Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Solder Paste Inspection Systems
Transcript
- Lucas: So you've got your PCB fully populated, components placed, boards headed into the reflow oven. Feels like the hard part's over. But if the solder paste wasn't right from the start, you're about to discover that in the worst possible way — after reflow, when rework is expensive and slow. Luna: And that's exactly where solder paste inspection — SPI — comes in. It's the gatekeeper before reflow, checking that every pad has the right amount of paste in the right shape. Lucas: Right. SPI systems use structured light or laser triangulation to measure each paste deposit in three dimensions. They check volume, area, height, and alignment. If a pad is short on paste, or if there's a bridge between two pads, the system flags it before the board ever hits the oven. Luna: And catching it there, before reflow, means you can wash off the paste and reprint. That's minutes of rework instead of hours of desoldering and replacing components. Lucas: Exactly. I talked to a hardware startup last quarter that makes IoT sensor modules. They were seeing about eight percent first-pass yield loss on their main board — mostly from tombstoning and insufficient solder joints. Turned out the paste deposits on small passive components were consistently twenty percent below the target volume. Luna: And they only found that after reflow, during AOI. So they were scrapping boards or spending huge time on rework. Lucas: They brought in an inline SPI system from Koh Young — the leading supplier in this space. Within two weeks they'd tuned their stencil printer parameters based on SPI feedback. Their defect rate dropped from eight percent to under five percent. That's a forty percent reduction in defects, just from measuring paste before it's too late. Luna: And that's not even the full story. SPI data can feed back to the printer in a closed loop — automatically adjusting squeegee pressure, print speed, or stencil wipe frequency. Lucas: That closed-loop capability is huge for startups scaling from prototype batches to low-volume production. You can't have an operator staring at every board. The SPI system becomes the quality gate that lets you trust the process. Luna: And it's not just volume. SPI measures paste height and area too. A tall, narrow deposit might have the right volume but poor geometry, leading to solder balling. Lucas: Right. The key metrics are paste volume as a percentage of pad volume, area coverage, height, and shape. Most systems use a pass-fail threshold — typically plus or minus twenty-five percent on volume, but tighter for fine-pitch components. And they also check for bridges, which are deposits that connect two pads. Luna: Bridges are the ones that really hurt. A tiny solder bridge on a 0.4-millimeter-pitch QFN can cause a short that's nearly invisible to the naked eye. Lucas: And SPI catches those before reflow, when the paste is still wet. You can just reprint. After reflow, you're dealing with a solid solder bridge that might require a hot air rework station and a lot of luck. Luna: So for a startup that's trying to get to reliable production, SPI is one of those tools that pays for itself in reduced scrap and rework labor. Lucas: And if you're sending boards to a contract manufacturer, you want to make sure they have SPI capability. Many CMs will offer it as an option, and it's worth the small per-board upcharge. Luna: I always tell founders: if your CM doesn't have SPI, ask why. It's a standard tool in any serious SMT line. Lucas: Look, if this kind of practical manufacturing detail is useful for what you're building or running, that's exactly what we're here for. And if today's episode was worth a coffee to you, honestly, that's the link — buy me a coffee dot com slash fexingo. Keeps the show ad-free and lets us dig into the specifics that matter. Luna: Yeah, it really does help. And we appreciate every bit of support. Lucas: So back to SPI. One thing that surprises founders is that SPI isn't just about catching bad prints. It's about understanding your process capability index, or Cpk, for solder paste deposition. Luna: Cpk tells you how well your process stays within spec over many boards. A Cpk of 1.33 or higher is the target for a stable process. Lucas: And SPI systems automatically calculate that for you. They'll show you a histogram of paste volumes across thousands of pads. You can see if your process is drifting — maybe the squeegee is wearing, or the stencil is getting clogged — before you produce a batch of defective boards. Luna: So it's preventive, not just detective. Lucas: Exactly. And the data is useful for more than just the line. You can share SPI reports with your stencil vendor to improve future stencils, or use it to validate a new solder paste formulation. Luna: What about cost? A standalone SPI system from Koh Young or Omron can run anywhere from thirty to sixty thousand dollars. That's a big investment for a startup. Luna: But if you're doing even moderate volumes, the savings in scrap and rework can justify it in under a year. Plus, some CMs offer SPI as a service, where you pay per board instead of buying the machine. Lucas: That's a smart entry point. You get the data without the capital expenditure. And you can use that data to negotiate better pricing with your CM, because you're giving them a cleaner process. Luna: One thing I've seen is startups that skip SPI entirely and rely only on AOI after reflow. But AOI can miss a lot of paste-related defects because the solder has already flowed and looks different. Lucas: That's the key insight. SPI and AOI are complementary, not substitutes. SPI catches paste issues, AOI catches component placement and solder joint issues. You really need both for a robust quality system. Luna: And if you're doing prototypes by hand, you probably don't need SPI. But as soon as you go to stencil printing and reflow, it's worth considering. Lucas: Absolutely. The rule of thumb I've heard: once you're doing more than a thousand boards a year with fine-pitch components, SPI starts to make financial sense. Below that, manual inspection might be enough. Luna: But even at low volumes, if you're doing a complex board with lots of small passives, the rework cost from a bad paste print can be huge. Lucas: Right. I think the takeaway is: SPI is the unsung hero of SMT quality. It's not glamorous, but it's one of the highest roi investments a hardware startup can make in their manufacturing process. Luna: And it gives you data to build a repeatable, scalable process — which is exactly what you need when you're moving from prototype to production. Lucas: Next episode, we'll talk about how to use that SPI data to optimize your reflow profile. Because the paste inspection is only the first step — the real magic is connecting it to downstream processes. Luna: Looking forward to it.