|
Application Notes:
Diagnosing Power Problems at the Receptacle |
|
 |
| |
|
|
| |
|
| |
Three quick measurements can tell you a great dealWhen there is suspicion on the reliability of the building's electrical supply, you are called in. Before you break open a panel, much less the three-phase monitor, you first grab your Fluke digital multimeter (DMM) and take a few measurements at the outlet nearest to the problem load. You take three quick measurements because there are only three measurements to make: hot-neutral voltage, neutral-ground voltage, and hot-ground voltage. Armed with this information, you are well on your way to answering these questions:
- Is the outlet mis-wired?
- Is the branch circuit too heavily loaded?
- Do sensitive electronic loads have the voltage they need?
Testing a three-slot receptacle for grounding, polarityMis-wired receptacles are not uncommon. A three-slot receptacle has a hot slot, a neutral slot, and a grounding slot. The short slot is the hot, the long slot is the neutral and the U-shaped slot is the ground. Are the hot (black) and neutral (white) wires reversed? Are the neutral and ground (green) wires reversed or shorted? These conditions can often go undetected for a long time. Many loads are not sensitive to polarity. In other words, they don't care if hot and neutral are reversed. Electronic loads generally are indifferent to ac polarity because they are just converting the ac to dc in their internal power supplies. On the other hand, sensitive electronic loads such as computer equipment and instrumentation do care about a clean ground; i.e., a ground with no voltage and no load currents on it. A single reversed neutral and ground can compromise the entire ground system.
Of course you made these voltage measurements during office hours and under normal load conditions. And what did you find?
- Hot-neutral.
- Neutral-ground.
- Hot-ground.
Are these readings normal? Is the outlet wired correctly?
The most common miswiring occurs if hot and neutral are switched, or if neutral and ground are either switched or shorted. How do you spot these conditions?
Let's investigate hot-ground voltage a bit more. Hot-ground reading should be the highest of the three readings. The ground circuit, under normal, non-fault conditions, should have no current and therefore no IR drop on it. You can think of the ground connection as a wire running back to the source (the main panel or the transformer), where it's connected to the neutral. On the receptacle end of the ground path, where the measurement is being made, the ground is not connected to any voltage source (again, assuming there is not a fault). So the ground wire is like a long test lead back to the source voltage. When there is a load connected, the hot-ground receptacle source voltage should be the sum of the hot-neutral voltage (the voltage across the load), and the neutralground voltage (the voltage drop on the neutral all the way back to its connection to the ground circuit).
Testing for voltage dropOn an ideal circuit, there should be no voltage drop: the less the voltage drop, the more "stiff" or reliable the source. But in reality, there is always some voltage drop through the system wiring.
- Wire gauge will affect voltage drop - the smaller the wire, the higher its impedance.
- The longer the wire run on the branch circuit, the greater the impedance and the greater the IR drop.
- And finally, the more heavily loaded the circuit, the greater the voltage drop (V = IR, so the more current, the more the voltage drop).
Since the first two factors are hard to change, it's the last question - is the circuit overloaded - that you are usually trying to answer. To measure voltage drop, we use the most "mysterious" of all these measurements, the neutral-ground voltage.
In most office environments, a typical reading of neutral-ground voltage is about 1.5 V. If the reading is high, above 2 to 3 V, then the branch circuit might be overloaded. Another possibility is that the neutral in the panel is overloaded. Check the neutralground voltage at the panel. What are we looking for? To accommodate computers and other electronic loads, the neutral should be at least as large as a feeder, and preferably twice as large. Notice how the hot-neutral voltage drop (5.2 V) about equals the sum of the neutral-ground and hot-ground voltage changes (2.4 V + 2.7 V). The combined black wire and white wire IR drops subtract from the voltage available to the load, the hot-neutral voltage. The white wire IR drop is easy to measure as neutralground voltage, but the increased current causes an IR drop on the black wire as well as the white wire. This black wire IR drop is measurable by the difference between the no-load and load hot-ground voltage (2.4 V). In the real world, it's not that easy to switch all the loads on and off to make this measurement. That's why the neutral-ground measurement is so useful.
Measure peak voltage The outlet is the farthest point in the wiring system from the source. That means it is the most vulnerable to voltage supply problems. But to the single-phase load, reliable or not, it's the only point in the system that matters. All our voltage measurements so far have been in rms values. You know that you need to measure the peak value as well. Many meters will specify a 1 ms peak or peak hold option. Since a half-cycle of 60 Hz is about 8.3 ms, the 1 ms peak function can capture the half-cycle peak. The normal peak, assuming that the ac voltage is a more or less perfect sine wave, is 1.4 times the rms. For 120 V, that equals about 168 V. Now why is it important to measure the peak? Because electronic loads care about peak value, since that is what they use to power their ac-to-dc conversion circuits. When almost all the loads on a circuit are electronic, they are all drawing power at the same time, from the peak of the wave. As a result, the sine wave tends to become "flat-topped." This makes it harder for electronic power supplies to charge. An rms reading alone won't spot this problem.
Click here to download the full pdf: Diagnosing Power Problems at the Receptacle » | |