Should You connect Transformer Ground to Building Ground?

Q. What are good transformer grounding techniques? I would especially like to know about connecting the grounding ring to the building ground. Because the neutral is grounded at the transformer and service switch in a typical low-voltage installation, the electrical inspector in our area isn’t allowing these grounds to be interconnected. He believes an interconnection provides a low-resistance parallel path for any unbalanced neutral current to flow from the service switch to the transformer. This creates a condition where current could flow on a ground conductor during normal circuit operation—an NEC violation. This scheme also results in a dual set of ground rods installed—often within 10 ft of each other but not interconnected. I’d like to learn more about this. —B.T.

A. The electrical inspector is correct—the neutral should be grounded at one point only. Otherwise, neutral current will flow over the grounding conductor. Some 20 years ago, when I was designing electrical systems for the Marine Corps, we made a point to be sure to connect the transformer ground to the service-entrance ground for the very reasons mentioned. —G.C.

A. When two “isolated” grounding systems are horizontally separated by only 10 ft of earth, they are not electrically insulated from each other. This medium-to-high impedance can lead to a possibly harmful potential difference between the grounding systems and the distribution system’s neutral and grounding conductors.

In B.T.’s example, there are two grounding connections of the system’s grounded conductor (neutral): one at the transformer and the other at the building’s service disconnect. Each is connected to its respective ground electrode. However, a reduced amount of objectionable current will still flow across the earth gap, separating the two grounding systems. Small voltages thus generated are more than enough to create possible destructive interference on digital control and communications networks and computer-operated devices. Preventing this requires one connection between the power system’s grounded conductor and just one of the grounding systems, with both grounding systems then directly connecting together. This is permitted by Sec. 250-21 of the NEC. —F.M.P.

Why Do Light Bulbs Fail Early?

In the August ’98 issue’s “Readers’ Quiz” column, readers discussed the early demise of incandescent light bulbs. I would like to add a few bits of information. First, life is expressed as average. Some have a short life and others long, given the average life listed on the product. Some manufacturers don’t list average life—probably because the tests for it aren’t under real-world conditions. Temperature is controlled, lamps are cycled on and off slowly, no vibration, no over voltage, only rated voltage, and so on.

Second, about L.D.E.’s suggestion to use only zero crossing switching to turn them on and off—why do the light bulbs in my house last years with standard “snap” on/off switches? If you are using 500W and above, a ramped turn on is best as you start them over 1 sec to 2 sec (60 to 120 cycles) not to full power in about 4.2 ms (1/2 wave of AC power at 60 Hz). Starting a smaller wattage lamp at the zero cross point over has little effect as far as I know. The best answer is to use a better bulb from a major U.S. manufacturer. I do!

Third, the response from T.F. about hot filaments: Low-voltage bulbs have a heavier filament wire, as less ohms are required at lower voltage. But at 120V or 240V, the amount of ohms required is higher. The easy way to get them is to use a thin wire. When hot or cold, thin wire will break faster and easier than heavy wires. Vibration is a killer; tungsten filaments are brittle and affected by hot or cold. A 100W, 120V, or 130V bulb wouldn’t make it down the street as a headlight in your car—but did you consider how long your headlights last? The wire gage and support structure is the major difference.

Why not do an article exposing truth about short life of light bulbs? —E.S.