a deal to rehome the machine. Some days I have zero disconnects and sometimes its after almost each tool path is completed. It never disconnects during a toolpath, always after. (the spindle would disconnect anywhere from when the spindle was turned on to a few minutes into a cut).
Thatās really interesting, if your controller PSU passes through ground the way the EU ones do this suggests that the Shapeoko frame is working as an antenna for electrical noise to the PSU ground.
Antenna or not, it works and is rock solid, and thatās good enough for me.
Controller and computer plugged into one circuit, router/vfd in another, no problems in multiple locations. Spaced controller allows for V1 probe use with aluminum plate and conductive materials.
Ah, so the machine frame is left floating and completely ungrounded, what do you do on your machines with a VFD spindle?
With the Dewalt if you leave the speed turned up and turn it off you can get disconnects but if you turn the speed all the way down before you power it off it is less likely to happen.
Is the board attached to the frame of the shapeoko?
Its been awhile since Iāve used a spindle on an S3 but it worked the same. Mostly relied on the anodized mount/z to sever any connection. The 1.5kw ive run ungrounded for awhile (also in the nomad), no issues (not the safest, but still).
Soft mounted, so frame/foam/superglue/controller heatsink.
@quicky06 Might be worth you giving this trick a try on your interference problem if itās still there.
Thanks Iāll add that to the list but Iām not a fan of un grounded machinery.
Iām currently searching for a grounded powered usb hub but coming up empty
I may order some nylon bolts so that I can still mount the controller to the rail if it does work that way.
Indeed, I wouldnāt run it un-grounded either.
I was thinking more keeping the spindle and machine frame grounded whilst isolating the controller board and itās local PSU 0V from the machine frame. The controller 0V is grounded through the power cable to the PSU anyway.
Separating the machine frame from the controller 0V cuts a transmission path and, I think, is worth a shot.
Iām seeing the same disconnects after my spindle upgrade as described in this thread and others on the forum but may have resolved it. My spindle cable is double shielded and grounded in the spindle (wasnāt from factory, had to solder it on myself). The cable is ran mostly away from signal lines. The spindle is speed is controlled via the PWM pin on the Carbide Motion board and I also have an airbrush compressor triggered off the coolant pin via an IoT relay. I have two feeds coming into the enclosure - the spindle is on one 20A 120V circuit and everything else (motors, control board, compressor, dust collector, lights, chiller) is on a separate one. Itās probably just be coincidence, but the disconnects didnāt happen very often until I changed out the distilled water in my chiller with a 50/50 distilled/anti-freeze blend. After that, it would disconnect almost every time as soon as the spindle turned on.
This evening, I detached the carbide motion board from the Y-frame and the USB connection hasnāt disconnected since. I plan on placing kapton tape between the aluminum plate and frame and printing some nylon screws to attach it. Iāll report back if the issue is fixed long term.
