Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Conformal Coating for Reliability
Transcript
- Lucas: So, Luna, you know that moment when a hardware startup gets their first batch of field returns? The boards come back corroded, maybe with some white crust around the pins. And the fix is actually something they could have designed in from day one—a thin layer of conformal coating. Luna: Conformal coating. That's the plastic-like film you see on some circuit boards, right? I always thought it was a nice to have, not a must. Lucas: Exactly. And that's the angle for today. A lot of founders treat it as an afterthought, but if you're building anything that lives outside, or even in a humid garage, it can be the difference between a product that lasts and one that dies in months. Before we go deep, though—you know we keep this show ad-free, and that's a choice. If you find these conversations useful for what you're building, the link is buy me a coffee dot com slash fexingo. That's the only way we keep this going. Luna: Yeah, and it genuinely matters. It's how we stay independent and keep bringing you these niche manufacturing details. Anyway, so what's the real-world failure mode? Lucas: Right. So picture an IoT sensor for agricultural monitoring. It's mounted on a fence post, exposed to rain, dust, and temperature swings. Without coating, moisture gets into tiny gaps around the microcontroller pins, and electrolysis starts eating the traces. I read a case study about a startup that had a 30 percent failure rate in the field within six months. They switched to an acrylic conformal coating, and that dropped to under two percent. Luna: Thirty percent to two. That's a huge swing for something that costs pennies per board. So what's the catch? Lucas: The catch is choosing the right material and application method for your product. There are three main chemistries: acrylic, silicone, and polyurethane. Acrylic is the most common for startups because it's easy to apply and, crucially, it's reworkable—you can remove it with solvent if you need to fix a board. Silicone handles higher temperatures and more flexibility, but it's harder to remove. Polyurethane is the toughest, great against solvents and abrasion, but again, not fun to rework. Luna: So acrylic is the default for prototypes and first production runs. When would you step up to silicone or polyurethane? Lucas: Good question. If your device is near a motor or an engine, you're dealing with heat and vibration, so silicone's flexibility and temperature range win. If it's going into a chemical plant or anywhere with aggressive solvents, polyurethane is your friend. But for most outdoor consumer or industrial IoT, acrylic does the job. And you can also get away with a thicker coating if you're not concerned about rework. Luna: Now, application matters too. I've seen people just spray it on with a rattle can. That's fine for a prototype, but for production, what are the options? Lucas: Right. So there's spray, dip, and selective coating. Spray is fast and cheap but can miss underside components and might not give you uniform coverage. Dip coating immerses the whole board, so you get full coverage, but you have to mask off connectors and test points. Selective coating uses a robotic nozzle to apply coating only where you want it—that's the best for boards with mixed components, but it's more expensive and has a longer setup. Luna: And I'm guessing most startups start with spray and then move to selective as volumes go up. But what about the masking? That's a pain point. Lucas: Exactly. Masking is where a lot of time gets burned. You have to cover every connector, switch, and test pad with a peelable mask or tape. That's manual labor, and it's error-prone. Some startups design their boards with coating-friendly footprints from the start—like using castellated pads or keeping sensitive areas clear. That way, you can skip masking altogether if you're selective. Luna: So it's a design for manufacturing decision, not just a process step. You have to think about it before you lay out the board. Lucas: Right. And that's where the cost really comes in. The coating material itself is cheap—maybe a few cents per board. But the labor for masking and the yield loss from bad coverage can add up. If you're doing low volumes, spray might be fine. But once you hit a few thousand units a month, selective coating pays for itself in labor savings and consistency. Luna: We should talk about testing. How do you know your coating is actually doing its job before you ship? Lucas: Great question. You don't want to wait for field failures. So you do accelerated life testing—we touched on that a few episodes back. For conformal coating specifically, you'd run temperature cycling from minus forty to plus eighty-five Celsius, with humidity. You can also do a salt-spray test if your product will be near the ocean. The key metric is insulation resistance—you want to see high resistance between adjacent traces after the test. Luna: And you can do that in-house with a simple chamber and a multimeter. You don't need a full lab for basic validation. Lucas: Exactly. A benchtop environmental chamber and a megohmmeter will get you a long way. But you also need to test for adhesion. If the coating doesn't stick well, it can peel off during thermal cycling. A simple tape test—like the crosshatch test—can tell you if your surface prep is good enough. And that's another thing: surface cleanliness. You have to clean the board thoroughly before coating. Any flux residue or grease will ruin adhesion. Luna: So what's the biggest mistake you see startups make with conformal coating? Lucas: I think it's underestimating the cure time and the process control. Some coatings take 24 hours to fully cure, and if you rush it, you get tacky surfaces that attract dust. And if you're using a spray booth, you need consistent temperature and humidity. A lot of startups treat it like paint and just go for it, but it's more like a precision process. Getting a coating vendor involved early helps—they can recommend the right material and process for your specific board. Luna: Vendor involvement early. That's a recurring theme on this show. But what about rework? If you have to replace a component, how do you deal with the coating? Lucas: That's where acrylic shines. You can dissolve it with a solvent like isopropyl alcohol or a specialized remover. But it's messy, and you have to be careful not to damage the board. Silicone and polyurethane are much harder to remove—you often have to scrape or use a chemical stripper. That's why many startups stick with acrylic until they're confident in their design and don't expect many rework cycles. Luna: So for a low-volume startup, acrylic is the safe bet. And for high-volume, you might invest in selective coating and accept the rework difficulty if you're confident in your design. Lucas: Right. And there's also the option of using a coating that's both moisture-resistant and uv cureable. Some newer acrylics cure in seconds under UV light, which speeds up production. But they might not have the same long-term track record. So you have to balance speed with proven reliability. Luna: I've also heard about parylene coating—that's a vapor-deposited film, isn't it? Why isn't that more common? Lucas: Parylene is a great coating—it's incredibly thin and uniform, and it provides excellent moisture and chemical resistance. But it requires specialized vacuum deposition equipment, which is expensive and typically done by a coating service. For a startup, the cost per board can be ten to twenty times higher than acrylic spray. So it's usually reserved for medical implants or aerospace, where the performance is worth it. Luna: So the cost is the barrier. But if you're making a high-margin product with strict reliability demands, it might be worth exploring. Lucas: Absolutely. And that's the kind of decision you make with your manufacturing partner. You don't have to figure it all out alone. A good contract manufacturer will have experience with conformal coating and can guide you. But you need to ask the right questions—like what's your typical coating thickness, and how do you inspect for coverage? UV inspection with a fluorescent dye is common, but you have to request it. Luna: Inspection is another thing startups overlook. You can't just assume the coating is there because it looks shiny. Lucas: Right. And that's why you see some startups invest in an automated optical inspection, or AOI, system that can detect coating voids. Again, that's an investment, but if you're scaling, it's worth it. Actually, let's bring it back to a concrete example. I remember a startup that built a smart lock for outdoor use. They had a high failure rate due to humidity. They switched to a silicone conformal coating and added a conformal-coated antenna—that didn't work, because the coating interfered with the RF signal. So they had to mask the antenna area or use a selective coating process. Luna: That's a great point. Coating can affect more than just corrosion resistance—it can mess with radio performance. So you have to think about what's on your board before you coat everything. Lucas: Exactly. And that's why you need to test with the coating applied, not just bare boards. Some coatings have a high dielectric constant at high frequencies, which can detune your antenna. So you might need to adjust your matching network or leave a keep-out area. That's a design consideration you have to plan for from the start. Luna: So it's not just about slapping on a coating—it's about integrating it into your overall design and test plan. But for a startup, how do you even get started with conformal coating if you're doing your own assembly? Lucas: Start small. For prototypes, you can buy acrylic spray cans from electronics distributors and hand-apply with a brush or foam swab. That's fine for a few boards. But as soon as you go to pilot production, talk to your CM or a coating service. They can recommend the right material and process for your volume. And don't forget to include the coating step in your BOM and your cycle time estimates—it's not just a material cost, it's labor and time. Luna: And it's also a lead time thing. If you're outsourcing coating, that adds turnaround time. So you have to schedule it into your production plan. Lucas: Right. And that's why it's smart to design your product to be coating-friendly—minimize the number of components that need masking. Use board-level shields or connectors that don't need coating. That can save you hours of manual work per batch. And it also reduces the risk of human error. Luna: So the key takeaway is: conformal coating is a low-cost, high-impact reliability measure that needs to be considered early in the design phase. The material choice depends on your environment and rework needs, and the application method scales with your volume. Lucas: Exactly. And the result is a product that survives the real world, which is what every hardware startup wants. I'd also add that you should always test your coating with your actual process—not just the material. The same acrylic can behave differently depending on how it's applied and cured. So validate with your specific board and your specific application method. Luna: That's a good note to end on. Conformal coating might not be the most exciting topic, but it can save your startup from a lot of pain. Thanks for listening, and we'll catch you next time. Lucas: Actually, before we go, one more thought: if you're between building and scaling, think about what field failures would cost you in terms of warranty returns and reputation. A few cents of coating is a cheap insurance policy. And if you have a story about coating saving your product, we'd love to hear it—reach out on the Fexingo community. See you next week.