The 7-Year Bug That Took 3 Minutes to Fix

I once fixed a bug with a similar timeframe. Many attempts by many people over the years. One day I finally fixed it and thought I'd let the user who'd raised it know that it was finally resolved. Unfortunately he'd died of natural causes in the meantime. That really hit home how long it'd taken.

I fixed a similar long-scale bug in... my parent's kitchen extractor fan about a year ago. They had a fancy custom-designed stainless steel extractor fan installed in their new kitchen about 12 years ago. The fan worked but it had one nasty habit: sometimes the controls would freeze up when the halogen lights in the hood were switched off. To get the thing to work again the power needed to be disconnected by either pulling the plug - in the topmost kitchen cupboard behind a hardboard shield, i.e. not really feasible - or by turning off the power to the whole kitchen for a few seconds. This became a regular thing: start cooking, press the button on the hood to switch the lights off, notice that the display got scrambled again and that it no longer responds to button presses, walk to the fuse box, switch off the power to the kitchen, switch it on again, continue cooking while cursing the damned thing. The installer came by and replaced the control board to no effect, the thing still kept on freezing every now and then. I had my suspicions about the cause - halogen transformers can produce nasty spikes when turned off, surely they put some protection in place? - but my father did not want me to pull apart the thing because the 'experts' had already looked at it. OK, fine, it is your problem after all - I live about 1300 km to the north of where they live and came by some 4-5 times per year - so I won't.

In 2019 my father died. My mother was quickly diagnosed with Parkinsons disease and could not really cook for herself any more so I increased the frequency of my visits and cooked enough food in a week to last her a month, freezing everything in portions. This meant I got to spend more time with the extractor fan and I decided to pull it down to see if my hunch was correct. Well, it was, more or less. The installer had installed a ferrite bead but he clearly did not understand why and where that bead was to be installed so he put it on the incoming power line. That's fine if the interference is expected to come from the outside or if the locally produced interference is to be kept from travelling down the line to the network but it does nothing to protect the control circuit from the interference coming from that big transformer which is connected directly to it. Swapping the ferrite bead from the incoming power line to the transformer power line took maybe a few minutes. The thing never locked up again, unfortunately my father never got to see the thing working as intended.

I bet somewhere in the history of that product there is a note from a tester saying we need to test with a real world cable...

I’m curious why they didn’t deploy diagnostics in the field if they couldn’t replicate in the lab?

Every few months for 7yrs is a lot of opportunities to iterate on collecting field measurements. And it could be done in a holistic way that doesn’t break the safety certification.

Of course you need a filter capacitor. That's basic electrical engineering. This post makes you look incompetent.

I can list several major problems I created that took many months to solve that make me look incompetent.

So I'm the pot calling the kettle black.

Also: hindsight is 20/20.