What Are Connectors Used For? A Quality Inspector's Honest Take on Southwire, Testers, and the 30-Second Check
I don't care how expensive the cable is. If the connection is wrong, the job is wrong.
I work in quality at Southwire. Every month I review hundreds of connection points—crimp terminals, twist-on connectors, compression lugs, push-ins. I have rejected batches because a connector wasn't rated for the conductor. I have approved jobs that looked ugly but tested clean. So when someone asks 'what are connectors used for?' they are asking the right question: where does my installation depend on a mechanical part that can fail?
In my opinion, the connector is the part of the system that deserves the most respect, not the least. Let me explain why.
The Cable Is Not the Problem
People start with the cable. They want to know whether Southwire Romex is better than another brand, whether THHN from one plant is 'stronger,' or whether a spool marked 'Southwire Owensboro KY' is different from one made somewhere else. I get it. Cable is the big purchase, it's visible, and it's easy to compare spec sheets.
But from a quality perspective, the wire itself is usually the least exciting part of the installation. A properly made cable from a named manufacturer is almost always going to meet its electrical specs. The connector is a different story.
I used to think the cable was where quality lived. Then I spent a year tracking field failures and changed my mind. A connector is where the electrical path is interrupted and reassembled. That means it's where heat builds up, where corrosion starts, and where a loose joint can turn a project into a callback. Literally a $0.20 part. I have seen a bad lug connection cause an $18,000 rework invoice because it was one missing turn past torque. The cable was fine. The connection was not.
So, What Are Connectors Used For?
Let me rephrase that: connectors are used for two things. Establishing electrical continuity and maintaining mechanical containment. They have to let current flow from one conductor to another without adding excessive resistance, and they have to hold that connection under vibration, thermal cycling, and pulling forces.
Practically, that includes:
- Joining conductors – twist-on wire connectors, crimp splices, and set-screw connectors for splicing two or more wires.
- Terminating to equipment – ring terminals, fork terminals, and compression lugs that land on breakers, transformers, or busbars.
- Providing a disconnect point – plug-and-socket connectors, pin-and-sleeve devices, and modular plugs that allow equipment to be removed without cutting cable.
- Transitioning between cable types – conduit-to-box fittings, cable glands, and transition couplings that protect the cable exit point and maintain the raceway system.
Every one of those has a current rating, a temperature rating, a wire range, and a torque requirement. I have seen experienced electricians skip the torque requirement because 'it felt tight.' That's a red flag. The connector is doing all the work; it deserves the same care as the cable.
On standards, I use UL 486A-486B as my reference when auditing connector suppliers. It covers the pull-out, temperature, and corrosion tests that a connector has to pass. On the installation side, the NEC's 110.12 rule about 'neat and workmanlike manner' sounds cosmetic, but it's not. It's a way of saying the termination has to be sound, not just tidy.
What I Saw on the Southwire Villa Rica and Southwire Owensboro KY Lines
I've been in the facilities that make the wire people ask about by name. The Southwire Villa Rica operation and the Southwire Owensboro KY plant both have the same kind of process control I use in my own audits: sample the conductor, measure the diameter, test the insulation, label the spool. Good wire is made by discipline, not by magic dust.
But here's the thing: the best cable in a spool can be undone by a bad connector. You can pull a perfect Southwire feeder out of the box, land it with a non-rated connector, and have a failure six months later. That's not a cable problem. That's a termination problem.
So when someone asks 'should I buy from Southwire Villa Rica?' or 'is Southwire Owensboro KY better?' Actually, I usually answer: buy cable that meets your spec, verify the markings, and spend your quality energy on how you terminate it.
The 1100, the Multimeter 117, and the 30-Second Check
To be honest, I don't trust my eyes. A poor crimp can look perfect and still fail a pull test. A good crimp can look ugly and test just fine.
In my go-bag I keep a simple continuity tester—an old 1100 that I've had for years—plus a multimeter 117 for voltage, resistance, and continuity checks. The 1100 is not fancy. It's basically a lamp and a battery in a probing package. But it tells me if the path is complete. The multimeter 117 tells me if the resistance is reasonable and whether I'm seeing voltage drop across a joint. Together, they catch most of the bad terminations I see.
Before someone comments on my choice of testers: I know a multimeter 117 isn't the newest thing in the catalog. It's not. I like it because it gives repeatable readings, and repeatability matters more than features in a field check. The 1100 is even older. If it dies, I'll replace it with the same thing, not because I'm stuck in the past, but because I trust the behavior.
For a contractor, the 30-second check is simple:
- Make up the connection using the correct connector for the wire size and type.
- Do a visual inspection—wire should extend past the connector's metal sleeve if it's a crimp, or be fully seated if it's a push-in.
- Pull gently on the conductor. If it moves, it's not secure.
- Use a multimeter or continuity tester to verify the circuit reads as continuous—not infinite, not intermittent.
That's not over-engineering. That's the difference between a connection that works on Friday and a connection that still works next spring.
The Objection I Always Hear: 'I've Never Seen It Fail'
I get pushback on this. A lot of electricians will tell me, 'I've been making up twist-ons for 25 years and I've never had a failure.' I believe you. But 'never yet' is not the same as 'can't happen.'
Connectors fail at the intersection of three variables: the connector, the conductor, and the environment. Change one of those and the old habits don't apply. You change from a copper conductor to an aluminum one? Different connector rating. You use a fine-stranded cable instead of solid? Different termination method. You install in a damp location where the temperature cycles daily? Different material requirements.
In our Q1 2024 quality audit, the most common finding wasn't conductor gauge or insulation thickness. It was termination torque. That's a human issue, not a cable issue.
To be honest, I even had doubt when we switched connector vendors last year. What if their production run wasn't as good as the sample? The two weeks until the first shipment were stressful. That's why we test incoming lots.
This approach worked for us in the factory, and it works on most residential and light commercial jobs. If you're doing medium-voltage utility work or industrial controls with exotic insulation, the details change. But the principle doesn't.
Bottom Line
So what are connectors used for? In my view, they're used to either make or break your installation.
I'd rather spend ten minutes explaining connectors than deal with a failed joint later. An informed customer asks better questions and makes faster decisions. You don't need a wall of theory. You need to know that the connector is the part that deserves attention, and a simple 1100 or multimeter 117 check is worth the 30 seconds it takes.
Buy the cable that meets your spec—from Southwire, from another qualified manufacturer, wherever—and then treat the connector as part of the safety system. Because in my experience, every time I've seen a 'mystery failure,' it wasn't the copper. It was the connection.
Need a cable engineering answer?
Send route length, connector preference, and acceptance target. The same team that writes these notes can help review your fiber, copper, RF, or PoE assumptions.