Near the end of last year, my digital wristwatch stopped working. The display no longer had any digits on it, like the battery ran out. Well, I’ve been using it for quite a while, over 10 years since I’ve last changed the battery, so maybe now is the time for a new battery change? Not quite.
But in any case, first I tried buying a new battery and replacing the battery. No dice. But, here’s the interesting thing. With the new battery in and playing around pushing the buttons to see if I need to enter some magic reset code to switch the watch on, I found out that pressing the LED backlight illumination button causes not only the display to light up, but the watch to also start ticking like normal. Hmm, interesting. Then, looking around and playing around with my watch a little more, I noticed that rotating and shaking it around causes the watch to intermittently work for a short period of time without backlight illumination, before freezing like it lost power. Interestingly, after pushing the backlight illumination button, the watch would continue from where it left off, like it preserved memory but did not advance the oscillator ticking.
Okay, okay, so thinking about this, it sounds like the problem is a loose solder joint. You rotate it around, and then the watch works because the loose joint makes contact. But then you rotate it again, and the loose joint looses contact, causing the watch to no longer work.
Okay, so now I had to get to work at a more complex disassembly. This was met with delays because I thought I had to remove the push buttons from the sides before I could completely remove the circuit board. I did remember that I was able to pop the buttons out once before, and surely they would be protruding inward to an extent where they would touch the circuit board but also prevent the circuit board from being removed. But, then I thought things through, and tried to be more optimistic. Well, maybe I can pull the circuit board straight out without removing the push buttons. So then I tried that, and it worked. Hooray!
Turning the circuit board over and looking at the front side, I saw the problem sure enough. Holy smokes! A capacitor is dangling by only one leg. Okay, okay, I definitely got to get this soldered back down. But… upon further inspection, the problem is more gruesome than I originally thought. The solder joints were intact, the leg actually became detached from the inside of the capacitor.
Okay, so now I’m going to need to be careful about this. Better not slap a soldering iron straight on the capacitor, else things can go awry really quickly. How do you repair a capacitor where one leg got pulled out of the capacitor itself? Surely, a web search would turn up results, right? Nope.
Failed search.
20200215/DuckDuckGo capacitor lead pulled out
20200215/DuckDuckGo capacitor lead disconnected
I end up looking at the general information about capacitor manufacture on the Wikipedia page. So, indeed it looks like the legs are not attached to the interior of a capacitor via soldering, but instead are attached via “bonding.” I assume this involves techniques using epoxy or resin to fasten the capacitor leg down to the interior of the capacitor. And, of course, the electrolyte inside the capacitor itself is sensitive to heat, please keep your soldering iron away from the body of the capacitor and only the legs, else you risk exploding your capacitor.
20200215/https://en.wikipedia.org/wiki/Capacitor
Okay, let’s go searching around for the capacitor part number in particular, I could read it off as “6971.”
20200215/DuckDuckGo 6971 capacitor
Okay, so I found this product info page, but it doesn’t look to be useful… not quite the part I am actually using. I thought for sure there is a metal body on the capacitor I’m looking at, but maybe it is just shiny gray plastic? The dimensions are somewhat similar, I guess.
20200215/https://www.csgparts.com/capacitors-fsc-5910/page/2705
20200215/https://www.csgparts.com/NSN/5910-01-090-6971_5910010906971.html
At the end of the day, I simply used a dab of Elmer’s glue to help hold the capacitor in a position where the legs would make proper contact with the interior of the capacitor. Unfortunately, the leg is still unbonded from the interior and merely in electrical contact. My hope is that this will buy me enough time to research a proper solution. It works, but I’m not sure how long the Elmer’s glue will hold. And, worst of all, this solution risks doing more damage in the meantime. Elmer’s Washable School Glue has a pH = 4.7. “NOTE: Not for bare metal” says the label.
But for the time being, it indeed works quite well. In order to truly be confident in the hardiness of this repair, I try deliberately dropping the watch 3 feet onto hardwood floor. QA PASSED - 3 ft drop test. Well, I guess I’ll keep the Elmer’s glue solution in place for quite some time.
In the midst of this, there’s a bigger lesson to be learned. Why did I have a broken capacitor leg connection to begin with? Well, I can answer that question with a bit of personal history. For much of the time that I have owned this digital wristwatch, I was under the impression that it is quite durable. It’s water-resistant up to 30 meters depth, and it feels quite solid in my hand. When I accidentally dropped it onto the ground the first few times, it sounded pretty solid when it hit the ground, and picking it up, it has always kept working for me. After the first few times, I’ve become accustomed to the assumption that “it’s durable,” so there’s no need to be extra cautious about dropping it, right? Wrong.
Presumably all those drops were slowly doing damage to it, it just wasn’t apparent until it stopped working. That capacitor? The way it is attached to the motherboard is by those two solder joints on the legs attached to one end of the “tin can.” The other end is left dangling. That allows for high frequency vibrations, such as those that might occur during a drop, to travel down the length of the capacitor. Depending on the specific conditions of the drop, those vibrations might be so intense that they snap connections apart. Matter of fact, even holding the bare circuit board in my hand and tapping on it made this quite obvious and evident, there was considerable vibration within the capacitor even from the slightest of taps.
Looking at other components on the circuit board, I noticed that some had hot glue to hold them down with extra security. The capacitor, in particular, had no additional hot glue to hold it in place. But, after I added my tiny dab of Elmer’s glue to it to help hold it in place, I did notice that tapping on the side of the circuit board seemed to pretty much have no effect on sending damaging vibrations through the capacitor. Indeed, that’s something that should have also been hot glued down during manufacture. Notably, there were two test points right next to the capacitor.
Alas, another thing I must admit is that the alarm speaker no longer works. It broke long ago. I blame the persistent dropping on the cause of its failure, of course. Looking inside, there is another component that looks out of whack, an inductor or transformer of a sort. Surely that is being used in the speaker path? I noticed that it did have a dab of hot glue on one side, but of course, the side opposite of the hot glue was the side that was unraveled.
So, the bigger lesson to be learned? Don’t be overly optimistic in assuming electronics are “durable” when they do not say so specifically. Chances are, even though they may be of a mostly durable design, they will have some simple design errors that effectively render them to be non-durable electronics.