Grounding cables

Am I correct that analog signal cables should be grounded only at the side of the cabinet (for cables going into the field) and that shielded cables between drives and motors should be grounded both at the cabinet and at the motor?

It is my experience that cables are to be grounded at one end regardless of the type of signal. This way the ground cannot complete a "circuit" of impressed voltage that may develop from nearby current fields from other wires. No current can flow in open circuit. Grounding both ends can causes current to flow and these currents (which are usually sporadic and createdfrom such things as solenoids energizing/de-energizing) develop their own H or current fields which will in turn affect the cable that was grounded in the first place. It can also act as an "antenna" with the length of the cable determining the wavelength it will be vulnerable to.

PLC, you are correct. Instrument shields are usually made of thin foil or small wire mesh and their purpose is to protect the internal conductors from ambient electrical noise. You can look at the cable and see that they are not designed to be conductors.

In some cases, and VFD/motor leads are a good example, there is a need to force two different grounds to the same potential. This can be done with a simple ground wire attached to the ground at both ends. This is usually called a ground bond since both ends were grounded before the bond was added. (The problem is that ground here may be at a different AC and/or DC potential than the ground over there)

When a cable is supplied with a ground wire that is clearly sized to be a conductor, then it can be used as a ground bond. It probably will also have a shield incorporated into it and it will serve as protection from ambient electrical noise, as well.

Grounding and shielding is a tricky business. People who are good at it are worth their keep. Since I am not one of those people, it seems to be more art than science. I'm sure that's not really the case.

What DickDV says +1

Check out Rule 8 in the attachment in This Post

I was also taught the same thing. The little low voltage signals need a ground at one point as their purpose is to catch induced noise from the outside in order to ground it safely away from sensitive electronics, and in VFD cable, the purpose of the shield is to keep all that trash in and drain it near all of it's points of generation so you don't let it spread to the sensitive conductors and electronics.

You know there's major noise in a VFD cable, so the shield keeps it in and drains it in large quantities.

Actual Experience:
Automatic Tire Assembly machine with 10HP drum drive using 1336 Impact, shielded cables, short distances.

After upgrading the whole control system, I was testing the build sequence in semi-auto and noticed the drum pressure reading wouldn't quite get to setpoint causing a hesitiation.

They were building tires, just had to go to manual and Step + once after the B&T package was inflated into place.

Hmm, nothing wrong with the PPC valve, and the analog gauge is right...scope shows 50 volt spike riding on top of the 0-10vdc in the shielded cable from the pressure controller...the drum is running, so I open that door to peek at the drive, and I see the shield wire already heat shrunk and labeled laying in the bottom of the panel not stripped or terminated. The electrician was standing there, and I pointed it out to him. He said, "I can land that while it's running, right?" I said, sure, "It's just a ground for noise". I asked, "Is it already landed at the other end?", He assured me it was, and indeed we later checked and the other end of the 25' cable had it's shielded bolted to ground.

When he stuck it in the terminal block firmly and quickly it still arced...a big thick spark too, yikes...yes, he turned is head around quickly, without losing the connection and gave me that "you sum---ch" look, but the analog signal imeediately went solid and the machine never hiccupped again.

That visual of a thick blue 1 inch arc gave me a whole new appreciation for VFD noise and shielding.

OkiePC, probably what you were also experiencing with that ground arc is the far ground being significantly offset in voltage from the place where the new connection was being made.

Grounding is intended for personnel safety and each local ground is good for that purpose because that's where the personnel are. But, the ground is NOT a good voltage reference or conductor. I've worked on equipment in two buildings only 100ft apart and have measured as much as 30VAC and 50VDC between the two grounds. That could have been the source of your arc---forcing the two grounds to come to the same potential.

That's why it's so important to include the ground bond wire between the VFD frame and the motor frame. If that is not done, the motor is out there jumping all around with respect to the VFD ground and the poor VFD can't figure out what it's doing so it just faults. And faults. And faults! That's true on grounded wye input power and even worse on floating or high-resistance grounded power.

Plc_User said:

Am I correct that analog signal cables should be grounded only at the side of the cabinet (for cables going into the field) and that shielded cables between drives and motors should be grounded both at the cabinet and at the motor?

Click to expand...

Here are recommendations from Siemens, MM4 EMC Guidelines
http://cache.automation.siemens.com.../MM4Appl_008_EMC_DesignGuidelines_V2_1_en.pdf

4.2.2 Signal cables
General
Signal cables should be protected against possible interference being coupled-in. This is the reason that shielded cables should also be used here.

Digital signal cables
The shields of digital signal cables, e.g. 24 V control signals, encoder cables, should be connected at both ends to ground through the largest possible surface area to ensure a good electrical connection. If there is a poor potential bonding between the shield connections then in order to reduce the current flowing through the shield, an additional bonding conductor – minimum 10 mm² – should be routed in parallel to the shield. It is permissible to connect the shield several times to ground. We do not recommend foil-type shields. Braided shields have a shielding effect which is 5 times better than foil-type shields.

Analog signal cables
In order to establish good potential bonding the shields of analog signal cables should be connected to ground at both ends. When the shields of analog signal cables are connected at both ends this can result in signal noise – the so-called ground loops (hum). In this case, one of the two ends is grounded through a capacitor (0.01÷0.1μF) which has a very low impedance at high frequencies.

Kalle

DickDV said:

OkiePC, probably what you were also experiencing with that ground arc is the far ground being significantly offset in voltage from the place where the new connection was being made.

Click to expand...

Yes, I also suspect that the shield on that motor lead was picking up noise and had built up quite a charge since it was rounted through about 20' of square duct with (i think) 7 other VFD cables, 2 of which had no shield, and may have been running at the moment he landed the wire. I felt pretty safe letting him do it with the machine running, since he was putting in in a terminal block alongside a lot of other motor grounds, and 480 volts was at least a foot away, but that was many years ago. I would not make the same decision today...shut it down, lock it out, land the wire, never make sparks unless absolutely necessary (like in the shop with the TIG!).

KalleOlson, those Siemens grounding recommendations are quite different from conventions in the rest of the control and drive world. I don't know why they are out there by themselves on this but that is there choice.

DickDV said:

KalleOlson, those Siemens grounding recommendations are quite different from conventions in the rest of the control and drive world.

Click to expand...

Hi Dick, 'quite different'? I don't follow you there...

Might be my fault, I omitted their recommendations on the motor cables section, here it is:

4.2.1 Motor cables
⇒ Motor cables radiate a high degree of interference. In order to limit the emission level, shielded, low capacitance, 4-conductor motor cables should be used (U, V, W, PE conductors and outer shield). Suitable cables include, for example, SIEMENS PROTOFLEX. The cable shield must be connected at both ends to the housings/frame (ground) through the largest possible surface area. Cables with a YCY copper braiding provide a good shielding effect. They are better than cables with SY steel armor.
⇒ Feeder cables for motor temperature monitoring (PTC or KTY) must be routed separately away from the motor cables! We recommend that shielded signal cables are used.
⇒ If it is absolutely necessary to interrupt motor feeder cables to connect reactors or terminals, then the non-shielded cable should be kept as short as possible.
⇒ If it is absolutely necessary to interrupt the motor feeder cable to connect contactors, switches, terminals then these must be separately mounted away from other terminal strips, switches etc.
⇒ A suitable M gland with shield contact can be used to connect the cable shield to the motor. It must be ensured that there is a low-impedance connection between the motor terminal box and motor frame (ground).
⇒ The motor terminal boxes should be manufactured out of metal and be electrically conductive (gray cast iron / aluminum alloy). It must be ensured that the terminal box cover is flush with the terminal box so that there are no gaps which could result in an antenna effect.

Kalle