[tzs]

This has long been a temptation for engineers. 40 years ago the textbook in the digital electronics class I took at Caltech had a chapter called "The Engineer as Dope Pusher" that talked about it.

It gave as an example clothes dryers. The way most home clothes dryers working back then was you put the clothes in, you turn a dial on a timer to the number of minutes you want the dryer to run, and you press start.

The mechanical timers were very reliable. There hadn't been any substantial improvement in their design in decades because there really wasn't anything to improve. There had been improvement in the materials used, and in the cost, but fundamentally mechanical timers was a solved problem.

If the mechanical timer ever broke the repairperson would have replacements in their van. Even if they didn't have the specific one for your dryer it didn't matter because they all worked pretty much the same. They could just put in another one. Maybe the mounting holes wouldn't be in the right place, but they could easily improvise some way to mount it in your dryer.

The book went on to say that somewhere there is an engineer designing a new clothes dryer, and instead of a mechanical timer that engineer is putting in a digital timer. It has a microprocessor, 7 segment LED digit displays for the time, some buttons for interacting with it (such as setting the time and correcting mistakes), and a power supply. And let's not forget that it has software.

That digital timer has no advantage to the user over a mechanical timer. But it has disadvantages. The interface will be worse. It will cost more. It won't be more reliable and possibly will be less reliable, and if it does need repair the repairperson probable won't have the parts. If they have another brand's digital timer on hand they probably won't be able to adapt it to your dryer.

So why is that engineer designing the new dryer with a digital timer?

Because mechanical timers are boring. Digital electronics was at the cutting edge of consumer engineering then, and so by using a digital timer the engineer got to play with exciting new technology.

[Kirby64]

Mechanical timers are fine, but digital circuitry is by far going to be more reliable if designed properly. There isn't anything physically moving, so the failure modes are much more restrictive. Also, digital provides advantages with offering variable timing on a dryer, for instance, based on the input of a moisture sensor. Mechanical methods for that are more complicated.

Additionally, I would be very surprised if the digital solution is not cheaper to make. Maybe not when first originally introduced, but nowadays it very likely is.

You're right that repair-ability is hurt in some ways... but the industry has moved to compensate. You can buy boards and replace them. They aren't inherently hard to service, because the form factor doesn't really have limitations.

[joelfried]

There is something physically moving: the machine itself. You can't wash or dry clothing without moving it around. Given that the entire machine moves (and on spin cycles, reasonably quickly), you need to make sure your circuitry is capable of handling the strain.

As a homeowner, I wish someone (anyone!) still sold reliable analog appliances that just did their job simply and made repair parts and schematics reasonably available.

[Kirby64]

Nobody would attach the timer to the actual moving drum, so the worst it has to deal with is physical vibrations from use (which, admittedly, can be quite violent if you have an unbalanced load). There are very few digital circuits that are actually meaningfully sensitive to vibrations. At worst, it's a manufacturing problem to make sure the PCB/solder joints don't crack from vibration.

In comparison, the mechanical timer is physically moving. A clockspring, or some sort of mechanism that physically sets the time remaining. Depending on how it's built, vibrations are a harder problem to solve. Not impossible, obviously, but it certainly adds cost.

Also, for most appliances we deal with today... they usually ARE simple to work on. Simple switches and mechanical contrivances. Parts are typically readily available... even PCBs, although possibly not at great pricing. There's certain appliances where you are basically screwed (fridges come to mind...), but that is mainly in my view because the typical failing part is the compressor. Nobody is rebuilding a compressor themselves.

[willyt]

In a fridge the typical failing part is the plastic shelves in the door. The compressor almost never fails and discarded fridges are a great source of cheap pumps if you need to scavenge them.

Source: The episode of 'The Secret Life of Machines' on fridges. Search it on YouTube.

[Kirby64]

Maybe for certain models. Personally I've never had a shelf break on any fridge I've used, which makes it sound like that's a user error to me.

On the compressors, there was a vast swath of Samsung and LG fridges that had known defects on compressors causing them to fail. Right now, the ice machines are probably most problematic. If you own a Samsung fridge with an ice maker you know what I mean.

[EdwardDiego]

On the Samsung fridge aspect...

A few years ago I was renting a house that came with a Samsung fridge that provided chilled water / ice. My kids loved the chilled water.

However, our usage of it caused the paint to start bubbling below the dispenser, and the owners of the rental wanted me to replace the door at a cost of ~$800 USD(!).

I argued that we were using the fridge as designed, so we weren't liable, instead they should discuss what looked to me like an obvious design flaw with Samsung.

They disagreed, so we ended up in court. My defence was about 12 pages printed from an appliance review site of people specifically complaining about this paint bubbling.

Easiest win ever, but seriously, how do you put a device that works with water into a fridge and fail to ensure it can't leak under the paintwork?

[adolph]

I do minor appliance repairs on occasion and the current time of manuals, videos and parts availablity feels like a golden age.

Granted, none of my large appliances are younger than 10 years, but I think I could build new ones (expensively) for the all the parts and schematics available, even wiring diagrams.

[fennecfoxy]

Yup exactly. And the electronics in a clothes washer allow for a lot more functionality than mech. When they say the "timer" was replaced with electronics, what they really mean to say is "the timer was replaced by electronics, which also allow you to select different temperatures/runtimes and whatnot based on the type of fabric you're washing or how fast you want it to be done, if you want stain removal or extra rinse then you can enable that, amongst other new options".

"Repairability" is becoming slightly nonsense because even as someone who is a programmer, who has done electronics at a hobbyist level myself, I'm not going to be able to fix a lot of stuff purely because you have to become an expert on it, the time investment is too high. As systems get more complex (to the overall benefit of all of us) the value of repairing something yourself vs getting an expert to do it, changes.

I think right to repair is good though, but purely meaning that companies to not intentionally attempt to thwart the repair of their devices and that parts/manuals are available where needed. Even so, this doesn't mean that every phone repair place is going to debug some sub-circuit inside some small part of the newest iPhone - they'll just identify the overall broken module and replace the entire board/module.

[gonehome]

Yeah, I’m always skeptical of “it was better in the old days” type arguments (even though I recognize the aesthetic appeal of analog).

People make similar claims about cars, but old cars broke down all the time and new ones are basically appliances that “just work” without the driver needing to know anything. Similar for computers to smart phones (though obviously both digital in that transition)

[tzs]

It wasn’t the old days. At the time the book was written and at the time I took the class mechanical timers in dryers were ubiquitous. Digital was new, expensive, and didn’t offer any advantages for that application.

Eventually digital became cheap, and enabled new features like dryers that had various sensors that could be used to optimize drying, but that was several years down the road.

[midoridensha]

Back in the "good old days", a car was ready for the junk heap after 50k miles. These days, that's barely broken-in. "But you could fix it yourself!" Who cares when the thing has such a short lifespan? It's really strange hearing people pining for the days of shitty old cars that you needed to constantly adjust the carb, set the points, etc. Insane.

[aSockPuppeteer]

That’s the point of designing it improperly so it fails early and only keeping spare parts available for a year or consistently out of stock.

The supply chain issues are hurting the servicing part, repair feasibility, and manufacturing part.

[conradev]

> If the mechanical timer ever broke the repairperson would have replacements in their van

> That digital timer has no advantage to the user over a mechanical timer. But it has disadvantages.

The mechanical timer is known to degrade over time, which is why the repair person has spares in their van. Does the digital timer really have no advantages? Will it ever fail and need to be replaced? How much more does it cost?

Yes, engineers are tempted to use shiny tools all of the time. Evaluating whether or not the tool is right for the job is _hard_. But it feels wrong to say that novelty is the only motivation behind upgrading tools?

The digital timer was made by an engineer, too, who designed it to be a more reliable replacement for faulty mechanical timers. It has both advantages and disadvantages compared to mechanical timers, which is why the engineer made it in the first place?

[JadeNB]

> The mechanical timer is known to degrade over time, which is why the repair person has spares in their van. Does the digital timer really have no advantages? Will it ever fail and need to be replaced? How much more does it cost?

Surely the expected lifetime of a digital timer is shorter than that of a mechanical timer.

[error503]

Surely is a very strong statement here. I see no reason that a properly designed electronic timer wouldn't have effectively infinite lifetime, which is not possible with a practical mechanical timer. It has no moving parts (other than switches, which can be substituted for capacitive touch). A mechanical timer has many small mechanical parts and wear points, and can get gummed up over time if it doesn't outright stop functioning.

In practice, cost engineering is going to mean neither is completely reliable, but it should be cheaper to make an electronic timer reliable enough. Especially today, where the cost of a functioning mechanical timer is probably an order of magnitude more than an equivalent electronic timer.

[JadeNB]

I'm willing to believe that the best electronic switches can last longer than the best mechanical switches, but there's so many more ways for an electronic switch to fail that it's a lot easier for me to trust an off-the-shelf mechanical switch than an off-the-shelf electronic one, especially if the cost of failure of the mechanical switch is just an easy swap in of another one.

[error503]

In what ways can a purely electronic, no moving parts, switch fail?

There are so many different mechanical things that can break, jam, get gummed up...

[shard]

Capacitors can die, for example. Anything with a battery backup, the battery can leak and damage components. Electronics are more prone to ESD and water damage than mechanical parts.

Going into the realm of unlikely scenarios, electronics are more susceptible to EMPs.

[heisenbit]

The thing is: It is moving as there are vibrations. There is a fair amount of acceleration and a high frequency. Then there are temperature cycles as the machinery is not perfectly isolated. There is migration of atoms at contact boundaries. Plenty things move.

[nuancebydefault]

What do you mean by "reliable" wrt mechanical timers? I'm pretty sure the clothing dryer is a harsh environment for such a component (moisture and heat can easily cause corrosion and mechanical stress). Also, I guess timing gets less accurate over time. If there are rubbers preventing moisture from entering and oil from leaving, those rubbers will wear out.

[Fatnino]

Like another commenter said, the timer is segregated from the harsh dryer environment.

But also, some slop in the timing accuracy is just fine. The user doesn't really know how long precisely the drier needs to run to dry their clothes. They just know that if they set this timer to 45 then the clothes come out dry.

[purpleflame1257]

The timer for the dryer is typically located outside of the dryer drum itself. It is not an actuating component.

[beezlewax]

I have one that is like 40 years old. The heating element is similar to those in an oven.

The only things that broke are plastic door hinges.

[cj]

I wish there were an easy way to screen these type of engineers out in the hiring process. It's very hard to judge whether a candidate's excitement over new technology is simply showing passion for what they do or a red flag. A certain degree of passion is desirable, but too much is not.

The worst engineers I've worked with are ones who, in their first week, fall behind on their onboarding plan because the company's compiler needs a rewrite.

[mrguyorama]

Ah yes, the mythical world where engineers are in charge of deciding which features ship instead of management.