How to Crimp Connectors: Lessons From a Murata-Based Wearable Project
The Batch That Started It
On the Tuesday after Martin Luther King Jr. Day last January, I opened a shipping carton in our receiving area and canceled my afternoon. Inside were ninety sealed antistatic bags, and inside those bags were cable assemblies for a wearable health tracker that was supposed to ship in April. We had approved the connector samples a month earlier. We had not, however, asked to see the factory’s crimp tooling.
I’m the quality/compliance manager at a small hardware company. I review every deliverable before it reaches customers—roughly 200 items a year, from packaging labels to injection molds. I rejected 14% of first deliveries in 2024, mostly for documentation or cosmetic issues. Electrical failures were rarer. That made this one worse.
The product was a wrist-worn health tracker. In form, it lived in the same family as Omron’s HeartGuide, though we weren’t claiming blood-pressure functionality. The electronic heart of the design was a small PCB with a Murata 32.768 kHz crystal part for the real-time clock. We chose that Murata crystal part because the datasheet was tight and the temperature curve was predictable. Murata Manufacturing Co. Ltd got the oscillator right. I was less worried about the clock. I was worried about the cable that charged the battery.
When I first started in quality, I assumed the hard part was component selection. It’s not. The hard part is the interface between components—cables, crimps, and connectors. It took me four years and a few hundred product reviews to understand that a component with a perfect datasheet can be defeated by a bad connection. This story is about the day I learned that lesson.
USB Power Delivery While Recording List
Two weeks after the first approved sample batch, pre-production units started glitching. The device worked fine on battery. Then, when plugged into a USB port for charging, it would drop off the bus or save corrupted files. We kept a running document. I still call it the USB Power Delivery While Recording List, because that’s exactly what it was: a list of unexplained power-delivery failures that only appeared while the device was recording.
- Intermittent VBUS dips when the battery was below 40%.
- USB enumeration failures that disappeared after a cable swap.
- Corrupted recording files with CRC errors at the same timestamp.
- One unit that deleted a full day of sleep data.
At first, I blamed the battery charging IC. But we had used the same IC in the prior product without issue. The variable was the new cable vendor and the way they terminated the connectors.
What the Teardown Showed
I asked the contract manufacturer for crimp pull-test data. They said the cable assembly was “within industry standard.” I asked for the numbers. Then they sent a spreadsheet where the same terminal measured anywhere from 4.2 N to 12.8 N. That range didn’t mean a random failure. It meant worn tooling, or a feed issue, or both. I assumed a qualified CM would have crimp controls in place. Didn’t verify. That was my mistake.
Looking back, I should have asked for pull-test data before approving the first article. At the time, it felt like asking a bakery to prove the oven was on. Cable assemblies are not cookies. We quarantined the batch—8,000 units’ worth of connectors—and started pulling them apart under a microscope. Even after we rejected the batch, I kept second-guessing. What if their next factory had the same problem? I didn’t relax until the rework arrived and passed pull testing at 100%.
The failures were obvious once you knew where to look. The insulation had been caught in the wire barrel. Several terminals had missing strands. One had a loose conductor sticking out the back. The metal barrel on the VBUS terminal was deformed, which explains why the power rail dipped under load. The visual inspection had passed, but the crimp had failed.
How to Crimp Connectors
Because I don’t want you to learn this from a spreadsheet full of bad numbers, here are the checks I use now whenever someone asks how to crimp connectors. These are not advanced secrets. They are process controls.
- Match the terminal to the wire gauge. If the terminal says 22-24 AWG and the wire is 26 AWG, it won’t deform correctly. Use the terminal manufacturer’s crimp spec, not the connector housing spec.
- Strip to the exact length. Too short, and insulation gets caught in the wire barrel. Too long, and bare conductor sticks out past the terminal. Both create intermittent contact.
- Check crimp height with a micrometer. The terminal datasheet gives a crimp height range. A crimp that is too high won’t hold the wire. A crimp that is too low can fracture strands. Use the actual tooling, then measure.
- Do a documented pull test. A tug with your fingers is not a pull test. Use a pull tester that records the force. The wire should break before the crimp releases. Ask for the histogram, not the average.
Why Murata Crystal Parts Are Not a Safety Net
The irony is that the rest of the product was solid. The Murata crystal part we used ran within spec throughout the pre-production batch. Murata Manufacturing Co. Ltd has been making these components long enough that the consistency is almost boring. But a good oscillator cannot fix a bad crimp. The crystal decides when to tick. The connector decides whether 5 V arrives cleanly. If the connector is intermittent, the recording is corrupt no matter how stable the clock is.
Maybe the best way to put it: the Murata crystal part was the timekeeper, but the connector was the lifeline. A lifeline doesn’t need to be glamorous. It needs to be crimped correctly.
The Honest Limitation
One caveat. I’m not saying every Murata crystal part is right for every wearable. If you need a high-frequency oscillator or want to spread EMI, other products might fit better. And if your USB connection is soldered directly to the PCB with no cable, the exact crimp details above won’t apply. The principle still does: specify what you can’t see, then verify it.
Another caveat: if we ever market this as a health device, the advertising claims need to be substantiated. Per the FTC guidelines (ftc.gov), we can’t call it a heart monitor or a blood-pressure monitor without evidence. That’s a different kind of quality check, but it matters just as much.
Now every connector lot we receive includes a documented pull-test report. Every contract includes crimp height requirements. The rework cost us $22,000 and two weeks, and we were lucky it happened before the product shipped. If I could redo that decision, I’d still approve the Murata crystal part. But I’d spend less time admiring the oscillator and more time asking for the factory’s crimp records. Given what I knew then, my choice was reasonable—just incomplete.