Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Lead-Free Solder Alloys
Transcript
- Lucas: If you are building a hardware prototype and you send your pick and place file to a contract manufacturer, there is a good chance they will ask you one question you are not prepared for: 'Which solder alloy do you want?' Luna: And most founders probably say 'uh, the normal one?' — which is how you end up with SAC305 on your BOM without ever thinking about it. Lucas: Exactly. SAC305 stands for tin, silver, copper — ninety-six-point-five percent tin, three percent silver, half a percent copper. It is the default lead-free alloy for the industry. It melts around two hundred and seventeen degrees Celsius, it is reliable, and every contract manufacturer has it in stock. Luna: So why would you ever choose something else? Lucas: Because that two-hundred-and-seventeen-degree reflow temperature causes real problems for certain products. I have been tracking a smart-sensor startup — they make a thin board with a large LGA package on one side and a bunch of connectors on the other. With SAC305, the board warped during reflow because the temperature gradient across that thin laminate was just too high. They were seeing ten to fifteen percent of boards fail visual inspection. Luna: So they tried a lower-temp alloy? Lucas: They switched to a bismuth-tin alloy — specifically, a paste that is forty-two percent tin, fifty-eight percent bismuth. Melts at one hundred and thirty-eight degrees. That is almost eighty degrees cooler. The board stopped warping. Their rework rate dropped by about twenty percent. And they passed a drop test they had been failing. Luna: Wait — the drop test was failing because of the alloy? I would have assumed it was a mechanical design issue. Lucas: So would they. But the root cause was micro-cracks forming in the SAC305 joints during reflow because the board was flexing as it cooled. The bismuth alloy is actually more ductile, so it absorbs that stress better. Plus, the lower peak temperature means the whole assembly sees less thermal shock. Luna: But bismuth has a reputation for being brittle, right? I have heard people say it is not reliable under thermal cycling. Lucas: That reputation comes from older formulations that had a problem with something called 'voiding' — gas pockets get trapped in the joint during reflow. But the newer bismuth-tin pastes have much better flux chemistry. The voiding rate is comparable to SAC305 now. The real trade-off is thermal cycling range. If your product lives in a car dashboard that goes from minus forty to plus eighty-five Celsius every day, bismuth might not hold up as well over ten years. But for a consumer device that stays in a climate-controlled room? It is fine. Luna: So the alloy choice is really about your use case, not a universal right answer. Lucas: Right. And there is another one to consider: SAC105. That is ninety-eight-point-five percent tin, one percent silver, half a percent copper. Less silver, so it is cheaper — about twenty percent less than SAC305. But it has lower fatigue resistance. If you are building something with a short product lifecycle — say, a smart home gadget you expect to replace in three years — SAC105 might save you money without hurting reliability. Luna: But your contract manufacturer probably won't suggest it because they don't stock it as often. Lucas: That is the hidden obstacle. Most contract manufacturers buy solder paste in bulk from one supplier. If you ask for an exotic alloy, they have to order a special syringe or cartridge. That can add lead time and minimum order quantities. One founder I talked to said her manufacturer wanted a five-hundred-dollar minimum for a custom paste — for a prototype run of fifty boards. Luna: Five hundred dollars for paste alone feels painful when you are bootstrapping. Lucas: It is. But here is the thing: if your assembly yield goes from eighty-five percent to ninety-five percent because you switched alloys, that five hundred dollars pays for itself in one production run. You have to do the math. Luna: And then there is the regulatory angle. If you are shipping into Europe or California, you are already RoHS compliant by using lead-free. But some people don't realize that California's Prop 65 actually limits the amount of certain metals in consumer products. Silver is not restricted, but if you use an alloy with cadmium or lead — even trace amounts — you need to label it. Lucas: That is a good point. And there is a growing movement in the EU to restrict bismuth in certain applications because of mining concerns. So the regulatory landscape is not static. I think the takeaway for a hardware startup is: do not default to SAC305 just because everyone else does. Ask your assembly partner what they have experience with. Get a quote for two or three alloys. And run a small thermal cycling test before you commit to a full production run. Luna: You mentioned tin whiskers earlier — is that still a real worry with modern alloys? Lucas: It is, especially with pure tin or high-tin alloys. Tin whiskers are these tiny conductive filaments that can grow out of a solder joint over time and cause shorts. SAC305 and SAC105 both have tin content above ninety-five percent, so they can grow whiskers under the right conditions — high humidity, mechanical stress, thermal cycling. The industry solution is to add a small amount of nickel or antimony to suppress whisker growth. Some manufacturers do that. But if you are buying generic paste from a low-cost supplier, you might not get that additive. Luna: So you kind of have to trust your paste supplier's formulation. Lucas: Exactly. That is why I recommend startups ask for the datasheet on the paste. Look for the term 'whisker mitigation' or check the IPC j std-006 classification. A good supplier will tell you their alloy has been tested to meet Class 2 or Class 3 reliability. Luna: Let me ask you about a specific scenario: say I am building a wearable device that goes through the wash — waterproof, ruggedized. Which alloy would you start with? Lucas: I would start with SAC305 and then evaluate the drop test and the flex test. If you see failures at the BGA or LGA joints, then I would look at a bismuth-based low-temp alloy because it reduces the thermal stress on the package itself. But I would also consider conformal coating to protect against moisture ingress. The alloy alone won't solve all your reliability problems. Luna: And if cost is the primary driver? Say I am making a disposable medical device. Lucas: Then I would use SAC105. It is cheaper, it reflows at about the same temperature as SAC305, and for a short-life product, the lower fatigue resistance is irrelevant. Just make sure you are not violating any medical device regulations — some disposables still require full lead-free compliance with specific alloy certifications. Luna: If today was actually useful to you, the way these stay ad-free is listener support — buy me a coffee dot com slash fexingo. Lucas: Yeah, and it is a small thing that keeps us from having to run ads. So if you got something out of this, that link is there. Luna: Alright, back to alloys: what about the new 'innolot' type alloys that are supposed to be for high-reliability automotive? Lucas: Innolot is a trade name for a tin silver copper alloy with added bismuth, antimony, and nickel. It is designed for high-temperature and high-vibration environments. Melts around two hundred and twelve degrees — slightly lower than SAC305 — but it has much better thermal fatigue resistance. The catch is cost: it is about thirty percent more expensive than SAC305. And not every contract manufacturer can run it because it requires a specific reflow profile. Luna: So for a startup making an electric scooter controller that sees a lot of vibration, is innolot worth the premium? Lucas: I would say yes, if you are planning to volume-produce. The cost of a field failure — recalls, reputation damage, customer support — is far higher than the thirty percent premium on solder paste. But for a prototype run, you can probably get away with SAC305 and just pot the board with epoxy to dampen vibration. Luna: Potting adds weight and makes rework impossible. Lucas: Correct. So that trade-off is exactly why startups need to think about alloy before they get to the potting stage. If you design for innolot from the start, you can avoid the need for potting altogether in some cases. Luna: One more thing: I have heard about 'solder preforms' — little solid pellets of alloy that you place on a pad before reflow. When would a startup use those instead of paste? Lucas: Preforms are great for applications where you need a very precise volume of solder — like attaching a large thermal pad on a power component. Paste can slump or spread unevenly, but a preform gives you a consistent joint height. They are also used in step-soldering, where you have two different reflow temperatures in one assembly. You use a higher-temp preform for the first side, then a lower-temp paste for the second side so the first joints don't re-melt. Luna: That is clever — but it adds a manual placement step, which increases labor cost. Lucas: It does. So it is usually reserved for tricky assemblies where paste alone won't cut it. I have seen startups use preforms for LED arrays where consistent standoff height is critical for optical alignment. Luna: Alright, so if I am a founder listening right now, what is the one thing I should do differently after this episode? Lucas: Pull up your BOM and look at the 'solder paste' line item. If it just says 'lead-free', ask your CM for the exact alloy designation. Then spend an hour reading the datasheet. That one datasheet can tell you more about your product's reliability than a dozen blog posts. Luna: Good advice. Next episode we are going to talk about something every hardware startup hits: how to choose between a rigid and a flex PCB for your first prototype.