Join 3,434 readers in helping fund MetaFilter (Hide)


Things are not as they seemed in EE101
March 22, 2012 10:05 AM   Subscribe

A series circuit of 1 battery, 1 resistor, 3 switches, and 3 LEDs. What happens when you turn on each switch?

Solution. Pause at 0:24, if it helps you.
posted by 0xFCAF (90 comments total) 47 users marked this as a favorite

 
How does it work?

Obviously, elvish magic. There are small elves in the wires. QED.
posted by Wolfdog at 10:13 AM on March 22, 2012 [1 favorite]


BRILLIANT! Thanks so much!
posted by JimmyJames at 10:14 AM on March 22, 2012


Yes, if you pause the solution at 0:24, you can clearly see the characteristic runes.
posted by Wolfdog at 10:15 AM on March 22, 2012 [2 favorites]


I clearly understand nothing about electricity and circuits, because I don't even see why this is supposed to be surprising. I haven't felt this baffled in a long time.
posted by King Bee at 10:17 AM on March 22, 2012


A certain segment of the population would look at this and say, "so what?" because they don't understand how electricity works and well, why wouldn't switches turn on lights like that? Isn't that how it works? You just wire stuff up and Bob's your uncle.

Another segment will look at this and say, "meh" because they're already knowledgable about RF and bandpass and all that good stuff and will have figured it out after a couple seconds.

But if you know just enough about electricity to know how to wire up a house, or maybe fix a broken lamp or something, this is magic.
posted by seanmpuckett at 10:19 AM on March 22, 2012 [27 favorites]


I was curious why LED symbols weren't used in the original schematic. I wondered maybe he somehow overlooked them.

Anyhow, I was hoping for some kind of paradox behavior rather than a trick circuit that can't be spotted on a youtube clip.
posted by 2N2222 at 10:21 AM on March 22, 2012 [7 favorites]


I think the magic is in the switches. If I had to guess there is an oscillation in the current (create by each switch) that puts a single LED in phase.
posted by kuatto at 10:22 AM on March 22, 2012 [1 favorite]


Another segment will look at this and say, "meh" because they're already knowledgable about RF and bandpass and all that good stuff and will have figured it out after a couple seconds.

This was posted all over electronics sites/boards, and almost nobody figured it out. The solution was posted a long time after the first video was. The consensus was that it was "photoshopped."

Once the solution was posted, the consensus was that this was the most epic physics troll, and also that the guy is a deadbug soldering GOD. Like, holy shit.
posted by Threeway Handshake at 10:22 AM on March 22, 2012 [14 favorites]


So, if I want to understand why this is "magic", where do I start?
posted by King Bee at 10:23 AM on March 22, 2012


I'm definitely in that "magic" segment. It's magic because we expect electricity to move like a flow (somewhat like water) where the switches act like gates on a path. No power gets through without all three gates being open. Once the three gates are open, the power should light up all three LEDs.

Any chance at a plain English explanation of how he's done this?
posted by Wulfhere at 10:24 AM on March 22, 2012 [1 favorite]


(deadbug meaning, that the circuit is a surface mount, usually soldered into a board where the pins will fit into holes, but it is flipped over and freestanding so the pins make it look like a dead bug)
posted by Threeway Handshake at 10:24 AM on March 22, 2012 [1 favorite]


To help explain the this is magic feeling to the "so what?" people, imagine if you had a garden hose with three people standing on it so the water wouldn't go, and three sprinkler heads on it.

If person A steps off the hose, sprinkler A goes off. But there are two other people still standing on the hose! Then person C steps off, and sprinkler C goes. BUT PERSON B IS STILL ON THE HOSE!

That's what it looks like, but with electricity.
posted by seanmpuckett at 10:25 AM on March 22, 2012 [7 favorites]


So he's good at soldering, and maybe that's a clever circuit... surely there are better uses of this than prop magic?
posted by cmoj at 10:26 AM on March 22, 2012


I am squarely in the 'magic' bucket because even with the solution schematic I have no fucking clue what's going on. I can understand how you might make a circuit that can do 1+2, 2+3, or just 2, but I don't understand how the 1+2+3 case works.

King Bee: The magic here is that each of the 3 switches in the top row should, when in the "open"/"off" position, interrupt the flow of current through the entire circuit, turning off all 3 LEDs. Once all 3 switches are on, current should flow through the entire circuit, lighting up all 3 LEDs simultaneously. In no normal configuration can you have 3 LEDs in a row ("series") where they can be lit and unlit independently. Think of old-timey Christmas lights where one dead bulb means your whole strand is unlit.
posted by 0xFCAF at 10:26 AM on March 22, 2012 [1 favorite]


More information on "dead bug" circuit building.
posted by filthy light thief at 10:27 AM on March 22, 2012


Why this seemed photoshop/magic:

It is in series. All switches on should make all the lights on. Any switch being off should make all of the lights turn off.
posted by Threeway Handshake at 10:28 AM on March 22, 2012


So the secret is lie on the original schematic... Got it.
posted by MrBobaFett at 10:30 AM on March 22, 2012 [9 favorites]


That's terrible soldering. A soldering iron is not a spoon for solder.
posted by le morte de bea arthur at 10:36 AM on March 22, 2012 [1 favorite]


Any chance at a plain English explanation of how he's done this?

None of the components are what he tells you they are in the original video. The "battery" emits a complicated waveform, the "switches" are switchable filters, and the "LED"s contain corresponding circuitry which lights them up only if their corresponding frequency component hasn't been filtered by one of the "switches". The impressive bit is that he has hidden the extra circuitry inside what appear to be ordinary components.
posted by hattifattener at 10:37 AM on March 22, 2012 [12 favorites]


He cheats. There's circuitry hidden in the battery plug.
posted by pracowity at 10:38 AM on March 22, 2012 [1 favorite]


So the secret is lie on the original schematic... Got it.

Yeah, pretty much. But I wouldn't put it quite so uncharitably.

While the actual circuit is clever enough, the real trick here is that the circuit is small enough so as not to be visible on a youtube resolution video. If someone built the circuit with conventionally sized components, it would still be clever, but the video would be a bit of a yawn.

I can't really comment on how the circuit actually works. It isn't very intuitive, in that the LEDs are driven by both a rf oscillator and a audio range oscillator. It seems by various clever polarity and filtering changes, the miniaturized circuit appears to exhibit paradoxical behavior. That's about where my chops end without hurting my head.
posted by 2N2222 at 10:39 AM on March 22, 2012


That's terrible soldering.

Everybody disagrees with you.

posted by Threeway Handshake at 10:39 AM on March 22, 2012 [4 favorites]


It's magic, because just like with magic tricks, the presenter fools the audience into accepting that certain things are plainly one thing, while they are actually another. Those switches aren't just switches, the LEDs aren't just LEDs, and the battery connector isn't just a battery connector. The diagram placed behind the circuit he is constructing reinforces the illusion, so to speak.
posted by jamincan at 10:39 AM on March 22, 2012 [6 favorites]


Remind me never to take you guys to a magic show. "Oh so he doesn't actually saw the woman in half? Got it. Cheater."
posted by 0xFCAF at 10:40 AM on March 22, 2012 [6 favorites]


Impressive. The solution is particularly enlightening.
posted by cloeburner at 10:41 AM on March 22, 2012


0xFCAF: "A series circuit of 1 battery, 1 resistor, 3 switches, and 3 LEDs."

1 * 1 = 0! Here is my proof!

oh yeah I secretly modified * so that it would mean -
posted by idiopath at 10:41 AM on March 22, 2012 [6 favorites]


the circuit is small enough so as not to be visible on a youtube resolution video.

Looking at the "making of" video it seems to me that he could give that thing to you in real life and you'd still be fooled. The quality of the deception far exceeds what is required to make that work on a YouTube video (e.g., melting plastic over the little doodads he adds to the bottom of the first LED--you don't even see the underside of the LEDs in the video).
posted by yoink at 10:44 AM on March 22, 2012 [2 favorites]


Remind me never to take you guys to a magic show. "Oh so he doesn't actually saw the woman in half? Got it. Cheater."

Oh great! Thanks for spoiling like every magic show ever dude!
posted by yoink at 10:45 AM on March 22, 2012


Crap soldering technique.
posted by ryanrs at 10:51 AM on March 22, 2012


[Jimi Hendrix video]

"He plays the guitar improperly."
posted by Threeway Handshake at 10:54 AM on March 22, 2012 [7 favorites]


"If you want to find components to this solution, please check advertisement that YouTube arrange to my movie."

See, this sentence has three verbs, three nouns and two items of punctuation, and yet this sentence does not do what it is supposed to do."

English major represent!

(Put me in the camp of IANAEE people who think flipping switches turns lights on and off and that is magic enough.)
posted by chavenet at 10:57 AM on March 22, 2012 [1 favorite]


[Jimi Hendrix video]

"He plays the guitar improperly."


Game, set, match.
posted by RolandOfEld at 10:57 AM on March 22, 2012


As someone who has only soldered something once in his life and did a terrible job of it, is there maybe some sort of disagreement-in-terms here where some folks are saying "he soldered really well at such a small scale, amazing" and other folks are saying "his solder points are messy/he shouldn't use the gun to hold the melted solder, sux" and y'all are just talking past each other, or is there something more interesting or complicated in terms of competing soldering dogmas going on? Help a layman out.
posted by cortex at 10:58 AM on March 22, 2012


Hackaday followup.
posted by pashdown at 11:02 AM on March 22, 2012


Not really. He's just sloppy because better technique would get in the way of the video. You'd need a tool to hold the components in place, which would interfere with the sparse "nothing up my sleeve" presentation.
posted by ryanrs at 11:02 AM on March 22, 2012


I've done some practical soldering and have hung out with people who were way more into soldering stuff up (EE students for the most part). My opinion, and that of the people I hung out with, has been that if it gets the job you were trying to accomplish done and lasts as long as you need it to, you've done fine.

This guy's hack works as intended, was a great demo, and seems to hold up just fine to being moved around a bit.

Haters gonna hate with regards to him putting solder directly on the tip instead of letting the heated points to be soldered 'wick' the solder onto them as it goes molten from conduction heat transfer I guess. To that I say, enjoy.
posted by RolandOfEld at 11:06 AM on March 22, 2012 [3 favorites]


His deadbug soldering technique to conceal all the SMD parts inside the 'big' components is quite good.

His soldering technique during the video is terrible, but probably purposely so as part of the misdirection.
posted by Confess, Fletch at 11:10 AM on March 22, 2012 [2 favorites]


The reason people get worked up about it is that in industry proper soldering is actually very important and somewhat difficult.
posted by ryanrs at 11:11 AM on March 22, 2012


I imagine it's kind of like seeing a mechanic pull out a crescent wrench. You just think, "gah, don't do THAT!"
posted by ryanrs at 11:12 AM on March 22, 2012 [3 favorites]


Question -- I've always wondered if there's a really good, simple "electronics simulator" for Windows or iOS that lets us build simple circuits and see what happens. I guess it would be limited to resistors, switches, lights, and transistors, though it would be cool to stick a quartz piece in there and suddenly be listening to Radio Canada on HF.

I've dabbled with Minecraft redstone, but it's not very realistic and it's kind of its own thing.
posted by crapmatic at 11:15 AM on March 22, 2012


(joking about the radio receiver bit; I know it's not technically possible)
posted by crapmatic at 11:15 AM on March 22, 2012


This isn't brilliant. It's just misleading.
This is how computer voting will work.
posted by weapons-grade pandemonium at 11:17 AM on March 22, 2012 [10 favorites]


I went through: super confused > excited by the possibility of an elegant bit of science > disappointed to find out it was actually a more complicated circuit > super impressed.

The only thing that makes it feel like a "cheat" is the circuit diagram. I like it when a magician works their patter out such that they never actually tell you anything that isn't technically true.
posted by lucidium at 11:19 AM on March 22, 2012


Ok, wow.

On the one hand, I suppose I'm glad to know that my understanding of basic electrical circuits isn't horribly flawed, and that it's a magic trick with disguised filters and whatnot.

On the other hand, I'm disappointed because my understanding of electric circuits is quite limited and I'd hoped there was some weird, deep, thing going on that real EE's know about and guys like me don't know.
posted by sotonohito at 11:22 AM on March 22, 2012


To explain it in non-electrical engineer terms, imagine the switches are dams on a river and the LEDs are water wheels. In order for the the water (electricity) to flow, you can imagine that you would have to open all of the dams for the water to flow to the water mills. If the water reached the wheels, all of them would turn, not just one. The idea of opening and shutting individual dams to turn different water wheels just doesn't make sense. To a EE, the circuit shown makes no sense.

The solution is the equivalent of not mentioning the elaborate system of underground pipes and switches that were installed in the dams and water wheels that you can't see. Using filters and

Good engineering, lame question.
posted by Argyle at 11:22 AM on March 22, 2012 [1 favorite]


Also, I am enjoying the soldering enthusiast grarr over technique.

Ah, the days of arguing over the proper amount of solder that should be left in the cup of a XLR connector. Good times.
posted by Argyle at 11:25 AM on March 22, 2012 [1 favorite]


This is in reference to the circuit shown at the 24 second mark in the solution video.

The trick is that the LEDs are driven by AC current instead of DC current. AC current will work just fine except that, since the LEDs are diodes, current passes through an LED only for half the cycle or 50% of the time, causing the LED to blink at the AC frequency. However, since the frequency is much faster than the eye can see, the LED looks like it is on steadily. Pushing twice the current through an LED half the time gives the same effective brightness as a steady current. This is how you adjust the brightness of your LCD screen that has an LED backlight. You just vary the duty cycle of the on time from 0% to 100% at some high frequency.

The AC generator in the battery connector combines two frequencies on top of each other 300 KHz and 5000 KHz so that the AC current contains both frequencies. The circuit is designed to respond differently to each frequency.

Capacitors in LEDs 1 and 3 and Switches 1 and 3 are high pass. They act like a short circuit for 5000 KHz and an open circuit for 300 KHz. So for the 5000 KHz signal, these components are just like shorted wires. It is as if LEDs and switches 1 and 2 and diodes aren't even there. However for the 300 KHz signal, the capacitors look like open circuits leaving just the LEDs and diodes. Note that Switch 1 and LED 1 have their diodes facing in the same direction while Switch 3 and LED 3 have their diodes both facing in the other direction. When both switches are open, the 300 KHz current is blocked in both directions by the opposite facing diodes in Switches 1 and 3 so both LED1 and LED 3 are off. When you close Switch 1, it allows current to bypass the diode in switch 1 and pass clockwise in the circuit. Since LED 1 has a counterclockwise diode blocking clockwise current, the current is forced through LED 1, turning it on. Meanwhile LED 3 has a clockwise diode that shorts across LED 3, so the current takes the path through the diode instead of LED 3, so it remains off.

When you close Switch 3, it allows current to flow counterclockwise, and similarly to above, it turns on LED 3. So LED 1 is controlled by clockwise current and LED 3 is controlled by counterclockwise current.

So what about Switch 2 and LED 2. The inductors are high frequency notch filters. They act as a short for the low frequency 300 KHz AC so they always allow 300 KHz through to control LEDs 1 and 3. It is as if Switch 2 and LED 2 aren't even there for 300 KHz, so that LED 1 and LED 2 respond only to the 300 KHz as described above.

However the inductors block the 5000 KHz AC. So when the Switch 2 is open, no 5000 KHz AC can get through to LED 2. Meanwhile the 300 KHz passes right through and goes to LED 2, but since LED 2 has a similar inductor that is a short for 300 KHz, the 300 KHz bypasses LED 2 and it remains off. When you close Switch 2, now the 5000 KHz can get through to LED 2 and since the inductor in LED 2 blocks 5000 KHz, the current is forced through LED 2 turning it on. Meanwhile the 5000 KHz just passes through the high pass capacitors on LEDs 1 and 3, bypassing the LED1 and LED 3 so they remain off.
posted by JackFlash at 11:25 AM on March 22, 2012 [23 favorites]


A series circuit of 1 battery, 1 resistor, 3 switches, and 3 LEDs. What happens when you turn on each switch?

Someone you do not know will die.
posted by Renoroc at 11:33 AM on March 22, 2012 [7 favorites]


obvious trickery is obvious
posted by charlie don't surf at 11:48 AM on March 22, 2012 [1 favorite]


So the secret is lie on the original schematic... Got it.

Yeah, pretty much. But I wouldn't put it quite so uncharitably.


Would you prefer "he completely misrepresented the facts in order to deceive"?

Since he doesn't flip the switches over, it amazed me for about 3 seconds. I didn't figure it out, but if the magician won't let me see the table underneath the hat, I don't assume the rabbit "vanished". I assume there's a hole; just didn't figure out the shape of it, or how he trained the rabbit.
posted by IAmBroom at 11:49 AM on March 22, 2012


crapmatic, SPICE (and variants) are the usual electronics simulator. Not especially simple.
posted by ryanrs at 11:49 AM on March 22, 2012


Both the idea and the assembly of the trick are unique and non-trivial. I don't understand why everyone is in such a rush to prove how unimpressed they are. (I'm not -surprised- that people need to have this dick measuring contest on a constant basis, but that doesn't mean I understand it)
posted by flaterik at 11:53 AM on March 22, 2012 [2 favorites]


Question -- I've always wondered if there's a really good, simple "electronics simulator" for Windows or iOS that lets us build simple circuits and see what happens.

How about a Java one? Analog circuit simulator on falstad.com (which was the subject of a 2005 MeFi post, which was probably where I learned of it).
posted by FishBike at 11:59 AM on March 22, 2012 [1 favorite]


I am not impressed by NASA putting a man on the moon in the 60's. They used computers and spaceships, not wings made of wax and feathers.
posted by Ayn Rand and God at 12:00 PM on March 22, 2012 [6 favorites]


This isn't a remarkable technical feat. Using highpass and lowpass to separate parts of a signal is not cutting edge, it is how basic electronic things have worked for over a century. Power supply filtering, radio, telephony, having separate speakers for the subwoofer and tweeter, tape recorders, electric phonographs, DAC, ADC, oscillators... come to think of it it, beyond light switches and toasters, it may be easier to list the things that don't use capacitors and/or inductors to filter a signal based on frequency.

The only surprise here is that what looks like (and is referred to as) a basic off the shelf standard electrical component has secret extra components hidden inside it. This is the electronic engineering equivalent of your grandpa pulling a quarter out of your ear.
posted by idiopath at 12:17 PM on March 22, 2012 [2 favorites]


This reminds me of that Teller article where he said "Make the secret a lot more trouble than the trick seems worth."

When I saw how he did it, wow. That's a lot of work for such a small light show. But it worked.
posted by fungible at 12:17 PM on March 22, 2012 [2 favorites]


Whether or not his soldering technique is correct is not something I cannot comment on, but damned if I don't find it relaxing to watch someone solder components together.
posted by Brackish at 12:26 PM on March 22, 2012


I'm an EE with steady hands and soldering experience, and I am highly impressed with his abilities.

I would love to know if his soldering during the troll video was done simply to save time and complexity, or intentionally done as a means of misdirection. Well played, EE troll, well played.
posted by I Havent Killed Anybody Since 1984 at 12:31 PM on March 22, 2012


BTW, you people who can't solder: buy a Kester 186 flux pen. Costs $5.
posted by ryanrs at 12:33 PM on March 22, 2012


That was lovely. It confused me utterly. The solution was lovely and complicated and invisible (to me).

0xFCAF says it best. Seems overly harsh to hate on his technique. Like lumberjacks critiquing the sawing technique of the magician. As someone who solders all the time, his technique, while not perhaps industry standard, was something I personally envied immediately. And I'm pretty good at it.

He set out to fool people with a deceptively simple circuit. Did so. He succeeded on every front he intended. If he didn't fool you then bully for you, but it doesn't negate the pleasure I felt at being fooled. Nor his at succeeding.

My hat is off to him. Haters gonna hate.
posted by asavage at 12:34 PM on March 22, 2012 [8 favorites]


For those who are still baffled: The reason the first video astonishes people who know about circuitry is because DC batteries normally cause LEDs to catch on fire. When the circuit builder operates the switches, the LEDs work normally, as if they were connected to an AC power source (alternating current, also called a "Marconi-Edison" source in honor of its inventors).

The people who suggested that the uploader had used After Effects to create the video clearly hadn't watched it. They got things backwards and thought he connected the LEDs to an AC power source, and later "caught them on fire" with a post-production trick.

The reason the circuit works is because of a trick component that has been built into the 9V battery connector. It's a miniature microwave emitter, which excites water molecules in the base of the LEDs, which he presumably hydrated with a good soak in saline solution (same stuff used with contact lenses). The excited water molecules transmit current at a 60 Hz AC frequency, causing the LEDs to work normally. This is why all of the analogies above compare current to water.

The debate over soldering technique is in my opinion entirely due to the circuit builder not using enough solder. Good technique means applying as much solder as possible while (and this is very important) avoiding the Farad ceiling for any individual component. Ideally, it should be a uniform mass of solvent connecting all parts of the circuit, but in a pinch you can isolate blobs here and there. You want enough solder to drown a bug, hence the term "deadbug."
posted by compartment at 12:53 PM on March 22, 2012 [9 favorites]


I'm just foolin, yo.
posted by compartment at 12:53 PM on March 22, 2012


vk2zay, the first guy to figure it out, has quite a nice youtube channel.

In this video he explains how he solved the mystery, and he explains how to build a version using large through board components.

While many smart people were arguing that this was done in aftereffects, vk2zay notices a glitch that lasts less than a second. At 2:05 in Fredzislaw100's video Fredzislaw100 touches the bare wire close to the LED with the inductor, and the LED dimly turns on for an instant.

I don't know if Fredzislaw100 intentionally left that glitch in as a hint, or if it was an error, but if it was not for the little glitch no one would have found the solution.

Pretty good detective work by vk2zay.

And his technique? I have a friend who is a pretty well respect retina surgeon. I have seen videos of his surgeries where he reattaches retinas using a microscope and tiny little cutting, grabbing and cauterizing tools. Fredzislaw100 technique does not look very different from that.
posted by Ayn Rand and God at 12:55 PM on March 22, 2012 [1 favorite]


As someone who has only soldered something once in his life and did a terrible job of it, is there maybe some sort of disagreement-in-terms here where some folks are saying "he soldered really well at such a small scale, amazing" and other folks are saying "his solder points are messy/he shouldn't use the gun to hold the melted solder, sux" and y'all are just talking past each other, or is there something more interesting or complicated in terms of competing soldering dogmas going on? Help a layman out.
I think some people just haven't seen the second video. Seriously, He is using a microscope to solder half millimeter surface mount components – like you find in a cell phone, together using wires rather then on a printed circuitboard.
(joking about the radio receiver bit; I know it's not technically possible)
Should be possible if the frequencies are low enough. software defined radio
The solution is the equivalent of not mentioning the elaborate system of underground pipes and switches that were installed in the dams and water wheels that you can't see. Using filters and
Not really -- all 'water' flows through what you can see. What you don't see is the hidden information in the stream.

It's more like, lets say you had a hose with sprinklers and faucets. The trick is, the faucets wouldn't really be faucets, but rather active pumps that turn on and off the flow of water every n seconds. The sprinklers would each have a clock in them such that they are on only for a 1 second every n seconds. The rest of the time, even if water is flowing, they're not.

So as long as the times for two sprinklers don't overlap only one is triggered by each faucet.

(i.e. if one sprinkler was 10 seconds, and the other 11 seconds, 110 seconds in, they would overlap – so the electronics version is a bit more complicated. You could simulate it by imagining a vessel that needs to pressurize by receiving multiple signals over a short period of time for the sprinkler to come on)

Now, obviously you'd be able to see this happening, and in the sprinkler example, it would take minutes for the sprinklers to cycle through all the possibilities.

But with electricity, it happens much faster. It could probably work more quickly with water if you had a higher pressure then a garden hose, as well.
posted by delmoi at 12:56 PM on March 22, 2012 [1 favorite]


I rewatched the first video. Are the naysayers only talking about that one? Because he is pretty clearly hamming it up there. It is in the solution video that has the holy shit stuff in it.

And if you are complaining about his technique on that one, then please post videos of you making that look bad so everybody can switch their hero worship.
posted by Threeway Handshake at 1:00 PM on March 22, 2012


Fantastic! I know just enough to be blown away by the first video. I studied the schematic in the second video for quite some time before saying "Different frequencies?" with the same cocked-head questioning that dog say "rowwr?".

Reminds me of a line by Teller: "Make the secret a lot more trouble than the trick seems worth"
posted by benito.strauss at 2:01 PM on March 22, 2012


Here's a free online circuit-simulator. Draw a schematic, turn it on, watch it work.

https://www.circuitlab.com/
posted by exphysicist345 at 2:15 PM on March 22, 2012 [1 favorite]


[C'mon folks "Help maintain a healthy, respectful discussion...."]
posted by jessamyn at 2:40 PM on March 22, 2012


When I saw this (& the solution) the other day, my first thought was "I wonder if he's a telco tech?".

Because this is very (i.e. almost exactly, but with surface mount components & LEDs instead of full-sized components & relays) similar to a circuit they made us apprentices build at the training school 30 years ago, as a simple demonstration of MFC generation & detection. And that was likely cribbed from old BPO or Ericsson training documentation or app notes. IIRC, the detection circuit we built was used essentially unchanged in the small(ish*) MFC decoder boxes we used for testing.

That said, I didn't twig until I saw the solution video…

('Small', as in 'smaller that a trolley AET'. They were ~15"x10"x6", with a hooked bracket to hang them on a rack & cords to plug into exchange power & the junction under test. They also contained a decade (hex?) decoder/counter chain & allocater, to drive mechanical "digital" displays so you could read off the signalling codes & digits. Really handy for when the exchange installers modified & re-commissioned routes before the signalling software / strapping was done & tested…)
posted by Pinback at 3:21 PM on March 22, 2012


Okay then, directed in a general fashion to anyone dismissing this out of hand because it's not advancing the state of the art in EE: you're missing the point entirely of why people are amused. I don't doubt anyone with anything beyond basic electronics knowledge could create a circuit that operated LEDs in that fashion, but creating it in such a way as to be hidden in that way and then creating the presentation is actually unique and interesting, and it both makes people think about electronics more than they usually would and demonstrates that video "proof" of something is rarely that.

And I'm offended to be told I'm acting like a child to have thought so.
posted by flaterik at 3:31 PM on March 22, 2012


I think a lot of the nay-sayers are reacting to the fact that this video isn't clever the same way cool electronics demos are usually clever. It's sort of like seeing a video of a really amazing card trick, then finding out it was made with video editing software instead of actual sleight of hand. It's nifty, I guess, but it's a different sort of nifty than what was expected.
posted by ryanrs at 3:32 PM on March 22, 2012


Well, I think the miniaturization is still pretty impression from a by-hand fabrication standpoint, but I think it's unfair to say it lacks all niftiness beyond that just because of the fact that it's not super fancy from a circuit standpoint.

I still think its a valuable demonstration for quite a large number of people, especially those that have a hobbiest level interest but only an entry level education on the subject - the population of which is growing rapidly.

and I also apologize for the tone of my justifiably deleted comment, both to idiopath and the mods
posted by flaterik at 3:40 PM on March 22, 2012


Idiopath: Not to nitpick but, 1 * 1 = 0!
posted by wobh at 4:04 PM on March 22, 2012 [3 favorites]


I think a lot of the nay-sayers are reacting to the fact that this video isn't clever the same way cool electronics demos are usually clever.

Oh it's clever all right. It's just stupid clever instead of smart clever. The instant he flipped the first switch, my reaction was "you don't really expect me to believe those are on-off switches, do you? There is obviously current running through them when they're supposed to be off."
posted by charlie don't surf at 4:10 PM on March 22, 2012


I dismissed all thoughts along the lines of the solution due to the schematic, so I feel a bit cheated. For those who'd call me a nay-sayer and a hater, if he had Photoshopped in the light, would you not feel the same way? It's a similar level of misrepresentation - unless the circuit is explicitly DC, this is trivial.
posted by springload at 5:29 PM on March 22, 2012 [1 favorite]


I have hand-soldered surface-mount chips (but only the largest ones) so as I watched him soldering in the first video I was sneering at his technique. Then I noticed that little flash from the LED when the circuit wasn't even completed -- aha! get ready, I thought, there's more than meets the eye here, maybe super-low currents, maybe RF. After the switches were demonstrated -- yikes! I was impressed.

Then I saw the how-to video and laughed out loud. I take back everything I ever thought about his soldering. I'm not impressed, I'm bowled over.
posted by phliar at 5:41 PM on March 22, 2012


For the many criticizing his "soldering" technique. He WASN'T soldering. He was "tacking".

Temporarily attaching two wires/components for a quick test/demonstration.
posted by notreally at 5:45 PM on March 22, 2012


I dismissed all thoughts along the lines of the solution due to the schematic, so I feel a bit cheated. For those who'd call me a nay-sayer and a hater, if he had Photoshopped in the light, would you not feel the same way? It's a similar level of misrepresentation - unless the circuit is explicitly DC, this is trivial.
Well, but here's the difference. If he had been sitting at a table with you, soldering that stuff in front of you, you would have seen the same thing. this was a 'real world' magic trick. Maybe if you'd have seen it in person, you would have been able to see the micro components he used to build the RF stuff.

I also thought the device itself was pretty impressive. Maybe the 'magic' wouldn't impress a real EE but I had no clue what was going on.
posted by delmoi at 6:14 PM on March 22, 2012


I clicked in expecting this to involve an anti-gravity field, a plane, and a treadmill at some point.
posted by BrotherCaine at 6:18 PM on March 22, 2012


I have to say, as a guy who soldered ~9 hours a day from 1993 to 1998, it's about damn time we have a soldering technique debate on mefi.
posted by glycolized at 6:46 PM on March 22, 2012 [7 favorites]


wobh: "Idiopath: Not to nitpick but, 1 * 1 = 0!"

This is not the first time, and surely not the last time, I will regret an unneccisary exclamation point!!
posted by idiopath at 7:02 PM on March 22, 2012


My previous explanation of this circuit was a little convoluted so I'll break it down, if that helps.

To analyze the circuit shown at the 24-second mark in the explanation video, you use two principles:

1. There are two signal sources, 300 KHz and 5000 KHz. The superposition theorem means that you can treat this as two separate circuits -- one operating at 300 KHz and another operating at 5000 KHz. This allows you to redraw the original circuit as two separate simpler circuits.

2. Use the simplifying assumption that the AC filters are perfect -- that they are either open circuits or short circuits at the two frequencies used. This is a reasonable assumption for our purposes for lighting LEDs since any signal lower than about 2 volts leaking through will do nothing.

Start with the simplest case, 5000 KHz AC. At this frequency the capacitors are selected to be short circuits and the inductors are selected to be open circuits. Now you can vastly simplify the original circuit. Since the capacitors are shorts, you can replace them with wires that completely bypass Switch 1 and 3, and LED 1 and 3 and all diodes. So take all of those elements out of the circuit. The inductors are open circuits at 5000 KHz so you can remove all of them, too. The only thing you have left in the circuit are Switch 2 and LED 2. That is a trivial case of a flashlight and a switch. You turn LED 2 on and off with Switch 2. Everything else in the circuit is gone. LED 2 flashes at 5000 KHz with a 50% duty cycle.

Now take the case of the 300 KHz AC signal. In this case, the capacitors are open circuits and the inductors are short circuits, just the opposite of the 5000 KHz case. Since the capacitors are opens, just remove them and throw them away. Since the inductors are shorts, you can replace them with wires which bypass Switch 2 and LED 2. So remove Switch 2 and LED 2 from the circuit. You are left with Switch 1 and 3 and their diodes and LEDs 1 and 3 with their diodes. When both switches are open, no current can flow either direction because of the opposite facing diodes on the switches. When switch 1 closes, current can flow clockwise, lighting up LED 1 and bypassing LED 3. When switch 3 closes, current can flow counterclockwise, lighting up LED 3 and bypassing LED 1. When both switches are closed, current can flow both directions, alternately lighting up both LEDs as the direction of the AC current changes direction at 300 KHz.

Put it all together using superposition, and you have three switches controlling three separate LEDs.
posted by JackFlash at 7:40 PM on March 22, 2012 [3 favorites]


I'm a techie, and that's really clever and slick. I had a good idea what was going to happen about halfway through and I still laughed in surprise when he did flip the switches.

When I saw the solution I thought, "It had to be something like that," but could I have "solved" it? No - and never could I have designed it either (but then, that's a hardware problem. :-D)

Full marks!
posted by lupus_yonderboy at 7:42 PM on March 22, 2012


Yeah, the limits of my circuitry knowledge had me thinking it was some kind of charlieplexing, but I couldn't figure out what the switches were doing. Neat trick!
posted by Kyol at 7:51 PM on March 22, 2012


I think everyone who is saying how simple this is should pay attention to how many paragraphs it's taking to explain it.
posted by flaterik at 9:09 PM on March 22, 2012 [1 favorite]


I think everyone who is saying how simple this is should pay attention to how many paragraphs it's taking to explain it.

And I didn't even get into how you select the capacitors and inductors to form the simultaneous high and low pass filters at two distinct frequencies or why you select those two frequencies.

Or why he used Schmitt trigger inverters as oscillators and how you determine their oscillation frequencies which depend on the high going and low going Schmitt trigger voltage thresholds combined with RC discharge curves.

Or how he connected the 300 KHz output in bridge mode to provide symmetrical push-pull currents for LEDs 1 and 3 so that they are the same brightness.

Or how he used AC coupling to superimpose the 5000 KHz signal on the 300 KHz signal.

Or the resistors required so that all LEDs are of similar brightness.
posted by JackFlash at 9:41 PM on March 22, 2012 [4 favorites]


JackFlash: Could you get into all that? Your previous explanation was really good, and now I want more.
posted by Ayn Rand and God at 11:19 AM on March 23, 2012


I think everyone who is saying how simple this is should pay attention to how many paragraphs it's taking to explain it.

JackFlash is giving the details of the execution and getting that right is not trivial - it's definitely well done! It doesn't take many paragraphs to explain the mystery though: "They're frequency-multiplexed" would suffice.

For me and I guess other electronics people who frown, the thinking went "Well that's strange. If it was AC, then you could... well but it's not, we have the schematic. Is he shunting it with his hands? No. But could the table be conducting? Do different parts of the circuit tough the table at different times? Maybe, but not really..." And at about that time we saw the thumbnail for the revelation video in YouTube's sidebar and realized that circuit wasn't what he had drawn. Then we felt cheated of the opportunity to see through it on our own and feel smart and brag in the thread.
posted by springload at 2:20 PM on March 23, 2012 [1 favorite]


A series circuit of 1 battery, 1 resistor, 3 switches, and 3 LEDs. What happens when you turn on each switch?
All three LEDs go on when all the switches are on.
A unspecified circuit of 1 battery, 1 resistor, 3 switches, 3 LEDs, an AC source, some inductors, and some capacitors...
Well, something else happens. It's a really cool demo, but the post should allude to the trickery...

</get_off_my_lawn>
posted by lalas at 3:04 PM on March 23, 2012


And if it was a video that said "here's some LEDs controlled via frequency multiplexing", none of us would have seen it, and metafilter would have been denied its soldering conversation.
posted by flaterik at 8:24 PM on March 23, 2012 [1 favorite]


« Older Since 2010, over 3,000 children throughout norther...  |  "The result feels like musicia... Newer »


This thread has been archived and is closed to new comments