How do you pass shielded servo motor cables through the cabinet wall/panel?

I think we all learn the hard way with this stuff. My experience is mostly with plasma machines. Plasma is a particularly noisy  EMI environment and high frequency start machoines rais that to the next level (which I have never played with). I have however had EMI issues in the past and even bought a USB Oscilloscope to monitor what was going on.  Hypertherm are the industry leader with Plasma. Their document I linked to is the gold standard for plasma cutter EMI management. Its not written for any specific motor type (eg stepper or servo).

I do know by adding their recommended witing to my plasma table togethere with a mains EMI filter on the control panel I solved my EMI issues.

When I looked at how to wire steppers, I thought It was accepted practice to attach ethe chield at one end and leave the other unconnected. I chose to connect the enclosure endIt did not make sense to me to not connect the shield to 

So I will stick to the Hyertherm recommendations as they have worked for me and the environment I work with. I think future cabinets I make will use cable grommets and where appropriate, I will add the ground clamps you have found as well as the HYpertherm stuff.

@Rod
The document by hypertherm you linked is certainly valid and should be followed. But in my opinion it doesn't cover how to connect the braided shielding of a cable to ground and that was the question I was referring to.
So to recap my input, it is not recommended to attach a wire to the cable shielding and run that to a ground post.
This is another example from www.focusondrives.com/more-about-grounding-and-shielding/

 

Yes I agree with you.  We travelled off track somewhat from how to pass cables throug the cabinet wall 
I think my learning from this thread is the cable should be passed though using a simple cable grommet (no soldering ), the insulation should be pared back and a cable ground clamp applied. The cable should continue direct to the driver. If required by the motor driver, a short pigtail should be used to connect the shield at this point. Seems like an extra shield clamp should be applied near the driver too. Did I get that right?

At the motor connection, if it does not have custom cable connectors, I will continue to use the QC connectors (no soldering). They seem to work fine for 1.5mm cables and there are larger ones available if you need higher current.  I use 4 pin housings but housings with more pins are available.

I might add this increases the size of the cabinet required fairly substantially, I struggled for room in a 500 x 500 mm cabinet. Using ground clamps would probably add another 200 mm in one dimension.

rodw wrote:

Yes I agree with you. We travelled off track somewhat from how to pass cables throug the cabinet wall

That's okay, my original question was born in large part from me not understanding how the motor cable shield works, electrically.

rodw wrote:

Did I get that right?

Not really. You seem to have skipped the parts you don't like. E.g. the part about connecting the shield properly at the motor side. And, as already explained, I'm not sure ground clamps at the drives help unless the drives have a low impedance mounting to the back plate - which yours don't. You are correct in that a very short pigtail is probably the best compromise available with those DIN-mounted LAM drives, but it's not a good takeaway that this is the way it "should be done".

If you want to learn something from this thread I think it should be to see the shield as another conductor between the drive and the motor frame - except this conductor should have low impedance for high frequency signals, which necessitates wide 360° connections at both ends (and an otherwise unbroken shield).

rodw wrote:

I might add this increases the size of the cabinet required fairly substantially

Indeed, this is problematic! It's very obvious in the Siemens video how much space the shield connection takes up at the drive.

I don't think there can be a one size fits all solution.
I suspect Seimens are talking about mains or 3 phase powered servos.  Steppers use lower voltages. I think also from some images i've seen that Siemens have provision to attach the shield at the motor end. Stepper motors do not have this provision. Here is a pic of a couple of  stepper motors. 
 

The Moon motor (shown on my machine) has provision for an earth point which I've used. Its not a shield attachment point. This is a high quality motor, not the usual Chinese junk! The attached wire goes back to a star ground point on the frame. There are many views about shield treatment but the conventional advice for steppers I have seen and the advice of my HF electronics advisor is to connect the shield it at one end. I'll stick with that as this machine has been super reliable. 

The ethercat closed loop stepper in my hand also does not have any provision and does not even use shielded cable for the encoder wiring.

rodw wrote:

I don't think there can be a one size fits all solution.

Neither do I. I'm not implying that there's anything wrong with your machines! They work, don't they? Good enough is good enough, and if it works it works. Like Aciera said, different machines will have different EMI requirements, and like you say, a machine with low voltage steppers will probably generate less EMI than a big industrial machine's drives.

(Although I have a Leadshine DM556 stepper drive that absolutely kills FM reception on a radio connected to the same outlet unless I wrap the 48 VDC feed of the drive around a ferrite core, so low voltage steppers sure aren't innocents! )

I just feel that it is a shame if you cannot absorb new knowledge just because your existing machines aren't perfectly in line with that knowledge. I believe that the better understanding you have of the principles (an idea on how you would make the perfect machine if you will), the better equipped you are to then make reasonable compromises in each case - and *every* machine will need compromises in different areas. And if you then get problems with EMI, you will have a much easier time fixing them if you understand the compromises you've made.

Now I am puzzled too by the general lack of provisions to attach the shield at the motor end. Yes, the large Siemens motor in the video has it, but just like you I've noted that most other motors don't - not my old Yaskawa servos, not my more modern Omron servos, and certainly not the cheap Ebay steppers. I have no explanation for this except perhaps that they leave it up to the machine builder to figure out how to attach the shield. It could also be that I've simply misunderstood how the shield works. In that case I hope someone with an intimate knowledge on motor drive EMI shows up here and explains what I've gotten wrong so that we can all learn from that.

I don't think I'm ever unwilling to apply new learnings but I can't implement the suggestions from this thread on existing machines.
I will say some of my learnings building a plasma table have been particularly brutal, in particular noise management....

Indeed all this servo shielding stuff is highly confusing.
Why most asian servo manufacturers use these crappy Molex connectors at the motor end while recommending shielding at both ends?! Yaskawa, Delta, Panasonic, ...

Meanwhile all European servos use proper shielded M23 connectors...