Fiber vs. Copper; What do we really need?

Fiber vs. Copper; What do we really need?


Remember that time I made a video about TOSLINK? I sure do, it was just last week! (almost) Well, before I ended that video, I left you with a question. Why haven’t we seen fiber… *THUNDER CLAP* [ominous music] You’ve angered the Hi-Fi gods! You said TOSLINK is impervious to electromagnetic
interference “but what does that matter in the digital space?” But you didn’t mention ground loops and
the devastation they cause! OK, OK, just, settle down! I’ll address that. Right now! When you connect two pieces of equipment with
an RCA cable, you also connect their respective grounds together. In some circumstances, this can create audible
humming or buzzing because even though the signal being sent is digital, the amplification
circuitry is not. Because it is connected to ground, if there’s
an imbalance between the two grounds of the sending and receiving equipment, that creates
a current in the ground of the signal wire which can affect the analog amplification
circuitry. Because TOSLINK is an optical signal, the
two devices are galvanically isolated and it doesn’t matter in the slightest if there’s
a ground imbalance. TOSLINK is, in effect, an opto-isolator and
this benefit should not be ignored. Now, I would like to provide an explanation
for my careless oversight. For every person who touts the benefits of
TOSLINK, there’s another that complains about clock jitter. Those individuals claim that if you’re encountering
a ground loop, “for the love of god don’t use TOSLINK with all its totally real imperfections,
sort out that problem you heathens!” Audio forums are worse than Twitter, each
of you likes what you like and anytime anyone presents any opinion regarding the pros and
cons of one system over another, y’all lose your minds and get real pretentious and I’m
just kinda over that. I, admittedly, chalked up the concern for
ground loops to be one of those opinions, but it’s not. TOSLINK can, does, and has many times presented
a meaningful solution to the problem of ground loops or other electrical noise issues, and
it may in fact be the entire reason Toshiba developed it. It is interesting to note, however, that the
ground loop problem seems to have gone away in the age of HDMI. Whether that’s down to better isolation
inside the equipment itself or some other thing, it suggests that the problem could
have been solved without going to an optical fiber. It may have just been easier (and, let’s
not forget, cooler) to use TOSLINK. Alright, now that that’s taken care of,
let’s get back to the question at hand. With TOSLINK being developed in 1983, why
haven’t we seen fiber… *THUNDER CLAP* You’ve angered the IT gods! You referred to CAT5 cable as Ethernet! You fool! Ethernet is a standard series of protocols
and fiber optic cabling is routinely used in larger networks! You’re right. I am but a lowly network plebeian, barely
capable of setting up a WiFi router. Forgive my callous equation of Ethernet and that cable. It won’t happen again. Wow, we’re almost three minutes into this
and we haven’t even gotten to the question at hand. So. Why haven’t we seen fiber optics in the
consumer space aside from TOSLINK? Actually, in some very niche circumstances
we have, and as the IT gods reminded us, it does exist in big networks, but I’m talking
on a broader, more basic scale. Your average consumer is probably familiar
with things like USB, HDMI, DVI, DisplayPort, Thunderbolt, Eth…, I mean, Cat 5 cabling
with 8P8C (often incorrectly referred to as RJ45) connectors, and maybe FireWire or SATA
or perhaps other things. All of these use wires, like some sort of
technologically regressive lazy person. Why not use light? Well, let’s start with the obvious thing. TOSLINK works great for connecting pieces
of audio visual gear together because each one will have its own power source. All that needs to get sent between them is
a little bit of data. But lots of things need power, too. Have fun using that keyboard and mouse connected
to your computer by nothing but light. And, as we moved away from the days of RS-232
serial connections and into the holy land of the Universal Serial Bus, suddenly we found
that we’re not sending little bits of power just to keep your mouse alive. We’re sending loads of it. I don’t want to go off on too much of a
tangent here… Oh who am I kidding I do this all the time. It seems like USB turned into the power delivery standard it is today kinda by accident. Remember that even in USB 2.0 days, officially
a device could only draw 2.5 watts, and that was considered high power. In those wild west days, different companies
were coming up with their own ways to determine the current a given port could supply, and
it wasn’t until 2007 that the first USB Battery Charging specification appears, finally
getting everyone on the same page and allowing for USB ports on computers to finally, officially,
provide more than 500ma when asked nicely. Anyway, now that USB is ubiquitous, we can expect
any modern device to provide a decent amount of power through its USB ports for things
like portable hard drives, charging your phone, and of course the monumentally important task
of powering all those RGB LEDs in that gaming keyboard you bought because, you know, +5
Agility. But you may recall that the switch from USB
2 to USB 3 didn’t just involve making the data transfer go faster. The 4 conductors of the USB cable weren’t
enough, so we had to add more of them for USB 3. Five more, to be exact. Add too many more and we might as well just
start using HDMI for everything. But with fiber optics, we could potentially
have much, MUCH faster data throughput using only a pair of optical fibers. Or, potentially, just one if you, for instance,
send data in each direction using two different wavelengths of light, and separate them with
a prism on each end. So sure, we need to send power to devices
in addition to data. Why not create some sort of composite cable (no, not that kind) a cable that is a composite of both optical and electrical? Like a USB cable where the data lines are
replaced with optical fiber? Then, the same cable could be used for nearly
everything! Alls we gotta do is just make the LEDs go
blinky blinky a little faster and we coulda been using the same cables since the ‘80s! Well, not so fast. Heh. Not so fast. Optical fiber is complicated. It doesn’t seem like it should be, after
all what are we really doing but flashing a light source and moving it somewhere else
so we can see the flashing and decode it, but thanks to … PHYSICS, it’s not actually
as simple as that. Stupid physics. Making everything difficult. I don’t want to get too far down this particular
path because its importance to potential consumer standards is arguably minimal, but it’s
important nonetheless. Since light travels at the speed of… light, it may seem as though a pulse of light down an optical cable will reach the other end
in exactly the same way that it was sent. But… it won’t. Thanks to a phenomenon called modal dispersion,
also known as multimode distortion, multimode dispersion, modal distortion, intermodal distortion,
intermodal dispersion, and intermodal delay distortion (jeez guys pick a name already) the signal actually gets a little smeared. Let’s imagine we send a pulse of light down
this optical fiber. Will it all get to the other end at the same
time? It seems like it should, it’s light after
all. But just because something happens as fast
as we know things can happen doesn’t mean geometry doesn’t apply. If the individual photons in that pulse of
light manage to stay perfectly parallel to the sides of the fiber, they will all arrive
at the other end simultaneously. But this is the real world, and nothing’s
perfect. Some of them are gonna enter the fiber at
an angle, which means they don’t take a straight path. They bounce off the sides, and although total
internal reflection keeps them from leaving the fiber, this zig-zag path is much longer
than a straight line so those meandering photons will in fact arrive later. And this limits the speed at which we can
pulse the light on and off and still have it be intelligible on the other end. This is that smearing of the signal. Even though on the sending end we’re putting
in a clear-cut, on-off signal, on the receiving end the photons who lollygagged and bounced
around in the fiber cause the signal to look like this. Some of the photons from the last pulse manage
to arrive at the same time as some of them from the next one. So, we have to limit the frequency at which
the pulses happen in order to make the received signal decodable. This means that optical fibers do actually
have a bandwidth limit. Just because we’re using fiber optics doesn’t
mean we have theoretically endless upgrade capacity. That said, you might have already surmised
that the bandwidth limit depends on how long the optical fiber is. No matter how poorly the light behaves in
the pipe, if it’s a short pipe then the time difference between straight-and-true
photons and disastrously off-course photons is quite small. And this is why discussing the bandwidth limitations
of a given optical fiber is not really a big deal if we’re dealing with consumer applications
where cables aren’t likely to go beyond 10 meters. But, since it’s interesting, let’s get
into how we deal with bandwidth limits anyway. In the land of fiber, there are two basic
kinds of fiber optic cabling. Single-mode fiber, and multi-mode fiber. Now, mode here doesn’t mean a method of
operation. Here it means the field pattern of propagating
electromagnetic waves. Ideally we want the light to travel only in
the transverse mode, and that’s what single-mode fiber allows us to do. It accomplishes this by having a very small
internal diameter, what we in the business call itty bitty. Generally it’s between 8 and 10 micrometers
across. This tiny size of the fiber allows the pulses
to remain distinct over longer distances because the light traveling through the fiber mainly
stays in the transverse mode. Because of this, long-distance links on the
order of thousands of kilometers are usually done with single-mode fiber. But, single-mode fiber is really finicky to
deal with, requiring special tools to make connections since it’s so darn small. So for applications involving shorter fiber
runs, like networking a building, we’ll use much easier to deal with multi-mode fiber. This thicker fiber allows us to make cable
terminations more easily at the cost of a messier signal on the other end. Still, it’s a fair bit of bandwidth. Using just LEDs, gigabit speeds are easily
achievable. And if we get into higher-grade fibers and
start using lasers as transmitters, we can get 100 gigabit speeds over distances on the
order of 100 meters. And remember, the shorter the run, the easier
it is to achieve high bandwidth. But if we’re talking about a fiber like
TOSLINK, well that’s nothing like even multi-mode fiber. It’s actually in a third category. Generally TOSLINK is a plastic optical fiber. It’s much thicker inside, about a hundred
times thicker than single mode fiber, so it exhibits much greater modal dispersion, and
thus its bandwidth limit is much lower. But how much lower is it, really? I mean, shorter lengths make the bandwidth
go up, so even though that fishing line in there has wicked bad modal dispersion, over
just a few meters it shouldn’t matter much, right? Right! In a 2009 paper by Yasuhiro Koike, the possibility
of using plastic optical fiber for high speed networking was explored. By using different modulation methods, it
was found that plastic fiber could easily achieve gigabit speed at lengths up to 100
meters. And, with the development of graded-index
plastic fiber, 40 gigabit speeds were achieved. Pretty impressive. But, now we’re stepping into messy territory. Let’s rewind a bit. Graded-index plastic optical fiber is a new
development. And the question I’m getting at here is
why haven’t we been using fiber forever? Imagine that in 1985 we had come up with one
standard cable, like a bidirectional TOSLINK cable with a pair of power wires running through
it. I’ll call it UniLINK. Could we have been simply been using the same
cable to replace the functions of all these, and had a future-proof design because the
speed could simply be boosted with each new generation of hardware? Well, maybe. A TOSLINK cable with a 10 meter length could
perhaps pull off 10 gigabit speeds. And so, if we had designed a cable like my
theoretical UniLINK cable, then perhaps we could have simply had one cable to rule them
all. Oh, and wouldn’t ya know it, I found this
product guide from Toshiba and it turns out that TOSLINK did have a few bidirectional
connectors out there. They are generally limited to professional
and niche applications like automation control, but this high-speed TOSLINK connection is
capable of a quarter gigabit at 20 meters. Not too bad, and I don’t doubt that could
be improved. But, well, now here’s where the heavy weight
of reality steps in. As the old adage goes, just because you can
doesn’t mean you should. It may seem silly to use such complicated
cables when commercial fiber optic links are now pushing past terabit speeds, but just
looking at the various connectors and cables we’ve used through the ages doesn’t really
explain what they’re doing. An HDMI cable doesn’t have 19 conductors
running through it for grins and giggles. Most handle the video data in various ways,
but others handle things like the audio return channel (and notably that pin was unused in
early versions of the standard), the display data channel which allows displays and whatever
they’re plugged into to get a nice introduction and learn their respective preferences (also
this is where HDCP runs along for the ride), a 5v power supply, the consumer electronics
control which is what allows other devices to turn on and off your TV for you, and many
of these pins have been given more and more tasks as HDMI has matured and improved. Could we do all of that with one single fiber
optic cable? Perhaps. Though HDMI 2.1 has a total bandwidth of 48
gigabits per second, which might not be possible with plastic fiber at all. Or at least, only over relatively short distances. See, it’s interesting to think about the
blazing fast speeds fiber optics allow, but a super fast serial datastream isn’t necessarily
useful for driving the pixels of an LCD panel in a logical fashion. And therein lies the problem. We have a bunch of different cables to deal
with because they’re all designed to do specific things in specific ways. The end! ♫ unexpectedly smooth jazz ♫ Just kidding! Though, that pretty much is the answer. At a very fundamental level, they’re all
carrying power and data. But how that data should best be transported
depends on what kind of data it is and what we want to do with it. As new standards came up, their cables were
designed to address those needs. And oftentimes the best way to take care of
a specific need is to add another conductor for doing just one thing. So, while we could just send everything over
a high-speed fiber connection, the extra processing that might be required on each end can make
the whole idea more complicated, and expensive. And so, new cables for new applications often
make the most sense. If you can simplify the data processing with
a more complicated cable, it’s often worth it. And now, I’m about to shoot that argument
in the foot. Did you know that there are HDMI cables which
are actually fiber-optic? If you need to run an HDMI cable over a very
long distance, one of the easiest ways to do it is to take an HDMI signal, convert it
to a fiber optic data stream, send it over a fiber however long you need, and convert
it back to HDMI on the other end. Commercially available products can do all
that in what looks like any ordinary HDMI cable that just happens to be very long. These cables actually cheat a little bit and
have four fibers going through them, probably one for each of the three sets of data lines
and the fourth for the clock signal that a normal HDMI cable carries, to make encoding
and decoding easier, but they demonstrate that the actual hardware required to convert
to fiber and back again isn’t that complex. It all fits inside these connectors. Back in 2014, LinusTechTips demonstrated a
USB 3.Optical cable from Corning that, well did the same thing but for USB. The tech to convert to optical and back was
a little more expensive at the time, but it still all fit in modules barely larger than
your basic flash drive. So, what’s the deal then? What exactly is stopping us from using a UniLINK
hybrid power and optical cable for everything? Um, nothing. Except of course for all the other cables
we already have. This is why I love looking at technology through
a historical lens. TOSLINK is surprisingly old for what it is, but at the same time it’s kinda in its own little corner. Consumers haven’t needed fiber optics for
bandwidth reasons until very recently, so unless there was some reason electrical isolation
was absolutely necessary, using a pair of copper wires was sufficient. So, throughout the rise of the digital age,
we just kept on going with copper. Because it worked. Lots of technological progress comes from
using existing things in new ways. Just look at how we first got the Internet
into our homes–using ordinary telephone lines! [dial-up modem sounds]
Just by calling a specific phone number and having your computer screech at another one, you’re online! And then, we adapted those phone lines into
basic broadband using DSL, and if you’ve got cable Internet, you’re getting your
memes over the same coaxial cable that’s been delivering …must-watch television programming
for decades. And it works! Backward compatibility is also partly to blame
for keeping optical tech in the dark. Like in the case of USB, more data lines were
added to the existing connector, rather than create an entirely new one. It was another case of using existing things
in new ways, though with a little more flair. The only practical time to introduce an optical
standard is when creating an entirely new standard. Speaking of creating entirely new standards,
Thunderbolt was almost optical. In fact, it was originally called Light Peak. But somewhere in the development process,
Intel realized they’d like to be able to send power through these cables, and while
they pondered stealing my idea and bundling copper wire with optical fibers, by 2011 they
gave up the fiber thing altogether, realizing that copper worked just fine. Though, it’s worth noting that in the development
stages, they were pushing 10 gigabits a second through plastic fiber, and believed they could
get to 100. Also of note is that, just like Corning’s
extra long optical USB cable, optical Thunderbolt cables were a thing, but this hasn’t yet
become a reality for Thunderbolt 3. Speaking of Thunderbolt 3, we do seem to be
headed towards a One Cable to Rule Them All future. Or at least, maybe. USB type C can do nearly everything we might
want a cable to do, and it can do it pretty well. Nevermind how incredibly confusing the whole
situation is right now because the connector is called USB type C but that doesn’t actually
mean anything in regards to what data can go through that port or any specific cable
because the USB Implementers Forum sucks at branding and bundling Thunderbolt into the
same thing as USB is making this mighty confusing. Get it together, people! So where does this leave us? To put it simply, it’s complicated. We do sort of see fiber optics in the consumer
space. Though most of those implementations are indirect
and just a way of getting around cable length limits. It doesn’t seem likely that we’d end up
with a TOSLINK-like cable that sends only optical data, because we need copper to send
power, anyway. And if we need some copper, why not just do
it all with copper? I imagine that’s the conclusion Intel came
to in the development of Thunderbolt. Will another attempt at creating a truly optical
standard find success? Well, I’m doubtful. Thunderbolt 3 and/or USB 3.2 or wherever we
are in this situation is probably way faster than what any casual consumer is gonna need
for a while. I won’t say ever, cause you just can’t
do that when it comes to tech, but then there’s this other elephant in the room I haven’t
mentioned yet; WiFi. [in a Valley Girl voice]
Wires are sooo yesssterday. Why use a wire when I could live life untethered? Ugh, so gross these wires. I hate to break it to you, but your average
consumer is not your average enthusiast, and so if there’s a wireless option, that’s
probably preferred. Why do you think headphone jacks are going
away? I mean, I was mad at OnePlus for ditching
it but today I rarely miss it. So yeah. There will always be a need for super-fast
wired connections in the professional space, and lots of you enthusiasts out there love
to live in that same tier. But let’s be honest. You’re not the average user. That’s not a bad thing, not at all! But it deserves to be said. For now, I imagine optical connectors will
stay right where they are. Simultaneously at the very bottom with TOSLINK, and at the very top with crazy fast networking equipment. For everything else, copper’s pretty OK. And for everything else else, microwaves are
fine, too. And for those weird niche cases where we are
running fiber as a Frankenstein solution to getting around a length limitation, well to
that I say, why not? Thanks for watching. There’s a lot I didn’t get to before ending
this video, so as these fine Patreon supporters start scrolling up your screen, I’ll bring
a few of them up. When we make long-distance fiber optic links,
we usually can’t send light all the way from one end to the other because even the
best optical fibers do attenuate the signal somewhat. In the past, we used electronic repeaters
which simply read the incoming signal and repeated it using another laser to enable
longer connections. However, these days we use optical amplifiers
which use some physics magic to passively make the light beam more intense without actually
creating a new beam of light. Because you don’t need a bunch of active
devices repeating the signal along the fiber’s length, you can speed up data transmission
within it. I don’t understand the operating principle
of optical amplifiers well enough to provide a good explanation of how they work, so if
anyone in the comments wants to take a crack at it please do. Also, I do want to point out a valid criticism
from the last video. HDMI sure does have more audio bandwidth than
TOSLINK, but what if you only want to send audio? Since HDMI is primarily a video interface,
audio-only just isn’t a thing it does. And, that’s a fair point. However, when we’re dealing with multichannel
high-resolution audio, I think it’s fair to say that we’re usually talking about
the home theater space, and that’s why I was ignoring that potential use case. Also! The Audio Return Channel was until HDMI 2.1
limited to S/PDIF, meaning until very recently TOSLINK and the ARC were essentially the same
thing, and since many people reported having issues with the ARC, TOSLINK provided an alternative
with just-as-good audio and with better compatibility. Now, though, the ARC is truly better, so assuming
you don’t run into problems, it leaves TOSLINK in the dust. Another fun little thing was that I did run
across some research looking into how to send electrical power over optical fiber. Yep, power over optical fiber is a thing,
but I can’t seem to find any specifications on exactly how much power can be sent. And I doubt it’s very much at all, though
I’m willing to be surprised. And lastly, following along those lines, one
of the things that CAT5 cabling lets us do is send power in addition to data over the
same lines. This is called Power over Ethernet, and … wait. It’s called power over ethernet, eh? Hmm. How interesting. Anyway, power over Ethernet allows for things
like powering wireless access points without needing an A/C power source at the location,
providing power to IP phones in office buildings, or any other such need. And up to 50 watts! That’s some powerful stuff, that Ethernet
is. ♫ abridgedly smooth jazz ♫

100 Comments

  1. Im an electronics enthusiast and FUCK BLUETOOTH!!! I and my family who arent 'techy' have had far more problems with that standard than just… you know…plugging my headset into a freaking port. You know, ensuring a physical electrical connection. So no sir, wireless is not always better to 'non-enthusiasts'

  2. You talked about the reasons it would be "hard", but I would ask "How about LiFi? Why is that not a popular connectivity option for areas with congested WiFi spectrum?"

  3. 13:23 "Super fast serial data stream isnt exactly useful for driving the pixels of an LCD panel in a logical fashion "
    What? All video interfaces are super fast serial data streams.

  4. The only really good way to connect to my old surround reciever is toslink. I was pleasantly surprised to find out toslink provides excellent sound quality.

  5. The reason that there will never be one cable to rule them all is because there will always be financial incentive to make something else.

  6. Modal Dispersion through Intermodal Delay distortion (6:31) all sound like they should be Portal elements within the walls of Aperture Science!!

  7. Single and multimode fiber are externally the same diameter because the cladding pads both up to 125 micrometers. Also, preparing and splicing them is the same. So practically, you can also use singlemode for in-house-application. From the price standpoint, the transceivers are a bit more expensive.

  8. In the broadcast TV space, we have a fiber-power combination similar in concept to your patent-pending UniLINK. SMPTE 304M is primarily used for powering and connecting broadcast cameras with one cable; it combines two single-mode fiber links (send and receive), four 20 AWG copper conductors for power, and two 24 AWG conductors for control voltage.

    While we also use triaxial cable (triax) for the same purpose – often muxing the bidirectional signal back to fiber anyway – we like the SMPTE connector for the bandwidth upgrade it provides over triax. In the sports world, it's used for super-slo-mo cameras that can be replayed with no juddering; triax doesn't have the bandwidth to transmit 180-240 fps (6x/8x slo-mo) video at 1080i/p or 4K resolutions.

    (At the same time, we hate the SMPTE connector for its ease of gathering detritus in the field and its relatively annoying cleaning method, usually having to be done minutes before air – go figure…)

  9. @19:28 OIC what u did there… made the concept of wired STEREO headphones appear as two angelic spirits reflected off the back of the … device … two yellow orbs, waiting patiently, to conduct their freqs once again…. YOU ELECTRORACIST!

  10. As a person currently studying for an A+ certification, I am 100% convinced the only reason tech people are so damn particular about the naming conventions is so they can make people understand second hand the trauma of memorizing every name of every cable and every bit and bauble on the off chance that the test ONLY labels it as a "Cat5" and not just an ethernet cable like -everyone else on earth does-.

  11. You did mention WiFi. That works well for the most part but sometimes the signal does not make it through all of the walls that are in its way between it and the device that is connected to it. Then you would want to make sure you have a decent brand of WiFi router if you plan to use that connection a lot. Maybe it is just me but WiFi doesn't seem all that great when it comes to speed. I have 100 megabit internet but barely get 6 megabits upstairs on WiFi. The desktop downstairs connected to Ethernet is so much faster that it is a night and day difference. All devices connect to the same router BTW so it is not like I have one for one thing and one for the other which would be kinda silly anyhow. Maybe I should try a wireless AC router since all of my devices are compatible anyhow. I will just hold onto the receipt just in case that for some mysterious reason doesn't solve my problem.

  12. USB is a really great invention/agreement. But…..I have no idea which USBs on my laptop are 2.0 or 3.0. Would a small nimprint be too much to ask Lenovo? BTW, my laptop is 4yrs old, with an AMD A6, some Radeon graphics stuff, 8gb & 1Tb. Cost $500. I just recently got my mum a new $500 laptop. Guess what? It uses an A4, lesser Radeon stuff, 8gb & 1Tb. So moore's Law is dead & buried. Her laptop is definitely inferior in every way where it is not identical. It's slower. So much for progress.

  13. Balanced signals going down a cable are able to do away with the grounding loop problem. When you see a "data +" and a "data -", it means a balanced signal. This setup makes the "data -" into a floating ground. This is why HDMI does not have ground loop problem.

  14. Power over fiber= Solar geared to light-wave produced by laser at end point. At the same time the data can be sent through the mini- solar panel while charging through a split signal solar cell. You could easily get the 5 volts you need this way for most devices filtered through a capacitor circuit to get a solid signal for the power. Old calculator solar cells were 1.5 v, imagine what they can produce now that solar has advanced. The only reason I know about this is I tried to build it. Have not gotten far enough with the prototype to test because its a hobby and life happens. I plan on attempting to power a house this way one day.

  15. There's no reason a serial data link should have any limitations on the type of data that can be sent over it, as can clearly be seen in the versatile USB type C/Thunderbolt carrying everything from HDMI to PCI-Express. Moore's Law's basically dead and they're still managing to increase USB-C/Thunderbolt speed so I doubt the cost of adding optical interfaces to everyday computing hardware will be justified anytime soon, though a universal fibre cable designed from the ground up to carry power would be great. The relatively short cables used in home scenarios should allow very high transfer speeds as the dispersion effect he mentioned only becomes significant over very much longer distances, one example I saw was of a connection speed of 10 gbps @ 30 km distance dropping to 2.5 gbps @ 500km distance, hardly a showstopper for even the most ambitious of LAN-parties!

  16. While I applaud your attempt to please the pedants in the audience, by mentioning that what folks commonly refer to as an RJ45 connector, is actually called an 8P8C—it still doesn’t quite tackle the can of worms you decided to open.

    First off, while you’re technically correct-ish, RJ45 was never actually the name of the wiring standard—RJ45S was, which was wired to an 8P2C connector, which LOOK like an 8P8C connector, with an extra nub on the side to make it keyed (more than it already was, I guess the FCC thought that was problem?)

    The RJ45S wiring standard is now obsolete and since it’s no longer in use anywhere, as it’s been replaced by other standards, like the 8P8C (which is a clunky AF abbreviation to actually say, and likely why even George Lucas wouldn’t have used it) which is typically wired using either the TIA/EIA-568 (A, B or both for crossover cables) which is what’s used in the most common “Network Cable” that you’re likely to encounter as an end user.

    The REASON that pedantic references to the name of the connector, that we’re all familiar with in most modern network applications, as an 8P8C isn’t necessarily a better option—is that with that EXACT SAME 8P8C connector, if a different wiring scheme than the TIA/EIA-568 is used, it would actually become an ISDN line that uses RJ48, RJ48C for T1, RJ48X (which is a jack that shorts when a cable is removed to create a nifty self-looping function for troubleshooting) or the RJ61 (that’s the next RJ standard that’s going to die, because it sucks for high speed Ethernet uses). [Honorable mention to the janky Ethernet (8P8C) to Serial (RS-232 connectors) that are required to configure Cisco hardware (which I believe uses their own proprietary pin-out, instead of the RS-232 standard that’s used by other serial devices.)]

    Again, while those are much less common uses of the 8P8C, they still serve a purpose in broadcast, telecom, and network applications—which is why they haven’t been completely phased out like the RJ45S has been, or RJ61 will be.

    The term RJ45, while NOT an official standard, was something that techs in the field started using to specifically refer to 8P8C connectors that are wired using the TIA/EIA-568(A or B) standard, which includes (but is not limited to) Ethernet. The reason it isn’t a recognized standard, is because there are already official standards that comprise the specific type of cabling required for the application that the unofficial term “RJ45” refers to—so it’d be redundant. That said, it GREATLY decreases confusion, when you’re also running separate cabling for voice and data, and they use the same type of cables, and connectors, but differ in the method that they’re terminated. Again, less common now that ISDN and T1 connections used much less frequently for new telecom installs—but sometimes it’s easier to re-run cables than it is to replace an entire system that’s still working, and ISDN is still widely used in broadcast applications. (Radio stations that are linked via ISDN lines have no perceptible delay or loss in quality, which is how syndicated shows are broadcast live in multiple markets, or in some cases allow radio hosts to work together on-air from completely separate studios. [Not to name names, but Bob & Tom stooped broadcasting from the same studio a long time ago, because they stopped getting along—and guests were told not to acknowledge the fact that one is them wasn’t actually in the studio.)

    So, to bring it all home here’s the TL;DR for folks who skipped ahead (or got lost in my pedantically winding trip). RJ45 is not an official standard, but it does serve a very practical purpose, as it simplifies which specific connector/jack, and wiring standard you’re referring to, when you’re trying to discuss the type of cabling that’s being used or installed for a large number of common applications.

    Other than THAT little bit of extra context and detail that I thought I’d fill in—I enjoyed the video.

  17. So, with the light dispersion in the cable, what if they were to coat the insulation of the cable with vantablack which effectively absorbs most light, would that theoretically take care of the issue? Just a thought 🤷🏻‍♂️

  18. This channel is amazing. I've been in wireless Telecom (fiber everywhere ) for a few years and this video basically tripled my knowledge of fiber optic cables (we just call them fibers).

  19. Nice video! BTW, your videos are awesome and I'm following your posts since the analog TV era. It would be really nice to do a "ground loop" video and explain why my amp and mainly my on-board audio interface hums … thanks!

    BTW, could please enable translations on your videos? Thanks!

  20. Interestingly, the smearing of the signal is what enables optical amplifiers – completely optical … really crazy! Brazil developed it's own fiber optics using the most modern standards in the 80s … with less smearing … which caused problems later on.

  21. Every time i turn on my tv my amplifier switches to a different audio input leaving me to have to swith it back over to the input i have it plugged into. When i switch it over to the input i have it plugged into it works fine but having to switch it every time i turn on my t.v. or even open a new app on the tv, is qute annoying. I tried plugging the red and white coax cables into the input that the amp swiches over to but it dosent work. Im not sure why as both are new, but this has led me into thinking about buying a toslink cable, seeing as there is no electrical connection between the two devices with that cable i feel it should be safe to assume it could no longer switch the amp over to a different input, although i questing the ability of standerd coax cabes to change the input on my amp im frustated and willing to try most anything. I wonder if the tv remotes infer red communicator might be tuned to the same input as the amp remote puts out to swich over to that channel, which would be just my luck. I dont know if anyone could help with this funky problem, but its quite irritating every time i go to use my tv.. ..

  22. Quick note, TOSLink being optical means the two devices are optically isolated, not galvanically. Galvanic isolation refers to power or signal transfer via electromagnetic field coupling, usually via transformer. Optical isolation is specific to power or signal transfer via photons. Both isolation techniques prevent ground loops as the video describes, so the difference here is more pedantic than anything. Ethernet is an example of a communications technology that features galvanic isolation.

  23. Sorry to correct you, but your explanation of modal dispersion is wrong. Do you want me to send you the correct explanation?

  24. The "no audio only" limitation of HDMI is something I have had issues with several times in my home cinema setup. Everytime I upgrade my display tech (from HD to HD 3D and now to 4K HDR) I've had to upgrade my surround receiver because it doesn't support the new standard. But if HDMI could send audio only I could just run one HDMI from my HTPC to the TV/Projector and a second HDMI-cable from to my receiver for the audio without having to have a "ghost" display active on my desktop where windows can get lost.

    At least next time I upgrade my display both the display and receiver will support eARC and have enough bandwidth to transmit surround sound with PCM which the HTPC needs for games as it can't encode audio as DD or DTS on the fly… for some unknown reason since my PS4 does it just FINE….

  25. You fool! The "speed of light" is not, in fact constant! Light travels through different mediums at different speeds! When one is referring to the "speed of light" generally one is likely referring to – or should be referring to – the "speed of light in a vacuum", or, perhaps more accurately, "the maximum speed of information!"

    P.S. You're not a fool. But this is fun.

  26. Nice channel. I often work in A/V and was wondering if you could do some videos about the history of various pro connectors and standards like SpeakOn, XLR, AES3, Digital Snakes, DMX lighting protocol standards and wiring, SDI (serial digital interface), etc. Example, lots of pro speakers use these SpeakOn connectors these days, you see them a lot when working in the AV event industry: https://en.wikipedia.org/wiki/Speakon_connector

  27. I guess RJ45 "connector" is technically ok, its the pin-out that is different right? So the pin-out would have a different name.

  28. Also why not call the "distortion" thing with a thousand names "optical diffusion" because it is light, and it is getting diffused basically right? Lol

  29. Couldn't you just use multiple strands? using that information and piecing it together at both ends in effect you could pulse the light in a particular order across the manifold of fiber? If the limitation is the speed at which the light can flash multiple strands could help right? Then again you gotta piece it together so it might negate any benefit idk.

  30. Astronomers use a system called adaptive optics to correct distortions of the light coming from stars due to our atmosphere. I wonder if a system could be devised to correct Modal Distortion based on the same principle of adaptive optics. Like some kind of gate filter to make the light pulse coherent again before being decoded.

  31. I laughed. I laughed hard. The first interruption caught me off guard since I hadn't seen it in previous videos. The second interruption was hilarious.

  32. I like to piss off audiophools by replacing my fiberoptic with rusty coat hangers soldered together (using plumbers flux and a blow torch) to hook up the digital out and marvel at the sound quality shh its the same

  33. 16:26 YOUR TELEPHONE TO MODEM ADAPTER IS FAKE !!!!!!!!!!!!!!1111ONE ITS A CUP HOLDER WITH TWO BEER COZY'S with a wire!!! Didn't think we would catch that but we did.

  34. Why do I feel like something funny was happening behind the camera at 2:58 lol. The way your eyes move over and you slow down what you're saying as if you were witnessing someone making funny faces at you. I spy a slight smile too right before you continued on. I need to know!

  35. 3:17 Wait, cat 5 isn't Ethernet??? What about cat 6,7?? What is Ethernet cable? What different between Cat 5,6,7 cable and ethernet cable?

  36. I have a question with a bit of prelude.

    I've been looking to pick up a charging dock for my Nintendo 3DS and one thing I like about it's dock and the Wii U Gamepad's dock is that, unlike the Switch's dock, they don't cause any additional wear on the charging port. Both have exposed copper contacts and when placed on their docks, they push down on a button which exposes and elevates prongs that hit the contact. That made me wonder if they could also send audio and visual data to a dock in a way that doesn't cause additional wear.

    Obviously they could just add additional contacts to send that info but watching this video was kind of illuminating (lol) because the 3DS has a IR blaster right next to the power contacts. Obviously it's low-bandwidth because it's intended to communicate with things over open-air but in a dock where said blaster could be pressed right up against an optical decoder, couldn't an IR blaster theoretically have the bandwidth to wirelessly send AV data?

    Obviously, the optical encoder would have to be able to blink the LED(s) fast enough and I'm positive the one in the 3DS can't but it's very line of site, a very short distance so it shouldn't have any of the setbacks that make consumer fiber optic cables impractical.

  37. We can do much beter and realible contacts with electricity than light, and why using optics for short distances when for coper we don't need any complex receiver like we need for light, and on short distances both can transfer data at the same speed, exept coper doesn't have unecesary ping and complex receivers and transmitters.

  38. great video!
    just one question… you said you don’t miss the jack plug anymore, but what about the cancer risk related to exposure to radiofrequency from bluetooth headphones? this still scares me a bit…
    anyway, keep the good work
    greetings from Chile

  39. My LG G7 has HiFi Quad DAC so I switched to Aux in my car because it sounds crazy better. It's about the best you can get for sound short of Android Auto or Apple CarPlay in the dash and a straight path inside to the amp. I would not like losing my DAC and therefore my headphone jack.

  40. For given distance, fibers will always have better bandwidth than wireless/wires Because the wave is enclosed in a tube, also there's less interference and obstacles, plus higher frequency of 'light', plus multiple wavelength can be used at same time.But our equipment isn't fast enough to handle it.

  41. Copper is good enough for any consumer application. Except for consumer laser cutters I guess. fibers simply have very little flexability and are simply too expensive.

  42. The reason Ethernet can deliver such high power is, the voltage can be very high… Like 48 volts which is high enough to make your muscle tingle.

  43. "All those rgb LEDs in your keyboard because, you know, +5 Agility." LMAO..imma stick to my old mechanical, thanks. LEDs on a keyboard make me nauseous for some reason

  44. I just laid two 20m copper HDMI cables into my wall during a refurb. Both won't take 4k over that length. Ordered another thinking we had broken it, no signal on that one. Finally went for a really cheap fibre HDMI and boom, working. Wish I used them before. It's sorted all my issues out and it's as easy to bend and work with as Cat5e

  45. I made this comment already and I might just be speaking against an echoy chamber that agrees with me on the other side but:
    Ethernet, is category of different cables of the same protocol but different abilities.
    Whoever who does not accept ethernet as a catch-all term for general RJ45 connector on both sides cable and insists it is Cat5e is way more wrong than the person that calls it ethernet because ethernet is not accurate but usable term while cat5e has a categorical chance of one in five of being wrong if you have not actually looked at the cable and seen what it is or tested it yourself to be sure what it's capabilities tell about it's category.

    Calling an ethernet cable an ethernet cable means you want just a cable that connects two endpoints in ethernet TCP/IP network (USUALLY) and you do not care about it's quality or it has no bearing on current conversation while calling a Cat5, Cat6, Cat6e, or Cat7 cable a cat5e is categorically wrong.

    To me the joke about IT gods is fun but I am sadly amused more by the fact someone probably did correct you with such a almighty attitude yet categorically wrong knowledge.

  46. As a prophet of the one true TechMoan, I think all cables are fine.. as long as they're new old stock from the 60's and somehow have a vacuum tube in it

  47. I still have my 1Gbit internet. I don’t care what my router and the other side of the fiber does, as long as it works 🙂

  48. Excellent video! As a network engineer, this was the best description of modal dispersion that I've ever heard.

    Just a quick note though – Multimode fiber can be (depending on the speed and grade of the cable) a few times longer than a copper ethernet run (300m vs 100m at 10Gbs), but still isn't usually going to be used for long runs through large buildings or across campuses. Where I work we have multimode connecting some devices in the same room, but use single mode for everything else (the vast majority of our fiber infrastructure). It's more expensive to buy and get installed, but once in place we use pre-manufactured patch cables (as does everyone), so it isn't any more difficult to work with.

    The best practices in networking actually are evolving into using singlemode for everything. The fiber is a bit more expensive, as are the optics. The advantages though are twofold – for one, singlemode fiber is capable of higher speeds over longer distances, and secondly, if technology improves, you can upgrade by simply replacing the optical transceivers, and reuse the same fiber. With multimode, for example, if you had wired your building with OM1-grade multimode years ago, and now wanted to go to 40Gb/s, you'd be out of luck – you'd have to run all new fiber.

    Just a bit of trivia there. Love the videos! Thanks!

  49. Way back in 2007 (before PoE, LEDs, and efficiency we’re really a thing people cared about), my boss and I looked over the specs for the best PoE injectors and lowest power consuming monitors with the dream of building a wall mounted thin client to connect nurses and doctors to the HIS with a single cable. We ended up settling on large, battery powered mobile computing carts connected to WiFi. He died five or so years ago and I still check on that technology from time to time. His idea was sound (and I think USB3 now provides the power necessary), we were just 10 years too early.

  50. toslink has a lower bandwidth than coax. thus why coax is still the prefered connector for most audiophiles. and HDMI blows both of them away. my sony 4k blu-ray player has HDMI audio only out along with the standard HDMI out. higher end equipment is implementing hdmi audio connectors now. fiber optic isn't going anywhere. 5g internet needs it to work. and some cable/phone companies are using it for their internet connection.

Leave a Reply

Your email address will not be published.


*