Lessons from a Failed Coating Project and Recovery

It was a Tuesday morning in late spring when our team at Precision Circuits gathered around the conference table, coffee mugs in hand, to review the details of what we thought would be our "trophy project." A medical device startup had approached us with a rush order: 500 PCBs for a portable patient monitor, requiring smt pcb assembly followed by conformal coating to protect against moisture and dust in hospital environments. The deadline was tight—just six weeks—but the client was willing to pay a premium, and the project would put us on the map as a go-to supplier for medical electronics. We were elated.

Our production manager, Maria, laid out the plan: "We'll start with smt assembly next week—all components sourced through our electronic component management software to ensure RoHS compliance. Then, conformal coating by the end of week four, testing week five, and delivery on week six. Piece of cake." The team nodded, confident. We'd done similar projects before, and our new automated coating machine had just been calibrated. What could go wrong?

Looking back, that overconfidence was our first mistake. But at the time, the excitement of landing a high-profile client overshadowed the tiny voice of caution in the back of my mind. We dived in, sourcing components, prepping the assembly line, and counting down the days until we could deliver.

Week three rolled around, and the smt pcb assembly phase was smooth sailing. Our electronic component management software had tracked every resistor, capacitor, and IC, flagging a minor delay in capacitor delivery that we resolved by swapping in an equivalent part from our reserve stock. By Friday, 500 fully assembled PCBs sat on the production floor, ready for conformal coating—the final step before testing.

Monday morning, the coating team fired up the machine. The process seemed routine: clean the PCBs with isopropyl alcohol, load them into the coating chamber, set the parameters (thickness: 50-75 microns, cure time: 30 minutes at 60°C), and let the machine do its work. By noon, the first batch of 50 boards emerged, glistening with a clear, protective layer. We high-fived, already mentally drafting the "project complete" email to the client.

Then came the inspection. Our quality control lead, Raj, picked up the first board and frowned. "Guys, come look at this." We crowded around, and my stomach dropped. Tiny bubbles dotted the coating, like a bad case of acne. In some spots, the coating had peeled away entirely, exposing the copper traces underneath. Raj tested a second board—same issue. A third? Worse. "This is a disaster," he said, voice tight. "These won't pass the client's waterproofing test. They'll fail in the field."

The room went silent. Maria's face paled. "How could this happen? The machine was calibrated last month!" We checked the coating material—expiration date was fine. The cleaning process—we'd followed the SOP. The temperature and humidity in the shop—within acceptable ranges. Panic set in as we realized: we had 500 defective boards, a client expecting delivery in three weeks, and no clue why our conformal coating was failing.

By Tuesday afternoon, we had to break the news to the client. I dialed their project manager, Sarah, and braced myself. "Sarah, we have a problem with the conformal coating. The first batch failed inspection—bubbles, peeling. We need to delay delivery by at least two weeks while we fix this." There was a long pause on the line.

"Two weeks?" Sarah's voice was sharp. "Our launch event is in four weeks. If we don't have these monitors by then, we lose our biggest investor. You have to fix this faster." I promised we'd work around the clock, but her tone made it clear: trust was fraying.,PCB,.

The financial toll hit next. We'd already spent $15,000 on components and labor for the failed boards. Reworking them would mean stripping off the defective coating (another $2,000 in chemicals and labor), re-cleaning, recoating, and retesting. And if we couldn't fix the issue, we'd have to start over from scratch—doubling our material costs and pushing the deadline even further. The team worked late that night, stripping coating off the first 50 boards by hand, but the process was slow and messy. By midnight, we'd only finished 10. "This isn't sustainable," Maria sighed, rubbing her tired eyes. "We need a better plan."

Wednesday morning, we brought in an outside expert: Lisa, a conformal coating specialist with 20 years in the industry. She arrived with a toolbox of magnifying glasses, pH strips, and a portable adhesion tester. "Let's start from the beginning," she said, calm as a storm. "Walk me through every step."

We recounted the process: cleaning with 99% isopropyl alcohol, air-drying for 10 minutes, coating, curing. Lisa nodded, then asked to see the cleaning station. She dipped a pH strip into the alcohol bottle and frowned. "This is 70% isopropyl, not 99%. Who bought this?" Our, Mia, winced. "The supplier was out of 99%, so I substituted 70%—the label said 'electronics-grade,' so I thought it was okay." Lisa shook her head. "70% has more water, which leaves a residue when it dries. That residue reacts with the coating, causing bubbles and poor adhesion. You cleaned the boards, but you left behind invisible moisture that's now ruining your coating."

It was a gut punch—such a small mistake, with such massive consequences. But Lisa didn't stop there. She checked our electronic component management software logs and noticed something else: the batch of PCBs had been assembled with a lead-free solder paste that required a higher cleaning temperature. "You air-dried at room temp, but this paste needs 40°C for 15 minutes to fully evaporate flux residues. The leftover flux mixed with the 70% alcohol residue—double trouble."

Armed with answers, we sprang into action. Mia sourced 99% isopropyl overnight. We adjusted the cleaning process: 40°C drying oven for 15 minutes, followed by a 5-minute cool-down. We tested a small batch of 10 boards with the new process—and this time, the coating was flawless. Raj did a adhesion test, scratching the coating with a blade; it didn't budge. Waterproofing test? Passed with flying colors. "We're back in business," he grinned, holding up a perfect board.

Over the next five days, we stripped, cleaned, and recoated all 500 boards. The team worked 12-hour shifts, fueled by pizza and the shared goal of making things right. By the end of the week, the final batch was inspected, tested, and packed. We delivered to the client with two days to spare. Sarah's email said it all: "The boards look perfect. Thank you for fixing this—we'll definitely work with you again."

To avoid repeating our mistakes, we documented every step of the recovery process. Below is a table of the troubleshooting steps we took, which now hangs in our production area as a reminder:

That failed coating project was a wake-up call, but it also taught us more than any successful project ever could. Here are the lessons we carry with us today:

A year later, we've refined our processes, and conformal coating projects are now our specialty. We even train other small manufacturers on best practices, sharing the hard-earned lessons from that chaotic spring. Our electronic component management software now includes a "coating compatibility" flag, and our production floor has a dedicated "pre-coating checklist" that every team member signs off on.

That failed project wasn't just a setback—it was a turning point. It taught us that success in manufacturing isn't about never making mistakes; it's about how quickly you learn from them, adapt, and come back stronger. And honestly? I wouldn't trade that stressful month for anything. Because now, when a client asks, "Can you handle a tough project?" we don't just say "yes"—we say "we've been tested, and we've come out better."

So here's to the projects that go wrong—the ones that make you sweat, argue, and stay up all night. They're the ones that turn good teams into great ones, and good manufacturers into trusted partners. And if your conformal coating ever starts bubbling? Check your isopropyl alcohol percentage first. Trust me on that.