Protective Earth cable for moving parts (e.g. CNC router spindle)

[TinkeringSteve]

As visible in my last thread in the tools subforum, I recently bought a chinese CNC router.

As mentioned in this here older thread:
https://www.eevblog.com/forum/reviews/china-cnc-6040-setup-testing-review/100/

they come with a spindle motor that runs on 230V AC 3-phase (coming out of Variable Frequency Drive inverter),
which is not mains-earth connected. Did I mention the thing is water* cooled and the O-rings inside are said to fail after some years...
That spindle motor is running on 3 (X,Y,Z) axes (ball screws), so there is quite some movement, bending connected cables back and forth.

While the necessity to fix that is mentioned, details of the proper way to do it are not spelt out anywhere (I've looked).


Now,
it may seem obvious to just use the same type of highly flexible cable that's used for the control lines and lay it out along the same path.
Or one that is even more expensive and has a "guaranteed" higher number of "bending cycles" than the other, so the signal lines are "guaranteed" to break first and when that happens, I also replace the mains earth cable.

The thing is, I don't buy into the "guaranteed", at least not strongly enough to bet my life on it.
Chances are, the earth cable will become suboptimally conductive at the wrong moment and whoever touches this will get a heart attack or whatever.
I don't like this at all.

How is this done properly?

- me

* via a submerged pump in a bucket of water, with 230VAC cable going into the bucket, what a sight! Yeah yeah, double insulation, I still don't like it Esp. when it says "CE" (china export).

-
Just for amusement: Currently, my makeshift earthing is outright horrible. The spindle is electrically connected, via screws into hits holder, to the ball screws (which are lubed with a nice insulator I guess), to another part of frame, via another ball screw... to the base frame where I have a screw with a cable attached to it that connects to earth, measuring ~ 4..5 Ohms from spindle case to earth, and who knows what happens if anything actually moves from the current position ;-)

[jbb]

Is it possible that the cable to the spindle is multi-core with a shielding earth braid?

I'm no expert, but I suggest:

• Using a Residual Current Device (RCD) / Ground Fault Current Interrupter (GFCI)
• Use a high flex earth conductor (e.g. many fine copper cores, silicone sheath)
• Use proper crimp lug or ferrule connectors
• Maybe use two earth wires?
• Inspect wire runs periodically

[CopperCone]

 :-/Oyou can get a metal pale or bucket and earth it. Not sure what to do with your moving parts

Real ones will use an air motor right?

[KrudyZ]

You could check that your protective earth connection in the spindle cable is still there by taking one of the control wires and attach it to the spindle housing.
Back at the controller you would feed a small current into this wire and measure its potential against the local protective earth.

[ajb]

Real ones will use an air motor right?

"Real" machining centers typically use multi-kW electric spindle motors and often supply liquid coolant (for the cutting tool, not the motor) through the spindle at 60+ bar.  But they also use sophisticated drives that will certainly be able to detect and fault out on any significant earth leakage.  Air-driven spindles (or spindle attachments) are typically only used where very high RPM is required.  As far as grounding, they probably have enough metal-to-metal contact that there's a low impedance path from the spindle to the base casting, but I wouldn't be surprised if additional ground bonding wires are included throughout the wiring harness. 

I suspect that adding a continuous-flex grounding conductor to the spindle and regular inspection of the condition of the wiring will give you reasonable enough protection.  Adding an additional ground sense lead for extra peace of mind as KrudyZ suggests wouldn't hurt, as long as it's done properly to avoid tripping any other ground fault detection circuitry.

[TinkeringSteve]

Hah. Interesting.
I have been thinking about adding electronics to regularly send a test signal in a loop that includes the spindle earth connection.

I wouldn't know how to design something that is guaranteed to not fail silently when actually 230VAC does leak to the thing, though, I guess.
Any pointers?

[SeanB]

2 grounding conductors in the cabling will work, size each one to handle the prospective fault current. Sensing is simple, just use a current transformer to inject a small current into the loop formed by the 2 grounding wires, and use another to detect the current in the loop. Low current and a good clamp on the other side ( actually on both so that the reflected fault current does not blow what is driving it) to limit the voltage applied to the detector and it will monitor the wires continuously for any breaks.

[C]

A GFCI would work even with no ground connection.

Put it between Variable Frequency Drive inverter and motor.
The GFCI would detect leakage from motor windings or wires.

https://en.wikipedia.org/wiki/Residual-current_device

C

[Richard Crowley]

Does your gadget use flex cable trays on each axis?



The traditional super-flexible and high-rel grounding solution has been a heavy-gauge copper braid.  Which would fit nicely inside the flex cable tray.

[KrudyZ]

A GFCI would work even with no ground connection.

Put it between Variable Frequency Drive inverter and motor.
The GFCI would detect leakage from motor windings or wires.

https://en.wikipedia.org/wiki/Residual-current_device

C

Not your run off the mill GFCI though, since it's three phase to the motor.
Also, you're likely dealing with 240 - 400 Hz to run the spindle.
You might get a lot of ground return current during normal operation and any GFCI would need to take this into account to avoid nuisance trips.

Edit: grammar

[C]

The basics of a 3 ph GFCI is just a current sense transformer where all three phase wires go through the current sense transformer.
The current sense here is a low amp value as there should be no current for normal operation.
Windings and wire often have leakage, so normal is often some low value of current.
If you build or buy, you could have a meter that shows the leakage. 
This meter can also be a benefit as it would show a motor going bad.

Yes for a VFD you have a wide frequency range, so current sense transformer has to be able to handle frequency range and allow large wires to be sensed..

C

[Gregg]

I have used CTs designed for ground fault sensing many times.  They are relatively cheap and can be connected to a relay to open a circuit single or three phase and it is fairly easy to wire in a test circuit with a momentary switch and resistor.  Many are available with adjustable thresholds like this one:
https://www.ebay.com/itm/NK-Technologies-AG1-NOR-120-FS-045-Ground-Fault-Sensor-40/142753983860
Route all 3 phase conductors through the current transformer and set the desired millamp threshold

[TinkeringSteve]

2 grounding conductors in the cabling will work, size each one to handle the prospective fault current. Sensing is simple, just use a current transformer to inject a small current into the loop formed by the 2 grounding wires, and use another to detect the current in the loop. Low current and a good clamp on the other side ( actually on both so that the reflected fault current does not blow what is driving it) to limit the voltage applied to the detector and it will monitor the wires continuously for any breaks.

Sorry for getting back at this late.
Do you know of examples of this, or could give hints of how to approach this?
I have no idea of how to dimension this. Can you use some small ferrite ring that fits around an average pencil, put some windings on it and stick the earthing cable through it, or how roughly am I to imagine what this would look like?