Why Your LED Bulbs Keep Dying (and It’s Usually Not the LED)

Why Your LED Bulbs Keep Dying (and It’s Usually Not the LED)

If you’ve ever bought a “lasts 10 years” LED bulb and had it die before you even finished living with the paint color you chose, you’re in very good company. We’ve heard the complaints from homeowners who feel duped, from builders who are sick of explaining it, and from electricians who didn’t sign up to be the official ladder crew for the next decade.

That’s why our recent BS & Beer conversation with lighting designer and educator Dave Warfell landed the way it did. He didn’t come in with a hot take or a gimmick. He just explained the system mismatch in a way that made all of our collective LED frustration suddenly feel… predictable.

Because the LED chip usually isn’t what fails.

The driver is.

If you want the full context and the best version of the story (plus plenty of laughs and some very honest trade talk), watch the episode here: BS & Beer with Dave Warfell

The uncomfortable truth about “LED bulbs”

A typical home is built around 120 volts. That’s not arbitrary and it’s not “old fashioned” for the sake of it. It’s simply the infrastructure that made sense when incandescent bulbs were the only real game in town, and it’s the infrastructure we’ve standardized around ever since.

LEDs changed the light source, but they didn’t change the house.

Dave’s key point is one of those details that sounds small until you realize it explains a lot. The actual LED chips that produce light don’t want 120 volts. They run on a few volts, usually in the neighborhood of 1.5–4V. Call it 3V for the sake of a clean mental model.

So every time we screw an LED “bulb” into a standard socket, we’re asking a 120V system to feed a 3V technology. That only works because there’s a conversion step baked in.

And that conversion step is where the trouble starts.

The driver is doing the hard work, and it’s the part we cheaped out on

The driver is the little power component that takes line voltage and converts it into something the LED can actually use. You’ll hear it called a driver, a transformer, or a driver/transformer combo, but the job is always the same. It’s the translator between a house built for incandescent and a technology that’s inherently low voltage.

In a perfect world, that driver would be robust, efficient, and easy to service. It would live somewhere that acknowledges reality: electronics fail, and we shouldn’t have to damage finishes to replace them.

But residential lighting rarely gets that perfect world. What it usually gets is a tiny driver crammed into the base of a bulb or into a compact fixture housing, because that’s what lets manufacturers hit the price point and the form factor the market expects. It has to look like a “normal bulb.” It has to feel like a commodity purchase. It has to be cheap enough that you’ll toss it in a cart without thinking too hard.

That tiny driver also tends to live in a warm, tight space, which is not exactly the environment you’d choose if longevity was your number one goal.

So when that bulb dies early, we all say, “the LED burned out,” even though the diode itself may be fine. Dave’s framing was simple and honestly pretty relieving: a lot of “LED failures” are actually LED driver failure.

Why this keeps biting us on real projects

Once you see it as a systems problem, the pattern feels almost obvious. We’ve built a lighting model that relies on thousands of miniature conversion devices spread all over a house, and then we’ve buried many of them in places that are annoying to reach and expensive to deal with.

That’s why the pain doesn’t show up as one big catastrophic failure. It shows up as a slow drip of little failures. A bulb here. A fixture there. A flicker that only happens when the dimmer is set at exactly the wrong level. A room where the “same” bulbs look slightly different because the replacement was a different batch. A homeowner who’s not mad at anyone in particular, but is definitely mad.

If you’ve done building science work, this will feel familiar. The problems rarely happen in the middle of the assembly when everything is continuous and well-supported. They happen at the transitions. At the interfaces. At the spots where we’re relying on one small component to bridge two worlds.

Lighting is no different. The driver is the transition layer. We’ve just been pretending it’s invisible.

Where low voltage changes the whole story

Dave’s broader argument wasn’t “buy nicer bulbs.” It was “stop forcing a low-voltage technology to behave like an incandescent.” When he talks about low voltage lighting, he’s talking about systems that are designed around 48V or less, which opens the door to a more sensible architecture.

The big shift is where the drivers live.

In many low-voltage lighting setups, drivers are centralized or at least placed in accessible locations, which turns a future failure into a manageable service task instead of a ceiling surgery. That single decision is what takes you from disposable to maintainable. It also changes how confident builders and electricians feel recommending a lighting approach, because nobody wants to install something that guarantees frustration later.

This is the part we wish more people would talk about. A lot of the “LEDs are junk” narrative is really “we designed this in a way that makes maintenance miserable.”

Low voltage doesn’t magically solve all lighting problems, but it often solves the one that makes everyone hate the whole category.

The fear is real: “We tried a system once, and it was a nightmare.”

We’re not going to pretend this is an easy sell, especially if you’ve been burned before. The controls side of lighting has a long history of big promises, weird compatibility issues, and apps that disappear faster than a limited-run IPA.

We talked about it openly on the episode because it’s the reason many tradespeople hesitate. If you’ve ripped out a “smart” system that stopped being supported, you remember the look on the homeowner’s face. At that point nobody cares about innovation. They care about whether the switch works.

Dave’s response was pretty grounded. He acknowledged the mess, pointed out that interoperability is improving (especially with standards like Matter gaining traction), and emphasized that the ecosystem is maturing. It’s still not as plug-and-play universal as 120V lighting, but it’s moving in the right direction and the industry is starting to look more stable than it did even a few years ago.

The practical implication is this: the risk isn’t low voltage itself. The risk is treating lighting and controls like an afterthought, then hoping it all magically behaves.

The one question we think every lighting plan should answer

If you’re building, remodeling, designing, or even just trying to stop replacing bulbs like it’s a hobby, ask this early and ask it out loud:

Where do the drivers live, and how do we service them?

It’s a deceptively simple question, and it tends to expose whether the lighting approach was actually designed or just selected. If the answer is “inside every fixture,” that doesn’t mean the project is doomed, but it does mean you’re accepting a future where a failure is harder, messier, and often more expensive to deal with.

If the answer is “in an accessible location,” you’ve already improved the long-term story. Even if everything else stays the same, you’ve shifted the system toward maintainability, which is where the real wins show up over time.

High-friction reality: what people get wrong (and what it costs later)

This is where we’ll be a little blunt, because the easiest way to keep getting bad outcomes is to keep pretending lighting is a simple line item.

What usually goes wrong isn’t one dramatic mistake. It’s a stack of small decisions that don’t look risky on paper, but add up to frustration in the field.

• Products that look identical at install but have very different driver quality and thermal management

• Drivers buried where nobody can reach them without damaging finishes

• Dimmers, drivers, and LEDs mixed together without a tested plan, leading to flicker or weird behavior

• Supply chain substitutions that change performance without anyone noticing until the client complains

• A general assumption that “if it fits and it turns on, we’re done”

If you’re doing high-performance work, you already know the moral of that story. The easy choice up front often becomes the expensive choice later. Lighting is just catching up to the same reality.

Who this approach is for (and who it’s not)

Low-voltage lighting and better driver strategies tend to make sense when the goal is fewer callbacks, better long-term performance, and a home that feels good to live in, not just bright enough to pass a walk-through.

This is a great fit if you’re:

• building higher-performance homes and want fewer callbacks

• trying to reduce maintenance risk in finished ceilings

• designing layered lighting that looks intentional, not just “more lumens”

• working with clients who care about comfort, durability, and long-term value

If the job is strictly lowest first cost, fastest install, and minimal coordination, you can absolutely stick with commodity fixtures and call it a day. Just be honest about what that usually buys you down the road.

A low-risk way to start without betting the whole house

One of the most practical suggestions from the conversation was to pilot this in a place you control. Your own house is the obvious candidate, but we also love the builder office or shop as a test lab. It gives you a place to learn the new language, make the mistakes, and come out the other side with a repeatable approach.

That’s how a lot of building science improvements spread. Not because everyone suddenly agrees at a conference, but because enough people try it, get their reps, and stop going back once they see the better option.

Want to go deeper with Dave?

Dave Warfell is doing the work to make residential lighting less confusing and more accessible, which we’re fully on board with.

• Dave on LinkedIn: David Warfell

• His site: Light Can Help You

• The full episode: BS & Beer with Dave Warfell

Bottom line

LEDs aren’t the villain. The way we’ve forced LEDs to live inside a 120V legacy system, then minimized and cheapened the conversion hardware, is what creates a lot of the pain that people blame on “LEDs” as a category.

If you’re tired of “forever” bulbs that don’t last, stop treating the driver like an invisible detail. Put it somewhere you can access, choose compatibility on purpose, and design it like you actually expect it to be maintained.

FAQ

Why do LED bulbs burn out so fast?

Often the LED diode is fine. The driver inside the bulb fails early because it’s a small, heat-stressed power component built to hit a low price point.

What is an LED driver?

It’s the power supply that converts and regulates electricity so an LED can run safely and consistently. When it’s low quality or trapped inside an inaccessible fixture, it becomes a common failure point.

Is low-voltage lighting safer than 120V lighting?

Generally, yes. Lower voltage reduces shock risk, which is why many low voltage systems fall under different wiring rules. You still need competent design and installation because electricity is still electricity.

Will low-voltage lighting cost more?

It can. You may spend more up front on design coordination and better components, but you often save pain later through easier maintenance and fewer finish-damaging replacements.

Can I mix low-voltage lighting with standard 120V fixtures?

Yes. Most homes will be hybrid for a long time, especially with decorative fixtures. The key is planning voltage, drivers, dimming, and controls so you don’t end up with flicker or mismatched performance.

What’s the easiest first step to reduce LED failures?

Design for access. Even if you stay mostly in a 120V world, choosing fixtures and layouts that keep drivers serviceable can dramatically reduce future headaches.