In previous articles, I have covered typical troubleshooting case studies, including sags caused by large- horsepower motors and heating elements, oscillatory transients caused by power-factor capacitor switching, destruction from lightning-induced impulsive transients and harmonics from rectified input switch-mode power supplies. Though those types of power quality problems are usually identified through straightforward analysis of PQ monitor recordings and basic knowledge of Ohm’s and Kirchhoff’s laws, I have had a number of calls for problems that didn’t even require that.

Information about what happened usually provides the starting point of the investigation, such as, “The lights blinked and then my computer stopped working,” or “I heard this humming sound before I smelled smoke.” Knowing the time/date and details about the affected equipment and recent changes in the electrical infrastructure and equipment are also typical starting points for the investigation. But those familiar with the TV show “NCIS” may recall Gibbs’ Rule No. 3: “Never believe what you are told. Double check.” This is equally as valuable in a PQ investigation as in a criminal one.

Lights behaving badly

Two recent investigations fell right in line with that rule. All involved lights not behaving properly. The first call was for the lights in two upstairs bedrooms. The ceiling light in one room was reported to have gotten suddenly brighter while the other bedroom’s light went nearly dark with a small amount of flickering. Having been in that dwelling when the homeowner first moved in, I knew several of the circuits had no grounds in the boxes and others had the line and neutral wires reversed. But the electricians had supposedly corrected those problems several years prior. The three-head ceiling fixture had LED bulbs that replaced original halogen bulbs.

Removing the mounting plate revealed oversized wire nuts on unnecessary jumper wires to the switched line and neutral wires and the lighting wires. However, the voltage to the light read 122V with no variations. Without having seen the light level before and after the reported sudden increase, there was no way to precisely determine why it changed, though it looked normal now.

The other bedroom light was, as reported, dimly flickering. The circuit with the light switch also powered a receptacle with an electric heater that was plugged in but not turned on. Voltage at switch and receptacle was 121V, but dropped to 114V when the heater was switched on. While a rapidly cycling heating element could cause flicker, this wasn’t the case, as the flicker occurred regardless. Removing the light fixture cover revealed two LED bulbs, one flickering and one with no light output. Before removing the mounting plate, I made a quick measurement of the bulb socket’s voltage: 121V. Since the bulb was out, I asked for a replacement for the flickering bulb.

A second replacement bulb solved the first problem. It was undetermined whether the low-voltage level when the heater was running contributed to the bulbs’ premature failure, so just in case, the heater was plugged into the same circuit the window air conditioner was plugged into during the summer months to eliminate a potential repeat of the issue. With the two bedrooms having different circuits, it was likely just a coincidence that both problems seemed to occur at the same time.

The next investigation started with a frantic call on a Sunday afternoon. The homeowner was afraid the house was going to burn down because the kitchen lights went out “suddenly.” The house was built in the 1940s and other wiring issues found during renovation had already put the homeowners on edge. I started the investigation by checking the breaker panel (none tripped or warm), then removing the dimmer switch that controlled the lights to measure the voltage feeding in and out of the dimmer. The 118V reading with dimmer fully on was typical for that dwelling in the summer. Varying the dimmer produced expected voltage levels. Removing the light fixture, I found a different situation. The old cable’s barely distinguishable hot wire was missing some insulation as it came through the unprotected cable clamp. Ah, the homeowner said, “See, I was right.”

After that problem was repaired, however, the lights still didn’t come on. All three bulbs were incandescents, so a quick ohm reading check found the problem—all three bulbs were dead. The most likely scenario is that they failed one at a time, and the homeowner kept turning the dimmer up to make it brighter in the room.

If the homeowners had checked the obvious potential problem first—bulb failure—they could have saved the expense of a weekend service call. However, it was likely a good thing they called to find the other hidden potential problems that might have had more serious consequences down the road.