September 5, 2026

The brief from the brand partner was two sentences long, and it changed everything about how we'd approach this product: "We want a mosquito killer lamp. It must connect to our smart lighting main line — not WiFi, not a standalone app. The main line."

Our optical engineer looked at the project lead. The project lead looked at the fluid dynamics team. Nobody said anything for a few seconds, because we all understood what those two sentences actually meant. This wasn't a request to build another UV mosquito trap with a fan strapped to it. This was a request to engineer a smart pest control device that could live inside an existing smart lighting infrastructure — and that constraint would end up shaping every technical decision that followed.

This is the design diary for the MW-120. Not a spec sheet walkthrough. A record of the thinking, the trade-offs, and the moments where we had to abandon assumptions and start over.

The Ecosystem Constraint Changed Everything

Most smart mosquito killer lamp designs start with a simple question: "How do we kill mosquitoes effectively?" Then, as an afterthought, someone asks about app connectivity. The result is usually a product with a bolted-on WiFi module, a generic companion app, and a user experience that feels disconnected from everything else in the room.

Our brand partner flipped that sequence. The smart lighting main line was non-negotiable — it was the platform, and the electric mosquito zapper lamp had to serve the platform, not the other way around. That single constraint forced us to think about the product differently from day one.

A lighting main line is a wired, always-on infrastructure designed for consistent power delivery and signal integrity. It's not designed to handle the high-voltage noise from an 800VDC electric grid, and it certainly isn't designed to accommodate a UV LED array that draws its own current profile. Every subsystem in the MW-120 had to be designed around this reality: power consumption within the main line's tolerance, electromagnetic compatibility with the lighting control signals, and a physical form factor that could mount alongside existing luminaires without looking like a medical device someone hid behind a sofa.

Phase 1: The Wavelength Question

The first real argument happened in the optical team's lab. Mosquito attraction through UV light is well-documented in entomological literature — the insects are sensitive to wavelengths in the 365–400nm range. But "well-documented" and "production-reliable" are separated by a canyon of engineering details.

Our initial instinct was to narrow the emission band as tightly as possible around 365nm, where research suggested peak mosquito photoreceptor sensitivity. The optics team built a prototype with narrow-band UV LEDs centered at 365nm. In the lab, it performed well. In the field test chamber with live Culex pipiens, it performed even better than expected.

Then we ran the cost analysis.

Narrow-band 365nm LEDs at the optical power we needed were expensive — not prohibitively so for a single unit, but at the price point our brand partner was targeting for Southeast Asian markets, the BOM didn't work. We went back to the literature and found something interesting: mosquito species common in tropical and subtropical regions (the actual target markets) show attraction sensitivity across a broader 365–400nm band, not just at the 365nm peak.

This changed the calculation. Instead of narrow-band LEDs, we could use a broader-spectrum UV array covering 365–400nm (±10% tolerance). The LEDs were more readily available, significantly cheaper, and — here was the unexpected benefit — the broader spectrum actually attracted a wider range of mosquito species, including Aedes aegypti, which is more selective about wavelength than Culex.

The lesson: the "optimal" engineering solution (narrowest possible band) wasn't the right solution for this product. The right solution was the one that optimized across attraction performance, component cost, and species coverage simultaneously. We locked the wavelength range at 365–400nm and moved on.

Phase 2: The Airflow-versus-Noise Problem Nobody Warns Us About

If the wavelength decision was an argument, the airflow design was a three-month education in why fluid dynamics is never straightforward.

The UV mosquito trap literature is clear: you need sufficient suction to pull flying insects off their trajectory and into the capture chamber. For mosquitoes, which are relatively weak fliers compared to moths or flies, the threshold sits around 34–35 CFM of effective airflow at the intake. Below that, larger species simply resist the suction and drift away. Our target became ≥34.5 CFM at maximum fan setting.

Achieving 34.5 CFM from a compact impeller is a fluid mechanics problem, not a "bigger fan" problem. The MW-120's total diameter is constrained to ≤150mm × 150mm — that's the physical envelope the brand partner's industrial design team had allocated. Inside that cylinder, we needed to fit the UV array, the fan module, the high-voltage grid, the collection mesh, and the control electronics. The fan couldn't be bigger. It had to be more efficient.

Our fluid team iterated through four impeller geometries before arriving at the final blade profile. The winning design used a swept-back blade with a variable pitch angle — more aggressive at the hub for pressure generation, more gradual at the tip for volume displacement. At 12VDC ±15%, drawing no more than 0.5A, the impeller delivers the target airflow at maximum speed.

Then the noise testing started, and it humbled us.

At full speed, the impeller hit the airflow target easily. It also produced noise that would have made the MW-120 unsuitable for bedroom use. We were seeing readings above 55dB at one meter — and the spec required ≤50dB. This is the moment where most products in this category quietly compromise: they ship with a louder fan and hope the user runs it on a lower setting.

We didn't want to do that. So we went back to the motor mounting. The noise wasn't coming from the blades themselves — it was coming from vibration coupling between the motor housing and the lamp body. A 3dB reduction came from redesigning the motor mount with a silicone-damped isolation ring. Another 2dB came from adjusting the blade tip clearance to reduce tip vortex noise. Combined, we got the maximum-speed noise down to ≤50dB@1m.

The trade-off: the isolation ring added 3mm to the total height and required a custom mold. Worth it. A smart pest control device that works but keeps you awake isn't a solution — it's just a different problem.

Phase 3: 800 Volts in a 150-Millimeter Enclosure

The high-voltage grid design was the subsystem that kept everyone up at night — not metaphorically, but because safety review meetings kept running late.

The grid needs to deliver ≥800VDC to ensure instant lethality on contact, even for larger species and even as residue accumulates on the mesh over hours of operation. 800VDC in an open field lab setup is straightforward. 800VDC inside a sealed 150mm-diameter enclosure that carries IP21 protection, sits within arm's reach, and shares infrastructure with a low-voltage smart lighting system — that's a completely different engineering problem.

Our high-voltage team's first challenge was the boost converter topology. The MW-120 runs on 12VDC, so stepping up to 800VDC requires a DC-DC boost stage with very specific characteristics: it must be efficient (the total system budget is ≤0.5A at 12V), compact (the 150mm envelope), and — critically — it must not inject switching noise back into the lighting main line's control signals.

We went through two converter topologies before settling on a flyback design with integrated snubber. The snubber was non-negotiable: without it, the high-voltage switching transients coupled into the lighting control bus and caused flicker in connected luminaires. This was one of those problems you can't see on a spec sheet — you only discover it when you integrate the full system and watch the lights flicker.

The mesh geometry was equally constrained. The grid spacing needed to be small enough to prevent finger contact (IP21 requirement: no access to hazardous voltages with a standard test probe), but large enough to allow airflow through to the collection chamber. We ended up with a concentric ring design — inner ring at positive potential, outer ring at ground — with 6mm gaps. The gaps are too narrow for a child's finger but wide enough for air to pass through with minimal resistance.

The IP21 rating also meant designing the enclosure to prevent water drip ingress. This sounds simple until you realize the collection mesh needs to be removable and washable — meaning there must be a user-accessible opening that also maintains IP21 integrity when reassembled. We solved this with a bayonet-mount mesh tray with a silicone gasket that compresses on closure. It clicks when it's properly sealed. No tools required.

Why the Main Line and Not WiFi

This is the design decision we get asked about most often, and it deserves a direct answer.

When we started the MW-120, the industry trend was clear: add WiFi, build an app, call it "smart." WiFi modules cost less than they did three years ago. Cloud platforms are essentially free to set up. The perceived barrier to a connected product has collapsed.

But we kept coming back to the use case. A mosquito killer lamp runs at night, in bedrooms, often for 8–10 hours straight. Its primary user interaction is: turn it on, leave it alone, clean the mesh every few days. How many WiFi-connected products in a bedroom is too many? At what point does "smart" become "another device competing for bandwidth and emitting radio signals near where someone is trying to sleep?"

The smart lighting main line, by contrast, is already there. It's wired, reliable, and doesn't add to RF congestion. The brand partner's existing main line already supports wired controller and remote control — so the MW-120 inherits those control interfaces for free. Users who already have a remote for their smart lights use the same remote to control the electric mosquito zapper lamp. No new app, no new pairing process, no new device on the router.

This decision also simplified the EMC (electromagnetic compatibility) picture enormously. A WiFi radio inside a device that contains an 800VDC boost converter is an EMC nightmare — the switching noise interferes with the antenna, and the antenna's harmonics can couple into the high-voltage circuit. By using wired communication through the main line, we eliminated an entire category of potential failures.

The MW-120 is, by design, a smart pest control device that's smart in the ways that matter — responsive to user control, integrated with existing infrastructure — without being "smart" in the ways that create complexity for its own sake.

From Prototype to KC and PSE Certification

The certification phase is where a product's real engineering quality gets stress-tested. The MW-120 needed both KC (Korea) and PSE (Japan) marks — two of the more demanding certification regimes for consumer electronics in Asia, particularly for products combining UV emission and high-voltage circuits.

The first round of testing surfaced an issue we hadn't caught in-house: under sustained operation at 40°C ambient with 80% RH, the high-voltage boost converter's efficiency dropped enough to push total current draw above our 0.5A budget. In our Shenzhen lab, ambient conditions rarely exceeded 32°C during testing. The certification lab in Seoul ran at 40°C. This was a thermal design oversight, not an electrical one.

We addressed it by repositioning the boost converter's heat-generating components away from the UV LED cluster — which was radiating its own thermal load — and adding a copper thermal via array on the PCB to spread heat more evenly across the board. The modification brought the high-temperature current draw back under 0.5A, but it also taught us something about our testing methodology: we'd been validating against Shenzhen's climate, not the product's actual operating environment of 5–45°C with humidity up to 80% RH.

The PSE review required additional documentation on the UV emission spectrum verification — confirming that the 365–400nm band stayed within ±10% tolerance across the full operating temperature range. This wasn't a surprise; Japanese certification is meticulous. But it did require us to add a UV calibration step to the production testing sequence that we hadn't originally planned.

Both certifications were achieved on the third submission cycle. The ≥1,000-hour lifespan rating was validated through accelerated life testing. The IP21 rating was confirmed via standard ingress testing. And the operating temperature range of 5–45°C, humidity ≤80% RH, was verified across both certification bodies' test conditions.

What This Product Taught Us

Every product teaches the team something the next product didn't know it needed. The MW-120 is no exception. Three lessons stood out.

Constraints are design tools. The smart lighting main line constraint felt like a limitation at first. In practice, it forced us to make decisions we otherwise would have deferred — about power management, EMC, control interfaces — and the result was a more cohesive product. The best products we've worked on at Angel Fluid have always been shaped by constraints that seemed uncomfortable at the outset.

Cross-domain integration is the actual product. The MW-120 isn't a UV light with a fan and a grid. It's the integration of optical design, fluid dynamics, high-voltage engineering, and smart lighting protocols — all within a 150mm envelope. The individual subsystems are each well-understood engineering problems. The integration is the hard part. Getting optical, fluid, and high-voltage teams to design in concert — not in sequence — is what separates a functional product from a great one.

The maintenance experience is part of the product experience. The removable and washable mesh collection tray sounds like a minor detail. But in this product category, maintenance is the moment where the user physically interacts with the product's engineering quality. If the tray is hard to remove, hard to clean, or doesn't seal properly when reinstalled, the user's entire perception of the product shifts. We spent disproportionate design effort on the bayonet-mount tray because that's where the user's hand goes every week.

The Team Behind the MW-120

Angel Fluid is a precision fluid control and intelligent electronic solutions ODM based in Shenzhen. Our engineering team spans optical design, fan dynamics, high-voltage circuitry, firmware, and system integration. Our Shenzhen engineering team oversees the full chain from prototype development through testing — full-process quality control from concept to certification.

The MW-120 represents our approach to product development: start with the hardest constraint, solve the cross-domain integration problems early, and never treat maintenance as an afterthought. If you're developing a smart mosquito killer lamp or UV mosquito trap and need a partner who thinks in systems rather than subsystems, we'd welcome a technical conversation.

Let's talk about your next project.