Ask Hackaday: What’s Your Worst Repair Win?

Dec 25, 2023

Like many of you, I’ve become the designated "fix-it" person for my family and friends. While it can be a lot of work — I just finished an oil change that required me to lay in a cold, wet driveway and I can't mention in polite company the substances I was bathed in while fixing a clogged pipe last week — I generally relish my role. I enjoy solving problems, I love working with my hands and my head, and who doesn't like saving money and time?

But for me, the best part of being the fix-it guy is the satisfaction that comes from doing something others can't do. I find this especially true with automotive repairs, which conventional wisdom says is strictly the province of factory-trained experts. A little bit of a hero complex, perhaps? Absolutely! After all, I don't get paid for my repairs, so I’ve got to get a little something for the effort.

This is why a recent pair of unrelated fixes left me feeling thoroughly unsatisfied. Neither of these jobs was a clear win, at least in terms of getting the rush of being able to do something that nobody else could. At best, these were qualified wins, which both still left me feeling a little defeated. And that got me thinking that I’m probably not the only one who has had marginal repair wins like these.

The first repair was for a friend of mine who lives deep in the woods. Somewhat ironically, his driveway is a very busy place, mainly with wildlife like deer, elk, and the occasional moose. Keeping track of their comings and goings is important for safety; one does not want to surprise a moose, after all. A fair number of cars find their way up his driveway too: most of them are innocent, but occasionally they come with ill intent.

So wisely, he installed a wireless driveway alert system that gives him a heads-up on intruders. During the recent cold snap, though, his system stopped working. He changed the batteries and tried a few basic diagnostics, but no dice — the transmitter wouldn't work. So I offered to take a look and see if maybe I could save him a few bucks on a replacement.

Now, this transmitter is a somewhat ad hoc assembly. Inside the weatherproof case is what appears to be an off-the-shelf PIR motion sensor, which is wired to a PCB that contains a microcontroller and a radio module in an RF can. When the PIR sensor triggers, it sends power to the radio module, where the MCU sends a recorded sound clip — "Alert zone one, alert zone one…" — to the transmitter, broadcasting it to a receiver inside the house. Simple, but effective.

His report was that he couldn't even get the LED on the PIR sensor to light up, so I figured I’d start there and popped the cover off. This is where I started having thoughts of heroics — perhaps one of the SMD components on the sensor's PCB was bad, and I’d be able to trace the problem and do a little microsoldering. Or maybe I’d have to do some reverse engineering of the firmware to figure out what was wrong. The possibilities!

Sadly, it was not to be. After I had removed the sensor — which entailed disconnecting the battery pack wires from screw terminals on the PCB — I noticed that the positive lead had broken off in the terminal block. Surely this wasn't just a broken wire? Where are the heroics in that? But alas, when I stripped the wire back and put everything back together, the whole thing worked like a charm. I felt cheated — no need to bust out the oscilloscope, the waveform generator, the spectrum analyzer, or even the bench power supply. At least I got to use my microscope. Just to make myself feel like I’d done something, I crimped some ferrules on the ends of the battery pack wire and gave it back to my friend. Yay me.

As if that wasn't disappointing enough, a day or two later my daughter texted me to come over and look at her stove. Alarmingly, one of the electric burners on the stovetop had started causing electric shocks through their cookware. The shocks ranged from barely noticeable to a little on the spicy side. Not good!

I rushed over with a multimeter and started poking around. My thought was that the burner element was cracked or otherwise internally damaged, and a short between the nichrome wire and the outer covering had developed. I did some continuity checks between the element and chassis ground, but didn't see anything. Voltage checks between the burner and ground were a little different, though — I was seeing 117 volts on the problem burner. Well, there's your problem, lady!

Unfortunately, the local big-box stores were all sold out of 8″ burner elements for GE stoves, so I couldn't replace the dodgy element right away. We decided to swap the other 8″ element on the stove, which wasn't shocking her, into the spot where the bad element was. Surprise! That showed 117 volts too. So it's not the element, but the spot on the stovetop? Confused, we swapped everything back to the original locations and that seemed to fix the problem — no voltage from either burner to ground. What?

Clearly, this one isn't a fix. There's still something wrong with the stove, and I’ll need to do more diagnostics. It was kind of a fix, I suppose — at least my kid isn't getting shocked when she cooks. But it certainly wasn't a satisfying fix, and even if I replace the suspect burner with a new element, I’m not sure I’m going to trust the repair.

I think we can all agree that neither one of these repairs is very satisfying. In the case of the driveway alarm, I barely needed to be involved at all — my friend would probably have found out what the problem was with just a little tug on the wires. It was a fix, to be sure — it wasn't working at all when I got it, and it's working because of something I did. But it's a boring fix, at best. The spicy stovetop is unsatisfying, too, but in another way: it's not really a fix, because I didn't replace anything or find anything that appeared broken. It's just back together the way it was, and working normally, at least for now.

A fix is a fix, but some fixes are just not worth the effort. So the question is: what's your least satisfying repair story? Have you ever had high hopes for a glorious repair, only to end up with something a toddler could have fixed? Or like my daughter's stove, have you managed to make a problem "go away" without actually having done anything? What do you do in cases like that? How do you know when you reach the point of diminishing returns in terms of finding the problem? And when do you — gasp! — throw in the towel and call in an expert? Sound off in the comments below.