I tend to think of ESD as the silent killer. It is all around, and if you don't take the proper precautions, it can destroy your design or at least hinder it enough to cause failure later. ESD issues cost companies millions of dollars each year. In spite of that, I think that many engineers (design, manufacturing, and quality) sometimes underestimate or just don't understand the need for proper precautions. As a result, proper action is put off.
It reminds me of a minor issue I had a few years back. I had a simple analog circuit that I used in many applications I had designed. It contained a couple of quad op-amps, a transmitter and receiver, and a bunch of caps and resistors. It was a very simple circuit, and not a lot could go wrong. The design seemed pretty solid and was very similar to a version that had been used for years.
One day, I had some calls from the quality department that a few systems were returned where the circuit was not functioning. Now, this board was used by the thousands, so having two or three failures didn't raise any red flags — until the next week, when we had a couple of processor board failures. Again, this board was used by the thousands, and we had only a couple of returns from quality. Even though the issue seemed small, I was still curious about what was going on, so I started investigating. Experience shows that issues like these can become huge and spread like wildfire overnight.
In my investigation, I put these boards on the bench and checked the power supply and signal voltage levels. Everything seemed OK — except that, on the small analog sensor board (the one with the op-amps), the op-amps were not functioning.
Moving over to the microprocessor board, I found that one of the transistors was not working correctly. On another board, the processor just seemed to be acting silly. In the back of my mind, I was thinking about tracking these components back through the contract manufacturer (CM) to where it had purchased parts. Then, maybe I'd check date codes, lot codes, production codes, etc., but I decided to keep the investigation local for the time being.
I discovered that both of the failed circuits were assembled on the same production line at our CM. Again, these boards were assembled on a couple of other lines, but the failed boards came from a common assembly line. I went to the area on the line where they were assembled. As it happened, the boards were assembled by the same operator in the same area. By this time, my mind was racing 100 mph. These totally different circuits failed within one week's time on the exact same production line, in the exact same area, by the exact same operator.
Now the investigation got a little deeper. I watched the operator assemble these boards into the enclosures. Everything seemed OK. The operator did not seem to be doing anything out of the ordinary, but then I noticed something that made my brain explode. Could this be the issue?
The operator was standing on an ordinary piece of plush carpet. When I asked why, the operator said it was comfortable and kept his feet from hurting. I quickly ran to the areas on the other production lines where these boards were being assembled. I saw no carpet at those lines. I went back to the suspect production line and asked the operator how long he had been using the carpet. He said he had been using it just a couple of weeks. Uh-oh, it all seemed to hit home.
Let's recap. These circuits had no history of issues. Similar designs had been used in the past. Both of the circuits were assembled on multiple lines. The failures came from the exact same line, the exact same area, and the exact same operator. I had tested the bad boards, and even though they were not working properly (or at all), I saw no obvious issues or damage. What was apparently the only issue in this equation? It was the small piece of plush carpet. How could that piece of carpet be the problem? The only answer I had was ESD. For one thing, my company had invested in little ESD mitigation equipment. In fact, it took pretty much no measures against it.
My conclusion was the circuits were getting damaged by an electrostatic charge building up as the operator scuffed across the carpet. At some point, when the operator was handling the boards, they got zapped with ESD. It wasn't long afterward that we implemented all the proper ESD floor and bench mats, wrist straps, and ground wires. And we trained everyone in proper PCB handling.
Perhaps I could have beefed up my circuit, but if that op-amp or processor were being touched directly and zapped, there was almost nothing I could have done to protect the design. I had many useful discussions with my management and the quality engineers about why spending the few thousand dollars to implement these ESD precautions could very well save us a lot of money in the future. Returns would be reduced. Design engineering time to try and figure out why good circuits failed would be minimized.
I wonder how many other engineers out there have had similar experiences with their designs or manufacturing facilities? Is ESD often underestimated? What's been your experience?
- ESD Protection for Analog Circuits
- Prevent Your Next High-Flying Product Launch From Being Grounded
- Meeting the ‘Rugged Design’ Challenge
- Can Technology Keep Up With Us or Are We Just Spoiled?
- Component Selection: How Do You Do It?
- Analog Designers: Today vs. Yesterday
- Is There Anything Left to Invent or Discover?
- When Engineers’ Worlds Collide
- What Changes More, a UL Engineer’s Requirements or the Weather in the Midwest?
- Are Today’s Design Tools Taking Away Our Critical Thinking Ability?