The Tester Said Dead. It Wasn't: Voltage Testers, Southwire Romex, and a $620 Multimeter Mistake

In September 2022, I lost a full Friday to a GFCI circuit that would not stay reset. My tester kept telling me the line side was dead, so I checked every box, pulled every wire, and replaced two perfectly good receptacles. At 3 p.m., I put an old incandescent work light on the “dead” circuit and it glowed orange. The line wasn't dead. My method was asking the wrong question.

The problem I want to talk about is not “cheap testers.” The cheap tester in your pouch is fine for what it does. The failure is assuming one tool can answer every question on a job site. I carry the Southwire Tools & Equipment RS 101A, and I use it. I just don't use it as a multimeter.

What Most People Think the Problem Is

A lot of electricians and DIYers blame the equipment when a reading doesn't make sense. “This voltage tester is junk.” “This multimeter is drifting.” Sometimes that's true. But more often, the tool is answering the question it was designed for while we wish it was answering a different one.

A voltage tester answers one question: Is voltage present? It does not tell you how much voltage is there. It does not tell you whether the neutral is solid. It does not tell you whether the circuit can handle a load. Asking a voltage tester to troubleshoot a GFCI is like asking a doorbell to diagnose your phone.

Why the Problem Is Actually Four Different Problems

1. You're comparing a voltage tester to a multimeter

A multimeter is not a more accurate voltage tester. It is a different kind of tool. A voltage tester is designed to beep or light up. A multimeter is designed to measure. If you grab a beeper when you need a measurement, you get a yes/no answer to a question that requires numbers. That mistake caused my GFCI disaster: I had presence of voltage, but I didn't have voltage under load, and I didn't have a solid neutral.

2. The “best multimeter for electronics” search gives you the wrong answer for branch circuits

The phrase “best multimeter for electronics” usually leads to meters optimized for low-voltage DC and small currents. For 5-volt logic, that's great. For a 20-amp branch circuit with an intermittent neutral, you need a meter rated for the environment, not just for the expected volts. That's where category ratings matter.

According to IEC 61010-1, a meter with a CAT III 600 V rating is built to handle the transient impulses you find on distribution circuits. A meter without a category rating can fail in ways that are dangerous to the user, not just inaccurate. It's tempting to think the only difference is accuracy. It's not. If you ask me, the phrase “best multimeter for electronics” is the start of a useful question, not the end of it.

3. Test leads fail more than the meter

I once chased an open neutral for an hour because my multimeter displayed OL on a connection. I replaced the wire nut, still OL. I finally did a continuity test on the leads themselves: the red lead was broken internally where the strain relief enters the probe. A $20 set of test leads cost me an hour of billable time. Now the first thing I check is not the circuit. It's the leads.

4. Ghost voltage is real

A high-impedance digital meter can read 50, 80, or even 120 volts on a wire that is disconnected because the wire is capacitively coupled to another conductor. The wire isn't live in any practical sense, but the meter can't tell the difference. If you don't use a low-impedance input or put a load on the circuit, you're chasing a number that doesn't represent anything.

Honestly, I still don't fully understand why capacitive coupling behaves the way it does in some cable bundles. My best guess is that the parallel wire geometry makes it worse. But I no longer argue with a ghost reading. I switch to a low-impedance tester (a solenoid-style voltage tester or a LoZ multimeter mode) or I load the circuit with an incandescent drop light. One of those will usually tell the truth.

What This Confusion Actually Costs

The GFCI call cost me about $620 in unbillable time and two receptacles I didn't need. A month later, I almost replaced a $340 breaker because a meter said it was bad. The breaker was fine. The issue was a bad neutral connection three boxes away. The only reason I caught it before ordering the breaker was a 5-minute checklist I had started after the first mistake.

I've made these mistakes mostly on residential and light-commercial jobs, probably 400 service calls and rough-ins over the past eight years. If you work on high-voltage switchgear or delicate lab equipment, your experience will be different. But the pattern is the same: we trust a reading without checking the tool, the leads, or the assumptions behind the test.

We use Southwire Romex on nearly every residential job. The wire is rarely the problem, but “the wire is fine” means you've verified it, not assumed it. I've caught nicked conductors, backstabbed connections, and one staple driven through a jacket. All of those showed voltage when I poked them, then failed under load. A no-load meter reading can hide a weak connection.

I also do low-voltage network work. That taught me a different version of the same lesson: a cable can pass a continuity check and still fail on crosstalk or return loss. For network runs, you need a tester designed for twisted-pair networks, not a general-purpose multimeter. Southwire Networks makes the cable side of that business; the troubleshooting gear is a separate category.

The Short Fix That Costs Five Minutes

I still carry a voltage tester. I just stopped treating it as a multimeter. Here is the checklist I use now, mostly because I got burned:

• Prove the tester before and after. NFPA 70E teaches it, and it works: check on a known live source, test the circuit, then check on the known live source again.
• Use the right category. For branch circuit and panel work, use at least a CAT III 600 V rated meter. No rating means no guarantee.
• When you see an unexpected voltage, put a load on it. An incandescent drop light can tell the difference between real voltage and ghost voltage faster than any meter mode.
• Check your leads first. If you get an open reading where there should be continuity, touch the leads together. If it still reads open, the lead is the problem.
• For network work, use a network tester. A multimeter is not a pair tester. I learned that after pulling 300 feet of cable, and you don't need to.
• When the symptom is a long run or a weak motor, use a voltage drop calculator as a second opinion. Southwire's is free on southwire.com, and it's already in my browser history.

“Test the tester. Then trust the method enough to double-check the situation, not enough to skip the next step.”

Five minutes of verification beats five days of correction. I know that now because I paid for it. The lesson is not that the RS 101A or any other voltage tester is a bad tool. The lesson is that every test starts with a question, and the tool has to match the question.