A little history on my machine. Shapeoko 3 xxl with a plate steel welded bed, 2.2kw spindle 220 in and out. Double shielded 16/4 cable with the shielding grounded at the VFD. Closed loop cooling for the spindle running through a surge protector. Power for control board running from a stand alone circuit.
Now for the problem, when the spindle is running, the old lost connection error stops me dead in my tracks.
Things I’ve tried, re-routed the power cable away from EVERY thing. Checked connections at the VFD and the spindle. Cut a foot of cable off of spindle end and reattached, just in case a break in the wire in the high flex area. Disconnected bit setter. Checked all grounds. Replaced spindle with a new spindle that I try to keep a spare on hand.
I can run a simple program on air cut with the tool removed and the spindle off, it runs as it should.Try it from the beginning with spindle on and within 30 seconds it stops. Start it with spindle off and anywhere along the program I can turn the spindle on and within 30 seconds…yep, it stops.
It’s been running good until 2 days ago. Just started all at once. Contacted Carbide support, tried what they suggested. Ya’ll folks have any suggestions, I’m all ears. Thanks.
Hi Basil - welcome to the community and all that.
The C3D people have probably covered most of the usual suspects. You don’t mention if you have dust collection, and if you do, if that uses a grounded hose - that can be an issue.
Other than that - are you in Colorado? That’s a possible issue (I’ve lived there - no disrespect intended).
I do have dust collection, but when trouble shooting this I’ve removed all of it including the dust shoe, hoses, pretty much anything that could hold or produce a static charge.
No sir, North Carolina here. Right out side of the big city of Albemarle.
Hmm… it does sound like you are on the ball with diagnosing these issues.
There’s conflicting information about grounding both ends of shielding. I have no idea what is correct. I have pretty much the same setup and have grounded as you have, though.
One simple thing that could look like EMF is vibration disconnections where the USB connector wobbles a bit and causes disconnections. Can you secure that to rule it out?
They are. There is a “disconnect” and “connect” sound when it is switched on and off, or the USB is disconnected or reconnected. When the error hits, there is no audible sound from the controller.
The “sometimes” issue with grounding shielding at both ends is if the two ends are at a different ground potential then you will now pass current along the shielding. This doesn’t usually happen in a residential situation as the ground is generally only in one place / potential… A commercial settings with multiple buildings and power sources can be a nightmare… speaking from experience on that one 
Thanks for that… makes perfect sense if ground != ground.
Here’s a thought from left field… could the Z stepper motor be weak and when the spindle is powered up, it’s picking up the EMI. The reason that came to mind is I’ve noticed they have been getting rather warm to the touch. Not so hot that you can’t touch them, but warmer than I care for. Just always figured that was the norm. Grasping at straws at this point.
I guess you could try disconnecting the Z stepper and turning the spindle on…
How are the brushes in the spindle doing?
Put a new spindle on this morning.
Perhaps a stupid question, but is there anyway to power the spindle directly and leave the VFD off?
I am stuck on that fact that everything is fine until the spindle is active. A new spindle is now in and the issue persists, so it can be potentially eliminated as the cause. This leaves the VFD to be tested in my mind…
The spindle is a 220 3phase and all I have in the shop is single 220. Going to try to run it with the Z stepper unhooked. If that works, then a new VFD will have to be ordered. That’s all that’s left.
Disconnected the stepper motor but can’t trick the motherboard to get it to home. Guess it’s time to visit the jungle an order a new VFD.
Any suggestions, probably be a few days before it gets here.
The Z-axis motor tends to run warmer than the others for two reasons:
- stepper motors work their hardest when holding position
- it’s not bolted to a steel plate to serve as a heatsink
That makes sense, thanks! Ordered a new VFD but won’t get here till Sunday.
Do you have a decent noise filter on the 220V power connection feeding into the VFD?
Not a surge protector, a proper noise filter?
No, where would one connect at?
Something like this
Which connects between the 220V supply and the power input to your VFD.
The VFDs generate a lot of EMI and inject it straight back down the power cabling, into your house wiring etc. These filters help damp that noise down, the closer to the source (VFD) the better.
That one has a circuit diagram on the top showing how it works, two LC filter stages, the inductors are in series with the power flow and the capacitors across L and N, this works to attenuate the high frequencies which cause the problem. Also note the discharge resistors to make sure the capacitors don’t just sit charged up for ever.
Edit - Also, do not put a filter on the output of the VFD between it and the motor, that is bad, some people put ferrite cores on the UVW cables and get away with it but…
So that one goes at the VFD where the 220 single connects in. Do they also make one that connects to the 220 3phase out going to the spindle? Or is that needed?
Just seen your edit, thanks and I’ll get on ordered.
Yep, that’s right.
Nope, please do not do this. The screened cable with good quality braid and foil screen, routed away from other cables is how this is dealt with. Also keep the spindle cable as short as reasonably viable.
The VFD sends pulses of power down the cable to the big inductor which is the spindle windings, they are designed to be able to do this and to interact with the motor. Putting filters in the way is generally bad for them and most of the better manufacturers specifically ask you not to do it.