How a car cigarette lighter works, in CT scans | Hacker News Reader

How a car cigarette lighter works, in CT scans

lumafield.com

30 pointsjonbruner29 comments

Seeing clever, simple, and reliable mechanisms like this make me think that digital electronics has really enabled modern engineers to be lazy.

Well, yes but.

Mechanisms like this can be extremely difficult to design. You have to go through many iterations to tune the performance and ensure it's reliable over a vast range of conditions. Then you have to do it all over again to make the design mass-manufacturable. The design is extremely sensitive to the quality of the input materials and could outright fail if not manufactured just so.

It requires a lot of education and ""real"" engineering. It takes a lot of very expensive engineer hours.

Meanwhile the digital equivalent is a weekend project at best, and can be designed by just about anyone with half a brain and one working eye.

I agree that mechanical mechanisms are almost always much more elegant and simple than a digital equivalent, but given that the digital versions exist and perform nearly as well (if not better), the pure mechanical solution just isn't practical.

I'm not entirely sure if this is a good or bad thing. It certainly feels like we've lost an art form, but are we simply moving from flint arrows to steel, or something more like the loss of Roman concrete?

I think it's more like Roman concrete. There are many creative engineering possibilities that simply aren't used anymore because there are so many jellybean parts on the shelf that can brute-force a solution without much creativity.

Much in the same way, I find many software engineers these days don't actually know much about how a computer or a computer network works.

Engineering is starting to become less about being masters of the physical world, and more like a discipline in gluing blocks together.

Computation on the bare metal of the universe, no abstraction or emulation required. I love it.

I think that analogy to programming then and now is very fitting.

> You might think such a system would involve a thermostat, a microcontroller, and an actuator, but this design [...] has bimetallic arms with steel on the outer surface and copper on the inner surface. As the filament heats up, the arms warm up, causing the copper to expand faster than the steel. This expansion pushes the arms open, releasing the lighter. The coil spring then causes the handle to pop out, signaling that the lighter is ready.

Maybe I am in a picky mood, but for me the intro made it sound like all three items (thermostat, microcontroller, actuator) would not be necessary, though of course only the microprocessor is missing from the design (thermostat -> bimetallic arms, actuator -> spring).

Mandatory link to another marvel in the same problem space (I found the 19min video worth the time):

"An Antique Toaster That's Better Than Today’s [video]"

https://www.youtube.com/watch?v=1OfxlSG6q5Y

https://news.ycombinator.com/item?id=21164014 [2019, 386 points|5 years ago|232 comments]

No, you have the correct take. The bimetallic bond is a thermostat. The spring is an actuator.

Heh, I correctly guessed the channel just by reading the title :)

It's a cool explanation, but I'm not sure what a CT scan offers that cutting it in half with a bandsaw wouldn't.

When all you are is a hammer manufacturer, all your demos look like nails.

In which case the question gets inverted:

Why of all common things you could CT scan and show people, you choose something that is kinda obvious to anyone who has thought about it for five seconds? I don't think average people struggle to understand the concept of "electricity makes wire hot", especially if they lived at the same time as normal car cigarette lighter usage.

Lumafield posts tons of these. I don't think the average person has thought at all about how these work, and even if they do, the appeal of these is to look at the pretty pictures (and demonstrate the analysis that is possible non-destructively to their target audience), not to understand "electricity makes wire hot".

They have a scan of a dang football, and it isn't because people don't know there's air inside: https://www.lumafield.com/article/ct-big-game-football

Those who have thought about it for more than five seconds understand that there is more going on inside of a normal car cigarette lighter than "electricity makes wires hot."

IR heat profile might have been cool.

It's a shame cars don't come with them anymore. A lighter can save someone's life and cost next to nothing. You can toss a lighter in but something about the designated electric plug is pleasing and invites you not to lose it.

Yes, very handy even for non-smokers. We have a 2015 vehicle that we bought used that came with a cigarette lighter. I removed it and put it in the glove box.

It's a scanner company showing off their scanners, they don't sell bandsaws. :p

That said, it's a kind of marketing I like, since it does have some underlying product-performance and educational virtue to it.

Same reason every coding tutorial -- no matter the capabilities of the language -- starts with a simple "hello world".

Learn with a simple toy example before moving onto the industrial-grade deconstruction that really shows the power of the tool. These guys have done CT scans of things like real vs. fake airPods and charger bricks to show what really justifies their (probably) 7 or 8-figure price tags.

This one is just content marketing of something anyone can wrap their head around.

> this design, developed in the 1950s, is entirely analog

Uhm, how is this analog? Perhaps "electro-mechanical?" Is there such a thing as a "digital cigarette lighter?"

It's not like the electric circuit is an analog of the act of lighting a cigarette; as opposed to what we commonly use the word "analog" to mean that something in the system is an analog to what's being recorded or processed. (IE, in analog tape the magnetic signal is an analog to the real signal, or in analog TV the signal is an analog to the brightness of the picture at a specific moment in time.)

It's a valid distinction of the author to make. I don't know if there's such thing as a cigarette lighter which uses digital circuitry, but there theoretically could be.

The way the lighter in the article works is both analog _and_ electro-mechanical.

No, go look at the definition:

https://www.dictionary.com/browse/analog

> of or relating to a mechanism, device, or technology that represents data by measurement of a continuous physical variable, as voltage or pressure.

"represents data by measurement of a continuous physical variable"

Did you read the article? The cigarette lighter represents data (pops out) based on measurement of a continuous variable (temperature)

Yes. The latch is a bi-metallic strip that moves.

There is nothing representing the temperature. IE, to meet the definition of analog it would need something like a thermometer resistor where the resistance is used by some electronics to turn off the main heater.

The pressure provided by the bi-metalic strip is representing the temperature, and it's turning off the main heater by causing the unit to pop out. The "signal" doesn't have to be voltage, it can be pressure or spacial position as well.

> Is there such a thing as a "digital cigarette lighter?"

Yes: https://www.walmart.com/ip/Flameless-Cigarette-Lighter-USB-R... :-)

It is an analog of a traditional burning fuel lighter. But they are using analog to refer to the fact it relies completely on analog circuitry and I suspect they actually wanted to point out the fact that it is 100% passive circuitry relying solely on the electrical and physical properties of a resistor (the coil) which is the interesting thing here, the resistor is its own temperature sensor. They sort of explain this and walk you through it all, probably trying to stick to language suited to their audience.

> it relies completely on analog circuitry

A resistor is not "analog circuitry"

> the resistor is its own temperature sensor

No, the clasp that holds the spring down is a bi-metallic strip that bends when it reaches a temperature. This is not "analog" because there is nothing for it to represent.

The article is merely a case of calling something old "analog" without knowing what the word means.

A resistor is not circuitry, the resistor is a part of a circuit which in this case is analog.

This circuit does not "represent" anything, and thus is not analog.

Look up the definition.

https://www.dictionary.com/browse/analog

> of or relating to a mechanism, device, or technology that represents data by measurement of a continuous physical variable, as voltage or pressure.

I am an OED man myself.

>Relating to or operating with signals or information represented by a continuously variable physical quantity, such as voltage, spatial position, etc., which in the case of measuring instruments can be displayed on a dial or other fixed scale. Also: designating a signal represented in this way.

But the bimetal strip represents the data that is the temperature of the coil and the coil represents the data of the current. We can use both as sensors and we do, the bimetal strip was ubiquitous in thermostats until the rise of the digital thermostat and still used, resistors as sensors are still common and used all over the place in electronics.

Reply on news.ycombinator.com