Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Overmolding to Seal and Strengthen
Transcript
- Lucas: You ever pick up a smart outdoor sensor — maybe a soil moisture monitor or a security camera — and wonder how it survives rain, snow, and the occasional drop off a ladder? Luna: All the time. I assume it's a rubber gasket or some kind of conformal coating, but those solutions have weak points. Lucas: Exactly. Gaskets can shift during assembly. Conformal coating is great for circuit boards, but it doesn't protect connectors or cables. That's where overmolding comes in — encasing the whole assembly in a second layer of plastic that bonds directly to the first. Luna: So it's like giving your electronics a custom-fitted raincoat that's also armor. Lucas: That's the idea. Let's look at a hypothetical case: a startup building a smart irrigation controller that lives outdoors in a planter box. They need it to survive direct hose spray, summer heat, and a five-foot drop onto concrete. Luna: I've seen those fail in the field. Usually it's the cable entry point where water creeps in. Lucas: That's exactly the weak link. With overmolding, you can mold a second layer — typically a thermoplastic elastomer, or TPE — right over the housing and the cable exit. The material flows into every crevice and chemically bonds to the substrate. No gasket, no o-ring, no gap. Luna: Is it the same process as two-shot injection molding that we talked about in episode 42? Lucas: Related, but not identical. Two-shot molding uses two different materials injected sequentially into the same mold — usually for multi-color parts. Overmolding is a specific variant where the second material is applied over an already-formed substrate. That substrate can be a rigid plastic housing, a metal bracket, or even a fully populated circuit board. Luna: Right. And the bond is key. If the overmold doesn't stick, you get delamination and a path for moisture. Lucas: Exactly. The bond can be chemical — like using a compatible polycarbonate substrate with a TPU overmold — or mechanical, where the substrate has undercuts, holes, or texture for the second material to lock into. Most startups combine both. Luna: What's the cost? I imagine a custom mold for a two-shot process isn't cheap. Lucas: For a simple overmold tool — one cavity, family mold — expect thirty to eighty thousand dollars. That's per cavity. If you need a four-cavity tool for higher volume, you're looking at a quarter million. But the per-part cost drops dramatically: a two-shot part might be thirty cents in material versus fifteen cents for a single-shot part plus gasket assembly. Luna: So it only makes sense if you're making tens of thousands of units. Lucas: Generally, yes. But there are ways to prototype cheaply. You can 3D print the substrate and then use a lab-scale injection press to overmold a few hundred parts. Or use castable urethanes. That's how our irrigation controller startup validated their IP68 rating before committing to steel. Luna: IP68 — that's the highest consumer waterproofing standard, right? Submersible beyond one meter. Lucas: Right. They tested to three meters for thirty minutes. With overmolding, they passed on the first try. Without it, they were failing at the cable gland. Luna: What about drop protection? The raincoat analogy only goes so far. Lucas: Overmolding also adds shock absorption. A TPE overmold with a durometer of 70 Shore A — about the hardness of a car tire — can absorb a five-foot drop onto concrete. The rigid substrate underneath stays intact. One startup I know dropped their device from a ladder twenty times with no damage. Luna: But is there a downside? I imagine it makes repair almost impossible. Lucas: That's the big trade-off. Once you overmold, you can't get inside without destroying the overmold. So if a battery dies or a connector breaks, the whole unit is scrap. That's fine for a disposable sensor, but not for a modular system. Luna: So it penalizes repairability. For a startup that wants to be sustainable, that's a tough call. Lucas: It is. Some startups design a sacrificial overmold — a thin layer that can be peeled off, with the electronics in a sealed sub-unit that can be re-overmolded. But that adds cost and complexity. Luna: Let's talk about design rules. What are the non-negotiable guidelines for a first-time overmolding project? Lucas: First: keep the overmold thickness under three millimeters. Anything thicker adds cycle time exponentially — you're waiting for the material to cool, and that drives up cost. Second: avoid sharp corners in the substrate. Radius them to at least 0.5 millimeters so the overmold flows evenly and doesn't create stress risers. Luna: And material compatibility? That seems like the biggest gotcha. Lucas: It is. You can't just pick any TPE. You need a material that is chemically compatible with your substrate. For example, polycarbonate works well with tpu based overmolds, but nylon requires special adhesion promoters. Always do a peel test — mold a test coupon, then try to peel the overmold off. If it comes off clean, you have a problem. Luna: A peel test is basically a 180-degree pull, right? Like measuring the force to separate the layers. Lucas: Exactly. You want cohesive failure — meaning the overmold tears before the bond breaks. If you get adhesive failure, where the layers separate cleanly, your bond strength is too low. Luna: That's something you can test with a cheap lab setup before cutting steel. Smart. Lucas: And you should. Because once you cut that mold, changes are expensive. Luna: What about insert molding? Isn't that a similar concept? Lucas: Related but distinct. Insert molding places a pre-made component — like a threaded insert or a circuit board — into the mold, then injects plastic around it. Overmolding specifically applies a second layer over an existing molded part. Insert molding is often used for adding metal fasteners to plastic parts. Luna: Got it. So for our irrigation controller, they'd overmold the TPE over the rigid housing after the board is assembled inside. Lucas: Yes. But there's a nuance: you can overmold directly over a populated PCB if the components are rated for the injection pressure — usually around 30 to 100 megapascals. Most surface-mount components can handle it, but tall components like electrolytic capacitors or connectors might get crushed. Luna: So you need a housing to protect them first. That adds a step. Lucas: Typically, yes. The PCB goes into a rigid housing, then the housing is overmolded. That way the sensitive electronics are shielded from pressure and heat. Luna: I've heard of liquid silicone rubber overmolding for medical devices. Is that relevant to hardware startups? Lucas: LSR overmolding is a whole different beast. The material is injected as a two-part liquid that cures in the mold at high temperature. It's great for biocompatibility and high-temperature environments, but the tooling is even more expensive — often double the cost of thermoplastic overmolding. For most consumer hardware startups, TPE is the right starting point. Luna: Alright. Let's say I'm a listener building a rugged outdoor gadget. What's the one thing to take away? Lucas: If you need IP68 and drop protection, overmolding is the most reliable way to get there. But plan for the tooling cost and the repairability trade-off. And always, always validate material compatibility with a peel test before you commit to production tooling. Luna: I'd add: don't forget cycle time. A two-minute cycle per part adds up fast if you're making a hundred thousand units. Lucas: Great point. That's where design for manufacturing really pays off. Every second you shave off the cooling time saves thousands of dollars over the product run. Luna: If these conversations are useful for what you're building or running, that's exactly who we make them for. Lucas: Yeah. And honestly, if today's episode was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. No pressure, just keeps the show ad-free. Luna: Totally. Appreciate you listening. Lucas: So coming back to our irrigation startup: after they validated with a prototyped overmold, they found that the cable relief was actually the hardest part to get right. Luna: Because the cable creates a leak path along its length, right? Lucas: Exactly. The overmold has to encapsulate the cable jacket and the housing simultaneously. If the cable has a smooth jacket, the bond can be weak. That's why many overmolded cables use a textured or braided jacket for mechanical lock. Luna: And the cable exit angle matters too. A 90-degree exit creates a stress concentration on the overmold. Lucas: Right. A gradual 45-degree exit or a strain-relief boot integrated into the overmold works better. Our startup ended up with a 30-degree exit and a thickened lip at the cable entry. That passed a thousand-cycle flex test. Luna: So overmolding isn't just about waterproofing — it's also about mechanical longevity. Lucas: Exactly. And once you get the design right, it's a very robust solution. The startup I mentioned is now scaling to fifty thousand units per year, and their field failure rate from water ingress is under 0.1 percent. Luna: That's impressive. I'd love to see a comparison with potting — maybe we can cover that in a future episode. Lucas: That's a great idea. Potting is another encapsulation method, but it's messier and harder to automate. Overmolding is cleaner for high volume. Luna: Alright. Next time we'll talk potting compounds and thermal management. Lucas: Looking forward to it.