The Inner Beauty of Basic Electronics
spectrum.ieee.org
spectrum.ieee.org
Many people find beauty in art. I find beauty in engineering perfectly suited to its task. Like the stunning beauty of a gas turbine engine.
The P-51 or Spitfire would lose to a T-28 or F8F, but the former look better to us. We are attuned now to designs like the Su-57 and F-22, so are less sensitive to the merits of systems that work a different regime.
Or maybe not so surprising. I had an eagle fly by me once at about arm's length, and at shoulder height. I was stunned by its deadly beauty. Right then I knew why so many countries adopt the eagle as their mascot.
Uhm...
> It is as if there were a natural law which ordained that to achieve this end, to refine the curve of a piece of furniture, or a ship's keel, or the fuselage of an airplane, until gradually it partakes of the elementary purity of the curve of a human breast or shoulder, there must be the experimentation of several generations of craftsmen. In anything at all, perfection is finally attained not when there is no longer anything to add, but when there is no longer anything to take away, when a body has been stripped down to its nakedness.
> It results from this that perfection of invention touches hands with absence of invention, as if that line which the human eye will follow with effortless delight were a line that had not been invented but simply discovered, had in the beginning been hidden by nature and in the end been found by the engineer. There is an ancient myth about the image asleep in the block of marble until it is carefully disengaged by the sculptor. The sculptor must himself feel that he is not so much inventing or shaping the curve of breast or shoulder as delivering the image from its prison.
> In this spirit do engineers, physicists concerned with thermodynamics, and the swarm of preoccupied draughtsmen tackle their work. In appearance, but only in appearance, they seem to be polishing surfaces and refining away angles, easing this joint or stabilizing that wing, rendering these parts invisible, so that in the end there is no longer a wing hooked to a framework but a form flawless in its perfection, completely disengaged from its matrix, a sort of spontaneous whole, its parts mysteriously fused together and resembling in their unity a poem.
-- Antoine de Saint-Exupéry, _Wind, Sand, and Stars_
Now things are holonomically controlled, every behavior is arbitrary and computer generated, nothing arises from necessity if we can avoid it.
With a spoon, anything you add will probably make it worse, you can sit and think and not find anything to change other than the material. If I give it adjustable length, the mechanism may be unsanitary.
There's no long list of features to pick from, once you've decided you're not making a spork, there's no real question of what to include, so the feature set(Be spoon) feels perfect, since I have no alternate design to compare it to.
Once you add a microcontroller you have so much possibility. Why can't my flashlight report battery level via Bluetooth and alert me when it's fully charged? A minor convenience, but it could be cheap enough I'd pay for it.
The tradeoff is a bit more code and a cheap chip. And I'm sure there's a dozen other features that I'd like to have if they were cheap.
Not only does the thought of a perfect harmony of only the necessary parts not always arise in a modern mindset.... but I wouldn't know how to recognize it if I saw it anyway. Is there a real engineering reason to leave out a feature? Did they not think of it? Or are they just appealing to love of simplicity for it's own sake? I don't know, I wasn't at the meeting, so I'll probably be annoyed and leave a comment saying they shouldn't have left this or that out.
Now, simplicity feels more like another feature to include or exclude, it doesn't just arise from the desire for low cost and reliability, since we sometimes but not always can make insane complexity cheaply that lasts decades.
The difference is that simplicity isn't compatible with a lot of other features, because it's almost like intentionally choosing to give up arbitrariness and holonomic control.
Respectfully, this is completely untrue. If you really study spoons you will notice they exist in various scales, shapes, depths and materials for good reason. Furthermore, some are single use while others are optimized for longevity, various surface properties, weight, reduced physical envelope, ergonomics, aesthetics, strength, rigidity or flexibility, handling of liquids-vs-solids, pouring, surface piercing, tasting, use by machine, specific length, safety, environmental impact, regulatory requirements, measurement, cost of manufacture, logistic concerns or some other set of requirements.
Some clearer examples to consider are a wooden spoon used for mixing in a traditional kitchen baking context, a foldable single-use polymer spoon atop a yoghurt container, a traditional English teaspoon with decorative enamel, a ground coffee spoon, measurement spoons and the entire category of ladle-like instruments.
(Source: Stepped out of pure software 7 years ago to work in food robotics.)
I'm sure we all have utensils we like and dislike, but the feature set is pretty much the same on all of them, especially within a category of specialty type, the difference is mostly decorative and ergonomic.
Wheras with tech, we often add things that are only tangentially related to the original purpose, and we add layers of indirection between user and the actual purpose, with millions of ways that mapping can happen.
Two variants of nominally the same idea can have totally different possible applications, like a phone with and without a GPS. These days all smartphones have GPS, but what about UWB? What about a stylus? Or wireless charging?
They're all smartphones, sold for basically the same general market rather than some specific application, but everyone has a different set of features they want, because it's not a single-function device.
Some people might like more or less ornamentation, and some people might prefer silver over stainless steel, but even most of the specialized types don't have extra features, and nobody picks them up and wishes they did, and rarely do people prefer spoons to be multi-purpose.
With point-to-point electronics at the time, the relatively large components were soldered into interesting 3-dimensional patterns protected by a metal chassis which normally had plain terminal strips and the underside of vacuum tube sockets to use as connection points.
With the proliferation of circuit boards, the components were lined up a lot more like Lego on a substrate but still more colorful.
Imagine every circuit board more colorful than a pile of random small Lego bricks today.
And it was even more obvious there had to be a pattern there that would be worth recognizing.
I liked building with Lego, but I liked building with this stuff a lot more.
That way, I could easily tell if the wiring was off (the posts are .1" apart, and it is easy to get off by one).
With these patterns, I rarely had the board go to smoke testing with a wirewrap error in it. This made me popular with people who needed a board done.
I did this also with boards to be soldered. All the resistors were lined up the same way, the diodes all pointed the same way, the chips had the #1 pin in the same place, etc. It all made for patterns that made checking easy.
I once was talking to a tech at a high end audio store. He said you could tell the high end boards because the resistors all pointed the same way. Same thing. A constructor contractor told me a proxy for quality work is the screw slots were all at the same angle.
I use the same technique with software. I try to lay out the code so it forms a visual pattern, this makes it easy to check for errors. It's also why I don't use source code formatters, as they disrupt the patterns.
I don't have to know the language to see that your code is built in a very orderly fashion and the components are intended to interact in a carefully-defined way.
Not unlike a well-engineered circuit board.
One picture I really enjoyed was the cutaway of a barrel jack power plug. The author explained how plugging in a power cable would physically change the power source from an internal battery to something external. For a CS person, this physical elegance blew my mind.
In many cases you can write your code in an analogous way. Sadly, most procedural code I see plods its way through transitions in a manner that allows errors, races etc (the old thermos joke comes to mind). Keep these positive hardware approaches in mind when designing your algorithms.
"You can put a cold drink in it and no matter how hot it is outside the drink stays cold".
The friend is suitably impressed.
"But that's not all: you can put a hot drink in it and no matter how cold it is outside the drink stays hot!"
Now the friend is puzzled. "But how does it know?"
I long ago became depressed by the amount of code I've read that is written this way. Nowadays I just expect it.
You can see they sorts of thing they were making here:
https://web.archive.org/web/20220816124638/https://www.compu...
Not sure if the store is still functional (or if they're still alive). (Server probably underpowered for HN attention):
The math for a flywheel and an inductor is the same. Springs are capacitors, and dampers are resistors. Math is the same.
(Dampers are still resistors in this model, though you must consider resistance to be how "loose" the damper is, rather than how "tight" it is.)
And, current and velocity are both motions, while force and voltage are both pressures.
Applying voltage to a capacitor is an incorrect mental model -- the voltage across a capacitor is a function of its charge [1]; it can't be both that and the voltage you apply to it.
In reality, when you apply voltage "to" a capacitor, it's to the series circuit comprising the capacitor and a (hidden or explicit) series resistance, thus causing a current to flow determined by the resistance and the net voltage across it.
The correct mental model is that you inject current into a capacitor. So long as you inject current, the voltage across the capacitor rises. viz. a flywheel -- so long as you apply a force, the flywheel speeds up.
> A flywheel speeds up as long as you apply a force to it; an inductor's current increases as long as you apply a voltage to it.
Likewise, a spring gains tension so long as you keep elongating it.
Inductors and capacitors are dual to each other, so -- ignoring the mechanical linkages themselves -- there's no inherent advantage for one mechanical analog or its dual. We are thus free to choose either to make a sound analogy of the physics of linkages:
> And, current and velocity are both motions, while force and voltage are both pressures.
This grammatical correspondence is not relevant to the mathematics. Kirchoff's laws [2] only hold for mechanical linkages when current is held analogous to force, and voltage is held analogous to velocity.
A more grammatically satisfying analogy can be found in hydraulic systems, with current : flow :: voltage : pressure. Flow, unlike motion, sums at nodes; pressure, unlike force, is equal at nodes.
If one is insistent on treating springs like capacitors and flywheels like inductors, you can do so, but only if you also take the dual of the full circuit, as seen here [3]. This is then a mathematically correct analogy, but personally I find it much less straightforward than simply swapping springs and flywheels.
[1] https://en.wikipedia.org/wiki/Capacitor#Current%E2%80%93volt...
[2] https://en.wikipedia.org/wiki/Kirchhoff%27s_circuit_laws
[3] https://www.ams.org/publicoutreach/feature-column/fc-2019-05
The large photos printed in very high resolution blow away what you can see on your screen.
https://news.ycombinator.com/item?id=32630740 (43 comments)
Eric previously designed the MOnSter 6502, built with Windell's help: a 6502 CPU with fully discrete components! https://monster6502.com/
Windell is one half of Evil Mad Scientist Labs: https://shop.evilmadscientist.com/
I have a giant chart of synth designs framed in my living room.
I've got a CS degree but from a university where CS came from the mathematical school, rather than the engineering school. Hence in four years, there was only one (optional) class where a computer was used, all other lectures where theorems and proofs with pen and paper.
But I would really love to be able to tinker with circuits, not necessarily with logic circuits, but being able to get some motors to do something useful or - more likely - fun has been something I've always wanted to learn. Mind you, it wasn't at the top of cool things I'm interested in, so I never really made too much of an effort. But that was also partly because I found it surprisingly difficult to find good resources to get started. Compared to say, programming, or other CS topics.
I'm sure this is partly because as a complete novice, it's already challenging to even assess what would constitute a good resource and what not. But I've just always failed to find a book or course that was at the right level of complexity (for me) and caught my interest sufficiently to actually stick to the topic. Either it's quickly over my head or it's something like 45 minutes of youtube videos just to get to Ohm's law, or so.
Eventually, I discovered some (older) books from the model train community that I bought for next to nothing from ebay. Those are quite okay, I like that they are quite practical in that they refer to actual electronic parts with the names/IDs that you can go out and buy. The downside, again, is that they're quite light on the theory - the kind of writing that tries very hard not to alienate the reader and thus doesn't dare to properly explain anything of more than trivial complexity.
Therefore, if anyone has any good recommendations for books, courses, videos, or other resources that you have found useful and that are tailored to beginners but dare to confront them with the full monty and not just some hand-wavy watered-down version of things AND (ideally) are hands-on so that when you're done, you'd be able to order some parts and be able to actually put something together, I'd very much appreciate some hints.
(Is that wish list simply too long?)
if you used the thick wire on both sides you'd get the same voltage but have a much heavier and more expensive transformer for an almost undetectable performance improvement
They also share a bit about the equipment and process used for making the pictures.
Some of the slides should be a part of every electronics course in every school and University.
It looks like something that wouldn't be out of place on the deck of the Discovery One in the movie Space Odyssey.
When they find deposits of these things, would they be able to figure out what they were used for?