Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Microinjection Molding for Tiny Parts
Transcript
- Lucas: So you're building a wearable device — something that sits on your wrist or maybe inside your ear — and you need a tiny plastic housing, say the size of a fingernail. Conventional injection molding can make that part, but the economics get weird when the part is that small. Luna: Weird how? I'd think smaller part means less material, so cheaper. Lucas: You'd think that, but the issue is the machine and the mold. A standard injection molding press has a screw — the thing that melts and pushes the plastic — that's typically 25 to 40 millimeters in diameter. When your part weighs less than a gram, that screw is moving a tiny fraction of its capacity. You lose control over shot size consistency, and you end up with flash or short shots. Luna: So you need a smaller machine. That's microinjection molding? Lucas: Exactly. Microinjection molding uses machines with screw diameters under 14 millimeters, sometimes as small as 8 millimeters. The whole barrel and injection unit is scaled down so you can meter out a shot of plastic that's literally a few milligrams. And the mold itself has to be precision-ground to tolerances in the single-digit microns. Luna: That sounds expensive. How does a startup even get access to that? Lucas: Great question. I talked to a startup last month — they make a custom hearing aid that fits deep in the ear canal. Their battery housing needed to be 0.15 millimeters thick, half of what they could achieve with standard molding. They went to a microinjection job shop in Minnesota that specializes in medical devices. The mold alone cost them $45,000, but each part ended up costing about 12 cents. Luna: Compare that to what they were paying before? Lucas: With standard molding, they were using a three-plate mold and getting inconsistent wall thickness. Reject rates were around 15 percent. After switching to microinjection, reject rates dropped below 1 percent, and they shaved 22 percent off their material cost because they could use less plastic per part. Luna: So the mold cost is the barrier, but once you have it, the per-part cost is really low. That sounds like a classic hardware startup trade-off. Lucas: Exactly. And it's not just hearing aids. Microinjection is huge for medical devices — things like endoscopic camera housings, insulin pump components, even the tiny gears inside some robotics actuators. If you've ever held a really small, intricate plastic part and wondered how it was made, microinjection is often the answer. Luna: I'm thinking about 3D printing though. Couldn't you just print those parts in resin? The resolution on high-end printers is getting insane. Lucas: You can, and for prototyping it's perfect. But for production, the economics flip. A microinjection cycle time is maybe 5 to 10 seconds per part. A 3D printer might take 30 minutes to print one part — or if you're lucky, a dozen parts at once in a batch. For volumes above a few thousand units, injection wins every time. Luna: And material properties? I've heard printed parts can be anisotropic — weaker in some directions. Lucas: That's true. With microinjection, because the plastic is melted and forced into the mold under high pressure, you get isotropic properties. Plus you can use engineering-grade materials like liquid crystal polymer or PEEK — stuff that 3D printers struggle with or that requires expensive post-processing. Luna: So if you're a startup at the prototype stage, do you design for microinjection from day one, or do you adapt later? Lucas: Ideally, you think about it early. Microinjection molds are expensive, but the design rules are different. You need uniform wall thickness — even more so than in standard molding — because the plastic cools so fast in a tiny cavity. You also need generous draft angles and careful gate placement. I've seen startups try to retrofit a 3D-printed design into a microinjection mold and they have to add ribs or change the geometry completely. Luna: What about the mold itself? Is it made differently from a standard mold? Lucas: Yes. Microinjection molds are typically machined on ultra-precision mills, not standard CNC. Some moldmakers use EDM — electrical discharge machining — to create features that are hundredths of a millimeter. And the mold temperature control is critical. You need heating and cooling channels that can hold the mold temperature within one degree Celsius, because the plastic's flow behavior changes dramatically at small scales. Luna: That's wild. So the mold itself is almost like a miniature piece of equipment. Lucas: It really is. And there's a whole ecosystem of micro-molders around the world — mostly in Germany, Japan, and the United States. A startup I know found a shop in Switzerland that specializes in micro parts for watch movements. They ended up using that shop for a robotic joint component. Luna: That's a cool cross-industry application. Watches and robots use similar tiny gears. Lucas: Exactly. And the tolerances are actually tighter for watch parts. So that shop was overqualified, but the startup got parts that were perfectly consistent. They paid a premium — maybe 20 percent more per part — but they never had a reject. Luna: Speaking of paying a premium — I want to circle back to something. These shows take a lot of research and conversations like the one you had with that hearing aid startup. If listeners find this useful, there's a small way to support the show and keep it ad-free. Lucas: Yeah, it's a buy me a coffee dot com slash fexingo link. If today gave you one concrete idea you can use, that's the place. No pressure, just a way to say thanks. Luna: And we appreciate it. Okay, back to microinjection — one more thing I'm curious about: what about multi-cavity molds? Can you make a bunch of tiny parts at once? Lucas: You can, but it's tricky. Multi-cavity microinjection molds require perfectly balanced flow. If one cavity fills faster than another, you get different part weights. Some advanced molds use valve gates or hot runners with individual temperature control. But for low volumes, single- or two-cavity molds are more common. Luna: So for a startup doing, say, 10,000 units a year, a single-cavity mold running 10-second cycles would be fine. That's about 3 shifts per week to hit volume. Lucas: Exactly. And if demand grows, you can either add more cavities or run multiple shifts. The mold itself is the asset. And because microinjection molds are so precise, they can last for millions of cycles if maintained properly. Luna: That durability is a big plus for a startup that might pivot or scale slowly. You don't want to rebuild a mold every year. Lucas: Right. One more thing that's worth mentioning: there's also two-shot microinjection, where you mold one material over another — say a soft-touch overmold on a rigid core. That's used in some wearable health monitors and hearing aids. But that adds complexity and cost. Luna: That's probably a whole other episode. For now, I think we've given listeners a solid handle on what microinjection is and when to consider it. Lucas: Yeah, I think so. If you're building something that fits in your palm and needs precision plastics, look into it. It might save you money and headaches down the line. Luna: Thanks, Lucas. And thanks to our listeners. Lucas: Thanks, Luna.