Last month, a contractor emailed me a video. Four brand-new Legrand sensor switches, all on the same job, and three were doing something strange. One wouldn't turn off. One flickered whenever the bathroom exhaust fan ran. One kept turning recessed lights on in an empty room.
His verdict was short: 'The product is garbage.'
I get the reaction. After four years of reviewing lighting controls and field returns, I've seen the same video dozens of times. Usually, the sensor switch is fine. The specification around it is not.
Quick intro: I'm a quality/compliance manager at Legrand. I review every reference installation before we sign off a control product—roughly 200 unique configurations a year. In Q1 2024, 11% of first-time install samples I tested didn't match the load, environment, or physical setup they were designed for. Not because the parts were broken. Because they were spec'd as if a sensor switch was just a fancier toggle.
The Surface Problem: 'My Sensor Switch Is Broken'
When an installer writes that, they usually mean: the switch doesn't behave like the simple light switch it replaced. Lights stay on when they should go off. Lights flicker on a dimmer. The sensor doesn't 'see' someone sitting still. Or it triggers every time the HVAC kicks in.
It's tempting to think a Legrand sensor switch is basically a regular switch with a motion detector bolted on. Here's the thing: that's the oversimplification that causes most callbacks. A sensor switch is a small computer with a relay, a lens, and sometimes a mesh radio. It makes decisions based on heat, movement, ambient light, and a pre-programmed timeout. Give it the wrong input, and it makes the 'wrong' decision—even though the hardware is fine.
The Deeper Cause: Matching the Sensor to Its World
The single biggest mistake I see is ignoring the load. And one of the easiest places to trigger that mistake is with recessed lighting.
The old '4 inch vs 6 inch recessed lighting' debate looks like a purely aesthetic choice: beam angle, trim size, ceiling footprint. It's also an electrical decision. A 6-inch fixture often uses a larger driver, with a different inrush current and standby draw. A 4-inch wafer light might draw so little current that it falls below the minimum load required for an occupancy switch to keep its relay engaged. I've tested a Legrand sensor switch that was rock-solid on six 6-inch housings and unreliable on six 4-inch wafers—same switch, same room, different driver.
That's why spec sheets list minimum load, maximum load, and compatible load types. It's not busywork. It's the difference between a system that works and a system that 'mostly works.'
There's also the neutral wire issue. A Legrand sensor switch that requires a neutral isn't being difficult; it's the price of always-on logic. If the box only has a switched hot and a load, the sensor can't power itself. I've rejected field returns that were 'dead on arrival' only to find the installer had connected the line and load screws and left the neutral nut untouched. The sensor was fine. The wiring was not.
Legrand Adorne Motion Sensor: Angles Are Not Optional
With the Legrand adorne motion sensor, installation angle and mounting height are just as critical as the wiring. The detection pattern has a defined shape. If you mount it too high, the 'dead zone' under the sensor grows. If you put it in a corner behind a door, it'll miss the main walkway. If you place it near a supply vent, moving air can trigger false occupancy.
And then there's the finish of the trim around it. Our test team has spent hours with different trim rings, gaskets, and spotlight covers. A matte white cover scatters IR differently than a glossy one. A deep cover can block part of the lens, turning a 180-degree pattern into a 120-degree pattern with a blind spot. I've seen a 'faulty' adorne sensor that was actually installed under a decorative cover its lens couldn't see past.
Zigbee and Humidity: The Invisible Variable
For wireless controls, the conversation often jumps to range and interference. What most people don't realize is that zigbee humidity is a real issue, especially in bathrooms, laundry rooms, and commercial kitchens. Zigbee operates in the 2.4 GHz band, and water vapor absorbs radio energy. In our environmental chamber tests, a stable link can start dropping packets when relative humidity climbs above 80%. The switch may still respond to a press, but the occupancy sensor's message can be delayed or lost.
That's not 'the wireless is bad.' That's 'the wireless environment changed.' Once the sensor switch moves from a dry hallway to a steamy shower room, the same protocol behaves differently.
The Cost of Getting This Wrong
Here's the part that makes me wince. Most of my field returns aren't defective. They're expensive examples of assumption.
In 2022, we had a 50,000-unit annual order where the installer substituted 4-inch recessed modules for the 6-inch units specified in our test plan. The occupancy sensors started randomly cycling the lights. The first instinct was 'bad sensors.' We reproduced the issue on the bench within an hour: the different driver minimum load. The rework cost $22,000 and pushed the launch two weeks. The substitute parts were 40 cents cheaper per unit (which, honestly, made the redo even more painful).
The most expensive sensor is the one installed twice.
A callback after a 'simple' sensor install costs more than the sensor itself: truck roll, labor, diagnosis, replacement, and the client trust you lose. The total cost of ownership isn't the unit price. It's the unit price plus the probability-weighted cost of a second visit.
That's also where the 'rush order' thinking fails. In March 2024, I worked on a project where the team paid $400 extra for overnight shipping on a specialized relay. The delay wasn't shipping. The delay was three days spent trying the wrong sensor combination in the field. The client didn't care about the $400; they cared about the missed deadline. What they needed wasn't a faster package. It was certainty.
I have mixed feelings about rushed installations. On one hand, deadlines are real. On the other, the fastest path to a deadline is almost always a pre-tested combination, not a guessed one. When a project is in emergency mode, the value of verified compatibility goes up, not down.
The Fix: Treat the Sensor as a System Decision
You don't need a lab to build reliable installations. You need a process.
- Read the spec sheet before buying. Check minimum load, neutral requirement, and sensing coverage. Match the sensor to the space and the budget—not the other way around.
- Test one bay before wiring the whole floor. Use the actual fixtures—not similar fixtures. If the plan calls for 6-inch recessed luminaires, don't test with 4-inch (or vice versa).
- Consider the environment. For Zigbee controls, check the humidity and the wall construction of the room as seriously as you'd check the electrical circuit.
- Default to the certified combination. When a deadline is tight, choose the Legrand system combination that has already passed our compatibility testing. It may cost a bit more; the certainty is worth the premium.
And if you're already on site with a misbehaving sensor? Don't send it back immediately. First, question the load. Question the trim and spotlight covers around it. Question the humidity. Most of the time, the switch is telling you something true about the installation—it's just not saying 'replace me.'