Brentwood, New York · Nationwide Stocking Distribution [email protected] · 1-800-SATCO-US

Planning a Legrand Smart Lighting Project? Four Scenarios, Four Different Answers (And the Mistakes I Made in Each)

Why there's no single answer

Ask any honest question about smart lighting, and the answer almost always starts with "it depends." And the thing it depends on usually falls into one of four buckets. First, whether you're retrofitting existing switches. Second, whether you're adding door and window sensors. Third, whether you need custom logic that can only be delivered by Zigbee relays or Zigbee microcontrollers. Fourth, whether you're running LED strip wiring, which is where the splicing question comes up.

I handled smart lighting retrofit orders from 2017 through 2021—roughly 40 significant projects. I made avoidable mistakes on at least half of them. Those mistakes added up to somewhere around $4,800 in rework plus several weeks of lost time. And they were distributed with depressing regularity across the four buckets I just listed:

  • Switch box depth mismatch (September 2022, $890 in rework)
  • Battery-powered Zigbee sensor dropouts (2022, barely any parts cost but a serious credibility hit)
  • Zigbee firmware version incompatibility (Q1 2023, $2,800 in extra gateways and modules)
  • LED strip voltage drop on a 16-foot splice (2021, $640 in rework)

So rather than looking for one answer, find which scenario you're in. That's what actually saves time and money.

Scenario one: You're retrofitting existing switches

This is how most projects start: replacing conventional switches with something smarter. You're basically putting Legrand adorne or radiant devices into existing electrical boxes. For a lot of jobs, this is the easy part. But "easy" comes with preconditions.

In September 2022, I ordered 30 Legrand adorne light switch units (dimmer versions) for a commercial retrofit. They arrived, and we started installing. Eight of them—all on the third floor—physically didn't fit. The switch box depth (the side templates eat into the box interior) was shallower than what the adorne template needs. We pulled eight boxes, reordered radiant models, and added box extensions in two locations.

Total damage: $890 in rework and a six-day delay on a job that was otherwise on schedule. All because nobody took a tape measure to a box before ordering 30 identical switches.

What actually works: before you order, do these three checks:

  • Measure one box interior depth (not just the mouth—the side templates narrow the usable interior).
  • Check neutral wire availability. Most smart dimmers need a neutral. If the wiring dates before about 1980, there's a decent chance you don't have one. And an adorne switch won't operate without a neutral unless you add wiring.
  • Check boxes on different floors or in different parts of the building. Box depth isn't uniform.

Here's the counterintuitive bit: the more visually refined adorne series isn't the most forgiving to install. It needs deeper boxes. The radiant series is more flexible in tighter applications. Both work fine, but your existing box depth might favor one over the other. And once you've settled on a series, pairing it with the Legrand IoT Connect gateway streamlines the Zigbee pairing step—though the older switch boxes still have the final say on whether you can actually mount the thing.

Scenario two: You're deploying door and window sensors

Next up: the Legrand smart door and window sensor. This is a battery-powered Zigbee device that reports open/closed status to a gateway and can trigger HVAC changes or lighting automation. Sounds straightforward until you try to mount one between metal door frames.

In 2022, we installed 18 of these sensors across a 12,000-square-foot office building. Two of them—both on the same corridor—kept dropping offline every few days. The gateway showed them as present. But automation calls just didn't fire. We eventually found out that both sensors were 32 feet from the gateway with two steel door frames between them. Heavy-duty frames, the kind built for fire-rated doors.

The mistake I'd made: I assumed the Zigbee mesh would cover the distance. But a mesh needs router nodes to route. Battery-powered door/window sensors are Zigbee end devices—they don't forward traffic. Per the Zigbee Alliance's public documentation, only mains-powered devices (smart plugs, switches, relays) extend the mesh. Nobody puts that on the spec sheet.

The fix: we added a Zigbee smart plug in the middle corridor (roughly $40 as of early 2025 pricing). That was it. Problem solved in a few minutes. But the diagnosis had eaten a couple of visits.

What actually works:

  • Map gateway location and sensor locations before installation. Flag every distance and every obstruction.
  • Plan for a mains-powered Zigbee router (a smart plug or a relay module) at every physical barrier crossing. Battery devices won't pass signal through anything.
  • Don't trust the claimed range. Trust the physical geometry of the building.

I'm not a licensed electrical engineer, so I can't speak to code compliance on running any of this in long reaches. What I can tell you from a project-management perspective is that the physical topology matters more than the protocol. Battery sensors don't extend networks. Plan accordingly.

Scenario three: You need custom automation logic

When your application goes beyond what the gateway's native app does—say, tying lighting to an external scheduling system, or triggering off conditions more complex than simple time/sensor logic—you're in scenario three. This means working with a Zigbee relay, Zigbee microcontrollers, or both.

Q1 2023: we bought 30 Zigbee relay modules for a conference room project that needed lighting to coordinate with a room-booking system. Three of them worked with our existing coordinator out of the box. The other 27 needed specific firmware versions, and one required a border gateway we hadn't budgeted for. Filling that compatibility hole ran us another $2,800 and three weeks of back-and-forth.

Here's the problem: not everything labeled "Zigbee" actually talks to everything else. Zigbee 3.0 is the protocol layer, but "revision 3" is the application layer. There are also custom profiles, manufacturer-specific firmware, and modules that ship with locked firmware. They all say "Zigbee" on the listing, but only after you've burned a weekend do you find out they're speaking different dialects.

Honestly, I'm not sure why some Zigbee modules simply refuse to pair. My best guess is locked firmware or application-layer profile differences. If someone has insight on this, I'd genuinely like to hear it.

On Zigbee microcontrollers (the TI EFR32 or Silicon Labs types): if you're doing custom logic, these boards need firmware flashed before they're useful. The typical pattern is a coordinator or border gateway, then custom endpoints built from the microcontrollers. The learning curve is steep. Unless you have a real development team, I'd start with turnkey relay modules rather than going the bare-module route.

This was true in 2015 when most gateways were closed and picky. Today, Zigbee 3.0 interoperability has improved a lot, but "improved" isn't the same as "plug and play." Check firmware versions before you buy. Check them before you commit the project to a schedule.

The bigger point: this scenario is overkill for some projects and the only viable option for others. If you just need a phone-controlled group of lights, you don't need a custom Zigbee relay or microcontrollers. If you need scheduling, zoning, and energy strategies across a building, this might be the only way to get there.

Scenario four: Wiring LED strips

And now the question from the title—can you splice LED strip lights? Technically yes. Practically, how far you can splice and how you should splice depend on voltage, controller, and wire gauge.

In 2021, we did a 40-light run. To reach a decorative soffit in the far corner, we chained six lights onto a 16-foot run from the primary feed. The first 34 lights were perfect. The last six were noticeably dimmer—about half brightness. The client spotted it immediately, which is the worst possible outcome.

The fix: we added a second feed point at roughly 12 feet with heavier gauge wire to reduce resistance. The rework cost was $640, plus a weekend of chasing the problem.

What actually works: For 12V LED strips, brightness drop-off becomes visible around 2–3 meters (roughly 6.5–10 feet). For 24V strips, it's around 5–6 meters (roughly 16–20 feet). Those numbers are copper-resistance limits, not strip-quality issues. Even expensive strips dim if you push them too far. Using 10 AWG or 12 AWG for the feed line extends those distances modestly, but don't expect to double them.

On splice points: solder and heat shrink is the most reliable connection. If you're using quick-connect terminals, verify the terminal's current rating against the strip's load. A marginal connector is just a service call waiting to happen. And if you're connecting strips underwater or in a wet location, use the manufacturer's wet-rated connectors—standard solder joints won't last.

One more thing—if you're driving the strip with a Legrand Zigbee controller, the controller itself is fine as a switching mechanism. It outputs a steady voltage and the strip behaves like any other 24V or 12V load. But the controller-to-strip current rating needs to match. Too much controller for a short strip is usually fine, but too little for a long run gives you exactly the dimming problem I just described.

Splicing itself is unrelated to Zigbee, by the way. The Zigbee relay handles on/off and dimming, not splicing. But if you're mixing the two in a project, you're back in scenario three territory—be careful with mixing and matching.

How to figure out which scenario you're in

Here's the quick checklist:

  • Measure your existing boxes. Depth plus neutral availability. That determines how smoothly scenario one goes.
  • Map every sensor location. Count how many need to cross metal frames or thick walls. Plan a mains-powered router at each crossing. That's scenario two.
  • List the specific logic you need. If it goes beyond on/off/dim/timer, you're likely in scenario three. Check that your coordinator actually supports the specific relay firmware version you're buying.
  • Calculate your strip length. From power feed to the last light, at 12V and at 24V. What's the actual length? Where will you need a re-feed? That's scenario four.

Most projects touch two scenarios at once. That's fine. Just separate them and handle them independently. Twenty minutes of this kind of checking routinely saves a week of rework—that's the practical version of efficiency improvement. Not fancier processes. Just fewer things going back out the door.

And if you're genuinely unsure: measure the box, count the sensors, calculate the strip length. The "it depends" answer works because the details dictate the approach. The details don't lie.