YT YouTube2Text

YouTube2Text → Transcripts → The Insane Engineering of Fiber Optics

The Insane Engineering of Fiber Optics — Transcript

by Veritasium · 6,412 words · 947 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00- In 1887, a young physics teacher attached
0:02a thin glass rod to the end of an arrow
0:05and loaded it into his homemade crossbow.
0:07He torched the rod's end until a molten glass bead.
0:10Roughly the size of a pin's head,
0:12it was barely hanging on.
0:13And then he fired the arrow.
0:15(air whooshing)
0:18(suspenseful music)
0:23Bright sunlight streaming into the room
0:25glinted off a tiny glass fiber.
0:28At its thinnest, it was less than a tenth
0:30of the diameter of a human hair.
0:32Now legally we cannot shoot a crossbow indoors.
0:36Luckily I am pretty quick.
0:37(Henry chuckles)
0:39(suspenseful music)
0:46That is cool.
0:47Look at that.
0:48Look how thin we got it.
0:49The physics teacher was using these fibers
0:51to measure very small forces,
0:53but in the decades that followed,
0:54other researchers found a different use for them:
0:57sending light signals.
0:59(quiet dramatic music)
0:59Glass fibers can carry light even through bends.
1:02Look at it. It's like that's 90 degrees.
1:04And during the 1960s,
1:06the telecoms industry was desperate
1:07for a better way to carry more information.
1:10As one American engineer said,
1:11"It should be possible to transmit
1:14"all of the conversations going on
1:16"anywhere in the world simultaneously
1:18"over a single thread of light."
1:21But these glass fibers were terrible.
1:24The light faded so much that you couldn't decode a signal
1:26after just one kilometer.
1:28No one thought they'd be practical.
1:30Except for one man.
1:32(dramatic theatrical music)
1:33His own boss called his dream "a wild goose chase,"
1:36but if he was right,
1:37if he could prove that fiber optics were in fact viable,
1:40and if someone could then figure out how to make them
1:42at scale, well, eventually they could connect continents.
1:45How much is in this warehouse?
1:47- Oh, there's millions of kilometers here
1:48that we're pushing through here every month.
1:50- It's approximately three times as hot
1:53as the surface of the sun.
1:54- That's crazy.
1:55And one day,
1:56it could build the internet.
1:59(dramatic music fading)
2:01All that was unimaginable in the 1960s,
2:04but even then telecom companies wanted
2:06a better communication system,
2:08just for phone calls.
2:10Between 1960 and 1973,
2:12the number of daily calls doubled in the U.S.
2:14from 285 million to 576 million.
2:18In that same time period,
2:19NASA sent people to the moon
2:21for a total cost of $26 billion.
2:24While AT&T had to spend nearly triple that amount
2:27just to install new telephone infrastructure.
2:29They just couldn't keep up with the demand.
2:32At the time, phone calls were mostly
2:34transmitted over copper wires.
2:36Now imagine I'm trying to phone my buddy Gregor.
2:38Now, when I pick up, a constant current
2:40immediately flows through the wire
2:41to supply power to the phone.
2:43When I speak, my voice vibrates
2:45a tiny diaphragm in the mouthpiece
2:46that modulates the current.
2:48The louder I am, the bigger the current swing,
2:50and my voice's pitch changes
2:52how fast the current oscillates.
2:54At Gregor's end, that current reaches
2:55a speaker that turns it back into sound.
2:58Now, this works for one call on one line,
3:00but phone companies aren't going to install wires
3:02between every single house.
3:04One cable needs to carry multiple calls.
3:06So phone companies figure out a neat little trick.
3:09Even though humans can hear frequencies up to 20 kilohertz,
3:12you really only need to hear the ones between
3:15300 and 3400 hertz
3:16to understand what someone's saying.
3:18So phone companies drop frequencies above four kilohertz
3:21or anything outside the voice band.
3:23But if everyone used that zero to four kilohertz band
3:26on the same line, well, then all the conversations
3:29would be on top of each other
3:30and you wouldn't be able to hear
3:31the person that you're talking to.
3:32Luckily, we can shift those electrical signals,
3:35so my conversation can use zero to four kilohertz
3:38while another person's can use four to eight
3:40and so on all the way down the same copper wire.
3:43At the other end, a filter separates out
3:45the different frequency bands,
3:46similar to how a radio can tune into one station.
3:49Then the signal is shifted back down
3:51to the zero to four kilohertz range
3:52before it hits the speaker,
3:54so it sounds like a normal voice.
3:56But as phone companies started
3:57using higher and higher frequencies,
3:59they started running into problems.
4:01That oscillating current carrying the signal
4:03generates additional fields,
4:04a magnetic field shown here in red,
4:06and little loops of current called eddy currents,
4:09which are shown here in blue.
4:10Now, because of their shape,
4:11those loops resist changes
4:13to the current at the center of the wire,
4:15and they reinforce changes along the surface,
4:17but eddy currents get stronger
4:19the faster the signal changes.
4:21And if you want more calls,
4:22you need more bands,
4:23which use higher and higher frequencies.
4:25So soon the eddy currents are strong enough
4:27to cancel out the oscillation in the center of the wire.
4:30But if the signal can only travel along the wire's surface,
4:33it's limited to a smaller cross section.
4:35So resistance goes up and higher frequency signals
4:37can't travel very far before they're absorbed.
4:40For a while, telephone companies were able to work around
4:43the high loss by boosting signals every few kilometers.
4:46But past 65 megahertz, well, the high frequencies died off
4:49in less than two kilometers,
4:51so they were too impractical to use.
4:53And this trade-off between signal power
4:55and data rate was actually discovered back in 1948.
4:59The Shannon-Hartley theorem described the maximum capacity
5:02of a communication channel based on
5:04the signal power, noise, and bandwidth,
5:06or the difference between the highest
5:07and lowest frequencies.
5:09The best way to increase capacity
5:10is to increase the bandwidth.
5:13But they'd already reached the limit for copper.
5:15The telecoms industry didn't have another alternative,
5:18(dramatic music)
5:19but then they saw the light.
5:20Visible light has frequencies in the hundreds of terahertz.
5:23As one researcher for Bell Labs put it,
5:25"the possibility of extremely high frequencies
5:28"and accordingly vastly wider bands of frequencies
5:31"appeared as a fabulous El Dorado,
5:34"always beckoning us onward."
5:37But if you're going to use light,
5:38there are two problems.
5:39First, there wasn't a strong enough
5:41light source to send a signal.
5:43A light bulb couldn't produce enough power.
5:45There just weren't any options.
5:47- And along in 1960,
5:48Ted Maiman invented the laser.
5:50- [Reporter] A science fiction weapon that melts steel,
5:53a death ray that could be more deadly
5:56than a nuclear bomb.
5:57It's a beam of light.
5:59This is called a laser.
6:02- [Henry] There's a running joke that the laser was
6:04"a solution looking for a problem,"
6:06but that's not really true.
6:08- [Donald] It rejuvenated the interest
6:10in possible optical communications.
6:13- As one Bell Labs engineer said,
6:15"Think of all this bandwidth!"
6:18But they couldn't figure out the second problem.
6:20Glass is not as clear as it looks.
6:23Only a small amount of light actually makes it through.
6:26In fact, only a year after the laser's invention,
6:28a Bell Labs scientist started looking into fiber optics.
6:32His boss told him to check its loss,
6:34or how much a light signal weakens
6:35as it travels through the glass.
6:38Now, signal loss is measured with decibels,
6:40similar to how you measure audio volume.
6:42Both are on a logarithmic scale,
6:44but while volume uses sound pressure,
6:46loss uses signal power.
6:48His boss said it would only be worthwhile
6:50if the loss was less than 20 decibels a kilometer,
6:53so it was comparable to copper.
6:54But when he checked the loss for the best fibers,
6:57a thousand decibels a kilometer.
7:01To understand just how bad that was,
7:02imagine a leaky truck carrying sand
7:05over one kilometer.
7:06If the truck starts with 100 kilograms,
7:08and it ends up with one,
7:10well, that represents a loss of 20 decibels per kilometer.
7:13That's only 1% of the signal left.
7:15It's what was happening with copper,
7:17bad but still manageable.
7:20A loss of a thousand decibels a kilometer
7:22is like starting with the mass of the observable universe
7:25and ending up with less than the mass of a single atom.
7:29To be practical, these fibers would need to improve
7:31by a factor of 10 the 98th.
7:33It sounded impossible. It would have to be one of the most
7:36absurd engineering improvements in history.
7:39So Bell Labs gave up.
7:40And fiber optics might have died right there
7:43if it wasn't for one man.
7:45Charles Kao.
7:47In 1960, he started working
7:48for an English telecoms lab
7:50that was looking for a way to communicate
7:51with higher frequencies,
7:53but not with fiber optics.
7:55Kao's boss said that it was "a wild goose chase"
7:57and the rest of the lab agreed.
8:00But Kao kept coming back to it,
8:01asking, "Why is glass so lossy
8:04"and can we remove those factors
8:06"or are they intrinsic to the material?"
8:08If it was the latter,
8:09then fiber would be dead in the water.
8:11But if not, then maybe it had a chance.
8:15So the question was:
8:16Why was the loss so high?
8:18Well, you might think it's because light
8:20is exiting through the sides of this fiber,
8:22but glass is really good at trapping light,
8:24even through bends.
8:26Here's how.
8:27I've got a bucket of propylene glycol here
8:29and a green laser.
8:30When I remove the tape,
8:32(gentle dramatic music)
8:36the laser bounces one, two, three times.
8:39And there are two things working together here
8:41that actually trap the light.
8:43See, light travels slower in some materials than others,
8:46and it travels as a wave shown here as a line of crests
8:49with the beam always pointing perpendicular to that line.
8:52When the wave meets a boundary at an angle,
8:54one end of the crest crosses into the new material first,
8:58and it changes speed before the rest.
9:00If light travels slower in that new material,
9:02then that end drags behind and the beam bends
9:05or refracts away from the boundary.
9:08How much it bends depends on the angle it hits the boundary
9:11and on how much each material slows light down,
9:13which is described by their index of refraction.
9:16This relationship is defined by Snell's Law.
9:19Now if light is moving from a material
9:21with a higher index of refraction
9:23into one with a lower index,
9:25like propylene glycol to air,
9:27well then one part of the wave crest speeds up
9:29and the beam bends toward the boundary.
9:31But as you increase the angle of the light,
9:33the refracted beam gets closer and closer to the boundary
9:37until it's running right along it.
9:39That's the critical angle.
9:41A little bit more and all the light
9:43reflects back into the glycol.
9:45That's total internal reflection,
9:47and you get the exact critical angle
9:48by setting the angle of refraction to 90 degrees
9:51and then rearranging Snell's Law.
9:53- The key of it is the bigger the difference
9:56in index of refraction,
9:58the more powerful this effect is.
10:00If I have a really subtle difference between these two,
10:03then the amount of bending that takes place will be smaller.
10:08- You can do the same demo with water,
10:10but propylene glycol works a bit better
10:12because it has a higher refractive index,
10:141.43 compared with 1.33.
10:16(liquid splashing) See, once the liquid is low enough
10:18for the angle of the stream to be much sharper,
10:20the light breaks through the water stream,
10:22but it stays trapped in the propylene glycol.
10:25And if you use something with
10:26an even higher refractive index,
10:28like the glass types Kao was investigating,
10:30well, then the critical angle decreases even more,
10:32which traps the light signals through even sharper turns.
10:36Okay, but the light has to move from
10:38a high refractive index to a low one.
10:40In air that works right up
10:42until something touches the glass.
10:44If you disturb that boundary,
10:45well, light can leak out the side.
10:48So about a decade before Kao,
10:50researchers realized that a fiber
10:51needed to have two types of glass.
10:53The middle, the core, has the higher refractive index,
10:56and around it goes a second glass, the cladding,
10:59with a lower refractive index.
11:01Together, they keep the light trapped.
11:03Now the fiber could be touched, coated in plastic,
11:06buried in a cable, bent around things,
11:08and the light would still travel down the core
11:10and come out the other end.
11:12(quiet dramatic music)
11:13So Kao knew that if the high loss
11:14wasn't from light leaking out the sides,
11:16well, it had to be from something else.
11:18He spoke with material scientists and glass experts,
11:20but no one seemed to know what was really causing it.
11:23He kept looking and eventually he landed
11:25on two potential sources:
11:27scattering and absorption.
11:29If you shine light through
11:30a perfect crystal that's transparent,
11:32there's going to be a little bit of reflection,
11:34but almost all of the electromagnetic energy
11:36will keep going down in the same direction that it started.
11:39But glass isn't a crystalline solid.
11:41The basic building block is silica or silicon dioxide,
11:45one silicon atom connected to four oxygen atoms
11:48in a tetrahedra or pyramid shape.
11:50In the glass, the pyramids are connected at the corners,
11:53so every oxygen is shared with another silicon.
11:56That's why it's called silicon dioxide.
11:58For every two oxygens,
11:59there's one silicon atom.
12:01This is an amorphous solid.
12:03Unlike a crystal, its atoms have no long-range order.
12:05Instead, its disordered liquid-like structure
12:08locks in as it cools,
12:10with the exact arrangement
12:11depending on the glass type and cooling speed.
12:14For example, if you quench the glass really fast,
12:17the structure locks at a higher temperature.
12:19The final solid has lots of dense regions,
12:21which re-radiate more light than the sparse ones.
12:24The light scatters, which weakens the signal.
12:26This is called Rayleigh scattering,
12:28and you can't get rid of it completely,
12:30but if you cool the glass slowly,
12:32those pyramids have more time to settle
12:34and pack together evenly.
12:35The structure locks in at a lower temperature
12:38and light scatters less.
12:40But Kao needed to know the minimum scattering.
12:43Was it 10 decibels a kilometer,
12:45a hundred, a thousand?
12:46He needed something concrete.
12:49Finally, he stumbled on a paper by Robert Maurer,
12:51a scientist from Corning Glass
12:53who measured the Rayleigh scattering
12:54across several glass types and cooling rates.
12:58Kao used his calculations to estimate
13:00that if the glass's structure was locked
13:02at a thousand degrees Celsius,
13:04well, its scattering loss would
13:05be around one decibel a kilometer.
13:08He was surprised.
13:09He expected scattering loss
13:10to increase with higher temperatures.
13:12That's what causes more density fluctuations,
13:15but a thousand degrees Celsius was pretty hot
13:18compared to other glass types,
13:20so then scattering couldn't be behind
13:22those crazy high losses.
13:23Absorption had to be the problem,
13:25but Kao didn't know what exactly
13:27was absorbing so much light.
13:29Was it the glass itself or something else?
13:32Well, absorption works like this.
13:34In a single atom, electrons sit at discrete energy levels,
13:37and there are usually some empty spaces in the upper ones,
13:40but pack atoms together into a solid
13:42and those levels smear up
13:43into wide bands separated by gaps.
13:46Now when light hits the surface,
13:47it can excite an electron to a higher band,
13:50but only if the photon has at least
13:51as much energy as this band gap.
13:54If it does, the photon is absorbed
13:56and eventually it releases that energy as heat.
13:58In something like silicon,
13:59that band gap is very small.
14:01It's only around 1.12 electron volts.
14:04Visible light is between 1.63 to 3.26.
14:07It's got plenty of energy,
14:09so it gets absorbed.
14:10That's why pure silicon looks opaque to us,
14:13but the silicon in glass is connected to oxygen atoms
14:15that pull more on the shared electrons.
14:18Now it takes a lot more energy to excite an electron.
14:21The band gap is much larger, around nine electron volts.
14:24Visible light doesn't have enough energy,
14:26so it passes right through,
14:28and that's why glass looked clear to us.
14:30But most glass types aren't pure silica.
14:33The raw materials for glass,
14:35from sand, soda, lime, and other ingredients,
14:38well, they contain impurities like trace metals,
14:40and these metals absorbed far more light than silica did.
14:44One of the worst impurities in glass is iron.
14:47Inside the glass structure,
14:48iron atoms form positive ions
14:50that strongly absorb violet, bluish,
14:52and infrared wavelengths.
14:54Of the visible light spectrum,
14:56impure glass doesn't absorb green light as much,
14:59and you can even see this in a regular glass pane.
15:02Looking at it from the edge,
15:03light travels down the entire width of the glass,
15:06so by the time it reaches your eye,
15:07only green light is left.
15:09So if you can see this attenuation
15:11after just a few centimeters,
15:13imagine a glass fiber trying to carry light
15:15for kilometers and kilometers.
15:17No wonder it had a crazy loss.
15:20Kao spoke with more experts
15:21who estimated that without those metal ions,
15:23the loss would drop to around 20 decibels a kilometer,
15:26comparable to copper,
15:27just by removing impurities.
15:29(quiet dramatic music)
15:30It was all theoretical at this point,
15:31but it was enough for Kao.
15:33In 1966, his team published a paper on their theories,
15:36concluding with "the crucial material problem appears
15:39"to be one which is difficult but not impossible."
15:42But nobody really paid attention to his paper
15:45because it really seemed impossible.
15:48A 10 to the 98th times improvement was unprecedented,
15:52and Kao didn't actually measure the loss in glass.
15:55He didn't have anything concrete,
15:57so he needed proof.
16:00Kao was trying to invent a new way to communicate,
16:02but even today actually staying connected can be hard.
16:05In fact, we just traveled to Spain
16:06for the eclipse a few weeks ago,
16:08but with so many other projects going on,
16:10we still needed to check in
16:11and send updates to the rest of the team,
16:13not to mention my mom. (laughs)
16:15We drove seven hours to Burgos,
16:17but the Wi-Fi when we got there was a little touch and go.
16:20We needed to have everything set before we arrived on site,
16:23so of course we used Saily, today's sponsor.
16:26After I downloaded their app and activated my eSIM,
16:28all I had to do was choose my data plan and country,
16:31and then I was set before I'd even reached the airport.
16:33It's available in over 200 destinations,
16:35and they even offer regional plans,
16:37which is great if I have an extended shoot
16:39in multiple countries.
16:40I'm automatically connected right when I land,
16:43ready to update my team, access maps,
16:45or get transportation to wherever I'm filming.
16:47I installed an eSIM once,
16:48and I use it almost anywhere in the world.
16:50There's no hidden charges,
16:51so it's extremely affordable and convenient.
16:54Actually, let me show you.
16:55(buzzer ringing)
16:56Wait, I can't get SMS messages right now.
16:59But with Saily you can now get a U.S. phone number
17:02right through the app for voice calls
17:03or SMS messages as an optional add-on.
17:06There's no contract or physical SIM card needed.
17:08So if you've got a trip coming up,
17:10and you still don't know where you're going to get your data
17:12well, you can scan this QR code for the Saily app
17:14and check out with code "Veritasium"
17:16so you'll be able to get 15% off your first purchase.
17:18Again, just scan this QR code,
17:20use the code "Veritasium" and get 15% off,
17:23and you can stay connected.
17:25Now, let's go back to Kao.
17:26(striking dramatic music)
17:27Kao needed to prove that his theoretical
17:29low-loss glass was really possible,
17:31so he decided to measure the purest glass
17:33anyone had ever made, fused silica.
17:36Now, the reason glass manufacturers
17:37had added all those extra ingredients
17:39was to bring sand's melting point down
17:41to around 1400 degrees Celsius,
17:43but fused silica was pure silicon dioxide
17:46melted at over 2,000 degrees Celsius,
17:49so it was a lot harder to make,
17:51and that meant that Kao couldn't get his hands
17:53on much of it.
17:54Measuring loss normally takes
17:56hundreds of meters of cable,
17:57if not kilometers,
17:59but he only had pieces less than 30 centimeters long.
18:02To measure the loss in something that small,
18:04he and his colleagues invented
18:05a new, very precise instrument.
18:08Kao spent the next three years working on it,
18:10but when he finally measured a sample,
18:12he found that fused silica had zero loss.
18:16(dramatic music continues)
18:17That couldn't be right.
18:18In a panic, Kao ran his experiment again.
18:21It turned out that the loss was so low
18:23that it was within their margin of error,
18:25a mere plus or minus four decibels a kilometer.
18:29He spent months fine-tuning his equipment,
18:30and in 1969, he finally reported that fused silica
18:34had a loss of less than five decibels a kilometer.
18:38Now, that got everyone's attention.
18:40One scientist at Bell Labs said,
18:41(audience applauding) "Kao gave everyone a jolt.
18:43"That was the first practical measurement
18:44"which said, 'Hey, you're not just whistling Dixie.'"
18:48Now everyone was interested.
18:49Telecom labs around the planet
18:51rushed to fund fiber optics groups,
18:53but Kao had only worked with small samples.
18:55Nobody knew whether kilometers
18:57of low-loss fiber were even possible.
18:59The problem caught the attention of one Robert Maurer,
19:02Corning researcher and author of the scattering paper
19:04that Kao had referenced.
19:06To make a practical fiber,
19:07he saw three major problems to tackle.
19:10First, he needed to decide which glass to use.
19:12He could try purifying raw materials for normal glass,
19:15or he could try using fused silica,
19:17which Kao had already proven to be very pure.
19:19Most labs were avoiding fused silica
19:21because of how hard it was to work with.
19:24But luckily, Corning had a secret technique.
19:27Back in the 1930s, another Corning scientist
19:29had ran silicon tetrachloride gas through an oxygen flame.
19:33In the heat, water vapor from the flame
19:35would break silicon tetrachloride
19:37into hydrogen chloride gas and silicon dioxide.
19:40The latter would then start clicking together
19:42in a fine glassy silica powder.
19:44Then the powder would be pressed into shape
19:46and melted to make fused silica.
19:49Corning was one of the few companies
19:51in the world using this technique
19:52called flame hydrolysis to make ultra-pure glass.
19:56So they had the supply,
19:58but there was a bigger problem with fused silica.
20:01It has a refractive index of 1.46,
20:03which is the lowest of any glass,
20:05but for total internal reflection to work,
20:07the core needs to have a higher refractive index
20:10than the cladding.
20:11So if Maurer wanted to use a fused silica core,
20:14well then finding another glass
20:15with a lower refractive index for the cladding,
20:18well, that'd be a nightmare.
20:19It simply didn't exist.
20:21So everyone abandoned fused silica except for Maurer.
20:25He said, "I had no reason to think fused silica
20:27"was any better, but nobody else was working on it.
20:30"If you do something that's different
20:31"from what everybody else is doing,
20:33"you've got two advantages.
20:34"One is you may succeed where they fail, of course,
20:37"but even if you fail, you will gather information
20:39"that they don't gather."
20:41All the other labs were thinking
20:42of using a fused silica core.
20:44After all, that's where the light travels,
20:46but Maurer decided to use fused silica for the cladding.
20:50Then he could easily raise the refractive index of the core
20:52by adding some impurities or dopants.
20:55- The only glass that we had
20:58with a slightly higher index than fused silica
21:00was this titanium-doped glass.
21:03- [Henry] But that would mean adding titanium
21:05right where light needed to travel.
21:08Kao worked hard to prove that glass could be pure enough.
21:11All the other labs were trying to get rid of impurities,
21:13and Maurer wanted to add some back in.
21:17But he wasn't completely crazy.
21:19Like silicon, the titanium in glass
21:20can easily share electrons with four oxygen atoms,
21:23forming the same pyramid shape,
21:25so it wouldn't disrupt the overall glass structure too much.
21:28(gentle music)
21:30Now that he had the composition,
21:31Maurer needed to tackle that second problem.
21:33How do you make a tiny fiber
21:35with that perfect core and cladding cross section?
21:38To help, he hired two more scientists,
21:40Peter Schultz, a ceramics engineer,
21:41and Donald Keck, a physicist.
21:44- Each time you find a problem,
21:45you put a development team on it
21:47and strike it down and make it happen.
21:50- To understand their technique,
21:52I went to the Bay Area Glass Institute
21:53for some hands-on demos.
21:55Here, our instructor, Erin
21:57is taking a colored rod of glass
21:58and then dipping it in a bowl of clear molten glass.
22:01In fiber optics, this large cylinder is called a preform.
22:05(gentle music)
22:06We have two colors of glass here.
22:07We have a color core and a clear outside.
22:11If you want your inner rod to, say,
22:12be a quarter of the diameter of your outer one,
22:15you just set that in the preform,
22:16and then all the length along your pole,
22:17the cross-section's going to be identical.
22:19And this is pretty similar to how
22:21hard candies with patterns are made.
22:23This sugar is an amorphous solid, just like glass.
22:26So the process is the same, just at a lower temperature.
22:30But there is one problem with this method.
22:33Resting glass on a surface like this
22:34introduces a ton of stress and impurities,
22:37but when I try and suspend it in the air, well, it sags.
22:41- [Erin] But yeah, from all that movement,
22:42it did create stress,
22:43so now it's a curved piece of, of glass.
22:47- Thankfully, there's another option.
22:49Now, gravity is pulling our glass for us,
22:52and the surface section is keeping it together.
22:54(light dramatic music)
22:55So now we get a much more straight and consistent fiber.
23:00That's awesome.
23:01(dramatic music continues)
23:02- With their drawing technique sorted,
23:04the corning team could tackle the final problem.
23:06How do you actually make the preform?
23:08Well, instead of starting with
23:10a core and adding the cladding,
23:11Schultz and Keck decided to start
23:13from the cladding and work in.
23:15They used the flame hydrolysis technique
23:18to make titanium dope silica powder,
23:20or soot as they called it.
23:21Next, they took a pure fused silica tube
23:24and tried spreading soot down the inside.
23:26If they could get an even layer,
23:27well, then they could melt and collapse the tube
23:29into a solid rod, a perfect preform.
23:32But the soot collected at one end
23:34no matter how they sprayed it.
23:36- So one of the technicians
23:38grabbed a vacuum cleaner from the corner of the room,
23:41and he held it up to the backside of this tube.
23:44And it sucked a stream of titania dope silica glass
23:48down through the center of the tube,
23:50coated it just like we'd hoped it would.
23:53- [Henry] The vacuum did not survive that experiment,
23:55but they tinkered with a new setup
23:56until they got the gas flow just right.
23:59(striking dramatic music)
24:00All through the summer of 1969,
24:02their furnace burned "hotter than Hades."
24:04They drew kilometers of fiber from preform
24:06after preform, testing various compositions,
24:09but none of it was good enough.
24:11Titanium was breaking down in the high heat
24:13and absorbing too much light.
24:15They ran the finished fibers through
24:17a second heat treatment in an oxygen environment
24:19before they finally started seeing results.
24:23(light dramatic music)
24:24A year later, on July 22, 1970,
24:27a preform slipped against the furnace wall
24:29after Keck only got a few meters of fiber.
24:31The preform was ruined,
24:33but he decided to test the short fiber anyway
24:36after its heat treatment.
24:37- When I slid it across
24:38and the laser light hit the core of the fiber,
24:41all of a sudden I had another beam coming back in my face.
24:45The core was brightly lit,
24:47and it took a while to realize that the light
24:50was going all the way to the other end of the fiber
24:52and reflecting off the far end and coming back into my eye.
24:57- [Henry] The loss was 17 decibels a kilometer.
25:00- The only word I could come up with
25:02for this auspicious moment in history was "Whoopie".
25:06So that's what's in my lab notebook, "Whoopie".
25:11- [Henry] But the fiber was much shorter than usual,
25:13only 29 meters to their typical 200.
25:15He added a careful "must remeasure" in his notebook.
25:19Team quickly remade the preform,
25:20but it took almost a month
25:21to get 210 meters of finished fiber.
25:25It was finally done by late afternoon
25:27of Friday, August 21st.
25:28It was a gorgeous summer day
25:30and almost everyone left early,
25:31but Keck didn't want to wait.
25:33The fiber was so fragile,
25:34small temperature changes
25:35might break it over the weekend.
25:37- It would break if you looked at it crossways.
25:40- So alone in the lab,
25:41he anxiously measured the loss,
25:4416.9 decibels per kilometer.
25:46(dramatic music continues)
25:47- I said "QED, quod erat demonstrandum"
25:51There was no, (chuckles) no eureka.
25:55- [Henry] Maurer published their results that fall.
25:57As Schultz described it, no one believed it.
26:00And telephone companies all demanded
26:02to see the fibers firsthand,
26:04so Maurer hand carried a spool of fiber
26:06to all the big telephone laboratories.
26:09He would literally guard to make sure
26:10that they didn't snip pieces off and keep them.
26:13(dramatic music continues)
26:14For two years, they refined their methods.
26:17And to make the fiber stronger,
26:18they switched to a germanium doping.
26:21Like titanium it easily swaps with silicon
26:23while still raising the refractive index.
26:26But unlike titanium,
26:27germanium fibers didn't break down in the heat,
26:29so they could skip the second heat treatment.
26:31When they drew their first germanium fiber,
26:33the end lit up with the light from the furnace.
26:36Even when it reached the bottom
26:37and wound up around a large drum,
26:38the fiber stayed bright.
26:40- And we kept winding and winding and winding
26:43and thinking the light was going to go out,
26:45and it never went out.
26:47- The loss was only four decibels per kilometer
26:50with excellent strength.
26:52Keck considers that the very first
26:53truly practical low-loss fiber.
26:57By 1977, AT&T trialed a fiber system underneath Chicago.
27:02Two years later, a fiber team from Japan was able to get
27:04a loss of just 0.2 decibels a kilometer,
27:08the standard up to today.
27:10Fiber networks continued to grow
27:12with the first transatlantic fiber optic cable
27:14installed by 1988.
27:16And it transmitted those phone calls
27:18differently than in copper.
27:19The audio wave was split up
27:21or sampled into discrete values,
27:23and then converted into binary
27:24to get a long string of ones and zeros.
27:26Then a laser flashed on for one and off for zero,
27:29280 million times a second.
27:32These pulses of light traveled down the fiber,
27:34but that wasn't all.
27:36Multiple conversations could now be weaved together,
27:39one bit from one conversation,
27:40then one from another and so on,
27:42slotting it together like a zipper.
27:45As long as the receiver knew the timing of each conversation
27:48and the sampling rate, well, they could
27:49reconstruct the original waves.
27:53(dramatic music)
27:54So this first intercontinental cable
27:56could carry 40,000 simultaneous calls,
27:5910 times more than the previous copper version.
28:02This insane capacity opened up a whole new world,
28:05and with digital signals people realized
28:06that they weren't limited to phone calls.
28:08It was easier than ever to send texts, images, videos.
28:13All building up to the internet,
28:15and since then demand has only increased.
28:18- This reel right here?
28:20(light upbeat music)
28:21- Yeah.
28:23- 50.477 meter kilometers,
28:25so 50 kilometers.
28:27So it's about 30 miles.
28:29- [Henry] That's crazy.
28:30Like how much is in this warehouse?
28:32- There's millions of kilometers here
28:33that we're pushing through here every month.
28:36- [Henry] To see how modern fibers are made,
28:37we visited Prysmian's North Carolina factory.
28:40It all starts with the preform.
28:42Each pass of this microwave source
28:44ignites a mixture of gases into a plasma
28:46which deposits a thin layer of glass
28:48on the inside of the tube.
28:50- And you see when it turns purple,
28:51that's when we're making a plasma with that gas.
28:54- So, how thick are these deposition legs?
28:58Like how many times are you doing this?
28:59How thick are you doing-- - Thousands of passes.
29:01- [Henry] "Thousands of passes."
29:02- Several thousand times it's doing this.
29:04Every time we may go back and forth,
29:05we could change the composition of that if we wanted to.
29:08- [Henry] Oh, okay.
29:09After all the inner layers are done,
29:10the tube has a thicker wall,
29:12but there's still a gap down the middle.
29:14- So we want to collapse that channel upon itself.
29:16So what we do is we heat it up.
29:18We have a furnace that goes back and forth,
29:20and just via surface tension,
29:22it collapses upon itself.
29:23- Interesting, I would've thought
29:24maybe you'd like pull it, or just--
29:25- [Sujeet] And I asked this exact question.
29:27I said, "Are you pulling a vacuum in there
29:29"or what's making that thing shrink?"
29:30- Exactly. - It's really just
29:32a glass phenomenon where the molecules
29:34are trying to pull together.
29:36You heat it up, you'll lower the viscosity of the glass.
29:38It just... (exhales)
29:39- [Henry] Now they need to add more silica on the outside
29:41to get the preform to the right proportions.
29:44The core is very small,
29:45just 7.2% of the total diameter.
29:47- And you have a several thousand C
29:50plume of plasma right there.
29:51There's grains of silica that's flowing into that plume
29:55and effectively melting it
29:57and spraying it onto that rod.
29:58- [Henry] Wow. (steam hissing)
29:59- That plasma torch is approximately three times
30:03as hot as the surface of the sun.
30:05- Plasma always blows my mind.
30:06This is absurd.
30:08Once the preform's done,
30:09it goes to the top of the draw tower,
30:10seven stories up to start making fiber.
30:15(steam hissing)
30:16It really looks like some sort of steampunk thing, you know?
30:19All these coils and then this hot glass going down.
30:22(gentle dramatic music)
30:23The bottom of the preform is tapered to a point
30:25and heated to over 2,000 degrees Celsius.
30:28It's like there's some heat coming off this.
30:31The end of the taper slowly drops down to the floor below.
30:35- [Man] We call it a gob, but it looks like a teardrop.
30:39- The fiber is a little thick,
30:40so it'll be thinned down to the final
30:42125 microns before it passes to the next level.
30:45(gentle dramatic music)
30:48It's like a little flashlight coming down.
30:51(glass clinking)
30:57I didn't expect it to be so violent.
31:00(gentle dramatic music)
31:02If the drawing,
31:04like the pull down to that final diameter
31:08happens in such a short amount of...
31:10- In a few meters, let's call it.
31:11- A few meters.
31:12Then why do you need six more floors?
31:16- You got to cool the glass
31:17before you can apply the coating.
31:18And if you cool the glass too quickly,
31:21it'll have residual stress
31:22and then have optical loss attenuation.
31:25(glass clattering)
31:28- Once the fibers are cool enough,
31:29they are dipped in two protective plastic coatings.
31:32At the bottom of the draw tower,
31:33the fiber is wound onto massive spools,
31:36each capable of carrying hundreds of kilometers
31:38at over 50 meters a second.
31:40Okay.
31:42Next, the finished fiber is tested for loss and strength.
31:45Fiber is stronger than a steel wire of the same thickness.
31:48I guess if I try to do it like...
31:50- [Sujeet] If you just try to do it like that,
31:51it's really hard to break.
31:52- First of all, it just slides down my hand.
31:53- [Sujeet] And that's a glass fiber with two coatings on it.
31:58- Ah, it's just hurting my hand.
32:01Oh, my god.
32:02Yeah, not easy.
32:03Definitely not easy.
32:04Ow.
32:05But it's weak to sharp turns. - [Sujeet] Tie it in a knot.
32:08- Tie it in a knot. - And watch what happens
32:10and see how much you can get before it breaks.
32:14(gentle dramatic music)
32:16Yeah, because the angle's so great, you can break it.
32:19- But after testing, fiber is bundled
32:21and wrapped into thicker cables
32:22so that it's harder for them to knot and break.
32:25These cables will join the more than eight billion
32:27kilometers of fiber already installed
32:29that carry information around the world,
32:31powering the internet and letting you watch this video.
32:35But some experts argue that even with fiber,
32:37the internet would not be possible
32:38without another key invention.
32:41The amplifier, that's what boosts signals
32:43across long stretches,
32:44and it revolutionized fiber capacity on land and sea.
32:48So if you'd like to see a video about submarine cables
32:51and fiber amplifiers, well, let us know in the comments.
32:54(quiet dramatic music)
32:55As for Kao, his dream went further than he ever thought.
32:58When he received the Nobel Prize in 2009 for his work,
33:02he said, "The information revolution that started
33:04"in the 1990s could not have happened
33:06"without optical fibers.
33:07"It is beyond the dreams of the first serious concept
33:10"of optical communication in 1966,
33:12"when even one gigahertz was only a hope."
33:16For me, one of the things I kept on coming back to
33:18in this story is that Kao was not a material scientist.
33:22He was an electrical engineer.
33:24And it was only because he was an outsider
33:26that he was able to ask the questions
33:27that nobody was asking.
33:29He was able to push the status quo,
33:31but he was not the only one.
33:32Maurer's team were also working
33:34with a very unintuitive idea
33:36that by making something seemingly worse,
33:38they actually were making it better.
33:40As Kao said, "Who knows what other forward thinkers
33:42"will dream of in the future?"
33:44Granted, he was talking about bandwidth,
33:46but I think it applies anywhere
33:48where people are able to question
33:50why things are the way they are
33:51and whether that's really the limit.
33:54(electric chirping)
33:57One last thing.
33:57These videos are only possible
33:59because of the incredible work done
34:00by our team behind the scenes.
34:02We have world-class editors, animators, researchers,
34:05and writers that make these videos possible.
34:07We'd love to shine more of a light on their work,
34:09so we organized a Q&A
34:10with the writer of today's video, Sulli.
34:12She discusses the fascinating things
34:14she learned while researching this topic
34:15and things that didn't make the cut,
34:17but that are really interesting.
34:18So if you want to learn more about that,
34:20you can sign up for our newsletter in the description
34:22and you can find her Q&A right there.
34:24(light gentle music)

About this transcript

This page contains the full transcript of The Insane Engineering of Fiber Optics by Veritasium, generated from the public captions YouTube serves with the video. The transcript has 6,412 words across 947 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.