correct grounding

[tryinghard]

To further clarify in the CNC area where people are doing stupid things.

If the Equipment Grounding Conductor (I like to call it "safety ground wire" to help the average person understand its purpose) is disconnected at the CNC and in its place a ground rod is substituted, then effectively the earth path back to the main panel becomes the EGC. This is totally unacceptable because of the almost certainly high resistance.

A journeyman electrician will never do this

Even if the EGC is correctly installed to a CNC the addition of a supplemental ground rod at the machine only has a marginal effect on ground path noise. In other words it seldom solves the problem.

As journeymen electricians our emphasis for grounding (earthing) is an electrode system as defined in 250 part III, our emphasis for equipment grounding (bonding) is a system as defined in 250 parts V & VI. Grounding and bonding are summarized thoroughly in 250-4 again this is where are emphasis is.

A correctly wired building will not have noise. This means no parallel neutral paths, balance load respecting harmonics, no common neutrals, data/comm cabling space from voltage... Noise or stray current on non-current carrying items are most often avoidable. Existing installations with high noise will need qualified electricians to help qualified technicians clean up bad wiring/cabling because supplemental work is a Band-Aid not a repair.

 

[gar]

080430-2050 EST

tryinghard:

Unfortunately there are cases where what I have described has been the so called solution, and plant electricians have complied with the CNC manufacturers to do this isolation. I do not agree with this and constantly point out to anyone suggesting this approach that it is an NEC violation, and why. On the other hand our I232 Isolation product solves almost any noise problem in a properly bonded and grounded system.

I would disagree on your noise comment by this one example. At either end of a directly connected, meaning no DC isolation, RS232 path I short a hot wire to the machine chassis. This produces a peak voltage difference between the two ends of the RS232 cable of about 1/2 the peak voltage of the hot wire source voltage. If the source is 120 V to ground, then the peak between the two ends is about 120*1.4/2 = 85 V. This will cause some data errors and will destroy RS232 components and maybe other components at both ends of the communication channel.

By contrast I can put a 1000 V RMS (+/-1414 V peak) voltage between the two ends with our I232 Isolators and maintain error free communication and no damage. Thus, I can clearly tolerate hot line shorts to chassis in any system under 1000 V. Our peak rating for isolation is +/-2000 V for 1 second.

.