← Back to context

Comment by 1970-01-01

16 days ago

This is literally evidence of stuff being designed to fail. An extra diode costs less than a cent at production scale. This was a manufacturing choice, not an error.

My microwave mainboard failed because I changed the range light bulb without unplugging the whole microwave first, which I would not have thought necessary. It seems that, without unplugging the whole microwave, the act of changing a light bulb will cause catastrophic voltage to delicate parts. Turns out to be a common thing with this brand.

I ended up replacing the mainboard with a part from no apparent manufacturer with new features (the blue LEDs dim after inactivity so as not to illuminate the whole room at night) and no connection for the thermistor. Works like a charm. It feels very much like the original manufacturer wanted the board to fail and be replaced, while some random Chinese circuit-board maker sold me a better quality board.

nah, this is just not something designer would expect to fail like that. The LED has datasheet, the datasheet have leakage current, it has no data on increased leakage over years, you plan for what you have.

What would help is not randomly planning for some of the segments to fail (they are multiplexed with other things, you'd have to put more diodes), but to just get slightly better/less cheap LED display

Only "choice" made here was sorting by price when buying components for the cheap device.

  • Sounds like this is far more common a problem with blue LEDs than others, and that was certainly a choice.

    As if I needed another reason to detest the eye-searing blue LEDs that have infested every device.

It’s not exactly designed to fail, they just don’t care. If they could add a one-cent part that made it fail sooner, they wouldn’t do that either.

  • How is designing it to fail and not caring about part longevity not the same thing for the buyer?

    • Avoiding products where the manufacturer doesn't care about longevity means looking for products that are well designed. Avoiding products that are designed to fail is trickier, because even otherwise high quality products can be designed to fail. Random example, I had a nice Brother laser printer where the toner cartridges had a hard limit on how many pages they would print. This was done with a mechanical mechanism, so it was possible to disassemble it, reset it by turning a gear, and keep on printing. This took time and money to build just to make the product worse.

I don't think it's so much an issue of designed to fail as trying to get it as cheap as possible.

Theres further issues with everything coming out of china and a brand slapped on it. No one is left to take responsibility on the engineering front. This feedback I doubt will get to the correct people at best buy, let alone going back to the microwave manufacturer. And then there's the question of if they care, as they aren't a customer facing brand.

Eh, I don't agree.

LEDs are diodes (Light emitting diode). Certainly this was a cost saving measure, but it's not a bad assumption that the LED wouldn't allow reverse current flow.

Don't underestimate the appeal of saving one cent per unit. So long as the costs are externalised, anyway...

  • The correct calculation is what proportion of your profit margin will be saved by that one cent per unit. It gets more complicated if that control board was outsourced. If it costs 10 USD for the sub supplier to make and they sell it with a 10% markup then saving that one cent adds 1% to their profit.

Capitalist profit motive strikes again. The invisible hand expands tech and the visible hand keeps making tech worse.

People usually respond to this by saying that it would be absurd to suggest the company did this for its own benefit, when anyone who engineers knows these are often caused by revising design to minimize costs... and increase profits.