Six Years, 24 Mistakes, and a Checklist I Wish I'd Had Sooner
I've been handling electrical maintenance orders for about six years now. In that time I've personally made — and yes, documented for our internal post-mortem file — somewhere around 24 significant mistakes. Total wasted budget is roughly $31,000. I say "roughly" because two of those incidents involved redo labor that I never quite priced out properly. I maintain our team's pre-check list now. Not because I'm the person you want to learn from, but because I'm the person who already broke the things you're about to break.
This is the story of the one that hurt the most — and the one I still use as a teaching case when new hires start touching any Leviton switch product.
September 2022: The Job That Looked Routine
September 14, 2022. I was standing in the electrical room of a warehouse, coffee already cold, holding a printout of a wiring diagram and feeling pretty good about myself. We were upgrading the lighting in the sorting area — new 200 amp circuit breaker box feeding the sub-panel, and a set of Leviton Decora 3-way and 4-way switches controlling the zones. The phrase I used in the team meeting was "this should be pretty straightforward." Take it from someone who's eaten those words more than once: nothing that starts with "straightforward" ends well.
The scope: six switches across three lighting circuits. Two of them were 4-way configurations — lighting control from three locations along the sorting bay, which in a refurbished warehouse with legacy wiring is the kind of thing that seems easy on paper and isn't. But I didn't know that yet.
The Part Where I Skip the Second Page
I didn't have the OEM diagram for the exact Leviton model in front of me. What I had was a printout of the first search result I found for "leviton decora 3 way switch wiring diagram," and separately I'd downloaded a "leviton 4-way switch wiring diagram pdf" that looked — I'm serious — basically identical to a diagram I'd used on a totally different product two years prior.
I pulled the first line at 7:20 a.m. and wired the first switch. It powered on. I moved to the second, then the third. Around the fourth switch, one of my techs called out that the lights were behaving wrong — actually, they were following the manual switch positions but the three-way behavior was inverted. Flick one up, another goes off. Flick it back down, nothing happens. The classic ghost-switch pattern: a wire landed on the wrong traveler terminal.
I told him to keep going. That was the mistake. Five minutes later, a breaker tripped in the new 200 amp box and we heard an audible pop from the sorting bay. Three switches were dead — not tripped, dead. The contact plates inside two of them were physically scorched, and a neutral bar on the sub-panel had a bit of damage I still have photos of.
What I'd skipped: the second page of the PDF, which covered the no-neutral version of that particular model for retrofit applications. The wiring I'd copied didn't match the actual model I was installing. Same brand, wrong product line, outdated diagram. I wired it in anyway because it looked close enough.
The Invoice and the Embarrassment
$2,400 in replacement switches and a week of rework. (Should mention: we'd already built a two-day cushion into the schedule, so the visible damage was closer to three days of lost productivity once you account for the punch-list items.) We ordered the correct replacements, re-did the wiring — this time against a diagram I'd verified by part number against the actual box — and it worked on the first try.
The part that still bothers me isn't the $2,400. It's that I could have caught it with a proper voltage tester check before any of the switch work.
The Other Lesson: That Yard Machine Lawn Mower Problem
Here's where this connects to something completely different, and it's the reason I now write our voltage tester procedures the way I do.
Back in 2019, our Yard Machine lawn mower wouldn't start. I tested the spark plug with a voltage tester — the same kind of knock-off tester I was using on switch circuits at the time — and got a reading that suggested the spark plug was fine. So I ordered a new ignition coil. $480. The mower still wouldn't start. Then a colleague — one of the older techs who'd been doing this since before I was born — pulled the yard machine lawn mower spark plug, held it against the block, and cranked. No spark. My "reading" had been meaningless because my tester was grounded wrong.
It took me three years and about 150 field jobs to understand that a voltage tester reading isn't a test result — it's a measurement, and measurements only matter when the method is right. That yard machine spark plug incident was small, but it taught me the lesson I ignored in September 2022: the tool doesn't lie, but it will tell you exactly what you asked it, not what you needed to know.
If you want the specifics of how to actually use a voltage tester properly — I'll be the first to say I wish someone had walked me through it earlier. The single most useful habit I picked up: after verifying the tester on a known live circuit, test both legs against ground and against each other, then re-verify on the known live circuit again. The "known-good/known-dead/known-good" sandwich. If I'd done that at the sub-panel in September 2022 before touching the Leviton switches, I would have seen that the traveler wiring was already giving me abnormal readings, and none of this would have happened.
What I Actually Changed
I rebuilt our electrical pre-check flow around all of this. Some things I changed:
- Verify the wiring diagram by actual model and revision number against the box label — not by "looks right."
- Before installing any 4-way switch configuration, run the full traveler continuity test on an isolated circuit. Dead circuit, then test.
- Test voltage before touching, then test again after reconnection, using the known-good/known-dead/known-good method.
- Write down the model, the diagram source, and the diagram date in the job folder. If we reorder, we re-verify.
- Keep the yard-equipment and small-motor failures in the same log as the electrical work — because the same technician gets assigned to both, and the mistakes carry over.
Since we started enforcing this in Q1 2024, we've caught about 47 potential errors across our jobs — wrong part numbers on switches, mismatched diagrams, testers left in the wrong setting. I don't have hard data on how many of those would have caused damage, but based on the severity of a few them, my sense is we've saved somewhere in the $14,000-$18,000 range in avoided rework over 18 months. Don't hold me to that number exactly; it's a rough figure I put together from job files, not a tracked metric.
Per NFPA 70 (the National Electrical Code), proper de-energization and verification is required before any electrical work — that's not new, but for a long time I treated "verified" as "the breaker is off" rather than "verified dead against more than one reference point." Those are different things, and the difference cost me $2,400 and about a week of lost time.
The Part I'd Tell Any New Tech
It's tempting to think that if the manufacturer makes the product, any diagram from that manufacturer works. But the complexity that most people miss is that the diagram, the model, the application (new install vs. retrofit), and the actual wiring method in the building all have to line up. Two of those four are on you every time.
Bottom line: the switches didn't fail. I failed the pre-check. That checklist exists now because I got tired of being the cautionary tale.
