If you’re watching something like, well
like this, on a modern display, you probably don’t think too much about what your device
is doing to make this happen. It’s not really that extraordinary in this modern world filled
with computers and microprocessors. In fact, I’m willing to bet you have a rough idea
of how this works. But I’ll explain anyway. If you get up close to a TV or monitor, you
find there’s a grid, made of millions of little squares called pixels. From far enough
away, these pixels blend into each other, and our eyes and brains build a coherent picture.
To actually create the image on the screen, each pixel can have some instructions sent
to it to tell it how much light it should emit. A series of controller circuits work
with millions of transistors to methodically give every one of these squares a specific
brightness value 60 or 120 times a second, sometimes more often than that. The instructions
are further divided into three separate values for red, green, and blue which when combined
together can create practically any color you can think of. But have you ever wondered how old-school
TV worked? We’ve been sending video signals over the air for a very long time, in fact
analog television predates World War II. There weren’t computer or logic circuits decoding
number values then, in fact there weren’t even pixels. Yet, this somehow worked. How?
What magic is going on to take the signal coming over this wire and turn it into a black
and white image of me, all without a single digital circuit? To find out, let’s make things really simple.
Instead of looking at millions of pixels, let’s look at just one. A single pixel is
really just a point of light. Without using any digital circuitry, how can we tell the
light what to do? Easy, by controlling how much power it gets. Here’s an LED hooked up to a power supply.
By simply varying how much voltage it gets, you can change how much light it emits. Using
radio technology, it’s really easy to build a circuit that can control the brightness
of this LED or another light source based on the strength of a signal transmitted over
the air. Of course, this is clearly not television, but it’s at the core of what makes it work. See, we really suck at seeing things that
happen quickly. Our eyes and brains are just no good at processing fast visual information.
Thus, it’s really easy to trick our eyes into seeing something that’s not really
there. If you take a light source and move it very quickly, you no longer see a single
light source, instead you see a continuous line that follow the path of the light. Our
brains can’t process the light’s fast motion, and it just blurs together into a
solid line. This phenomenon is called persistence of vision. Now, if you manipulate the voltage
of the light while you move it, you can make patterns in the line. You can find a lot of toys that exploit persistence
of vision. This old 20Q game works using this principle. A small circuit board with a few
LEDs on it spins in a circle really fast, too fast for your eyes to keep track of. If
the LEDs were lit up all the time, it would just appear to be a continuous circle. But
the game uses sensors to track where the LEDs are, and turns each of them on and off at
very specific times. By manipulating the brightness of the LEDs and timing it with their motion,
it can draw simple text and graphics using just eight points of light. But a CRT television like this has only one
point of light to work with. After all, this is 1920’s technology, and multitasking wasn’t
really a thing yet. So first, what actually makes the light? Well, CRT stands for cathode
ray tube. The name comes from cathode rays, which were discovered by Johann Hittorf in
1869. William Crookes had created these goofy tube things that were really important to
early scientific discovery. He was able to evacuate nearly all the air from these tubes,
which allowed electrons to move freely within them, though no one yet knew what electrons
were. When electric current was sent through these tubes, something caused them to glow.
Johann Hittorf was the first person to piece together that whatever was causing this phenomenon
travelled in straight lines from the cathode, or negative electrode, observing how a stencil
between the cathode and the surface of the tube cast a shadow. Eugen Goldstein gave them
the name Cathode Rays, just like rays of sunlight. J J Thompson would later use these tubes to
work out what these cathode rays were actually made of, and in doing so he happened to discover
the electron. So good on him. Before we continue, SAFETY WARNING: Exploring
the innards of a CRT television can be quite dangerous. A set as small as this can generate
over a thousand volts through the flyback transformer, and the CRT’s glass can store
a lethal charge. I know what is and what’s not OK to touch, and since you likely don’t,
don’t try this at home. If you’ve ever messed around with antique
radios you’ll have seen vacuum tubes, which are the precursors to transistors. These electronic
components have the air evacuated from them so electrons can move freely, just like the
crookes tubes. Using a heater filament to induce thermionic emission from a cathode,
they can manipulate electric current in a bunch of ways. A television CRT is really
a specialized vacuum tube that has had its top blown way up and out to form a screen.
It’s then mounted sideways in a cabinet, and your eyes stare at the front of it. That’s
what brought about the phrase, watching the tube. And it also explains the name of this
site. Because it has no air inside, it has to be
pretty strong to counteract the force of the atmosphere always trying to crush it. That’s
why larger tube TV’s are so heavy--the glass needs to be quite thick on larger sets. Most
of the tube is empty space with the meat and potatoes being at the very back. Here you’ll
find the awesomely named electron gun. This component generates a stream of electrons
and they are shot straight out to the front of the tube. The flyback transformer generates
an extremely high voltage in the anode to attract the electrons to the front of the
screen. Coating the inside surface of the tube is a special powder known as a phosphor.
When the electrons sent from the gun hit the phosphor, it gets all excited and emits light,
via fluorescence, but only in the spot the electrons are hitting it. Here’s a working CRT with one of the critical
components to television removed because we haven’t gotten that far. Don’t worry,
we’ll get there. So, the CRT is doing a bang-up job producing a stream of electrons
and they’re going straight to the front of the screen, and colliding with it to make
it glow. And, this is the result. So fascinating. But hang on, there’s more to it. The vast
majority of the signal coming into this television is simply telling it how bright to make this
point of light. Therefore, a signal that alternates between bright and dark will make this happen.
Amazing. That doesn’t do that much good. Ah, but you see, the point of light can be
moved. One of the things Crookes and others noticed
when mucking about with his tubes was that a magnetic field can bend an electron beam.
In other words, you can use a magnet to alter the path the beam makes through the tube.
Watch. Here’s an ordinary strong magnet used for a nametag. When I move it around
the neck of the tube, the point of light moves around the screen, also. Mind bending, more
like beam bending, amiright? So then, here comes the other bit. This little
bundle of wires is called the deflection yoke. This is responsible for moving the beam really
really quickly, and fooling your eyes. The yoke is made of two electromagnets that surround
the neck of the tube, and they work together to move the electron beam around in a set
pattern. It does this by creating a fairly strong magnetic field which will deflect the
beam’s path. First, I’ll turn on the horizontal deflector. Now, rather than a point of light,
we see a line. This line is being drawn on the screen thousands of times a second, way
too fast for your eyes to notice. Just like the POV effect from the toy, if we carefully
control how bright this line is as it moves left and right we can create patterns in the
line like this. But the yoke contains another magnet that
can move the beam up and down. Let me switch to that one. We now have a vertical line being
drawn on the screen, and we can control its intensity just like the horizontal line to
draw patterns. This vertical movement happens much more slowly than the horizontal movement,
with the line only being drawn 60 times in a second. Now, since we can point the beam
from left to right, as well as up and down, we can point it anywhere we want on the screen.
Let’s turn on both electromagnets at the same time. We now have an image on the whole
screen. Pretty neat, huh? By carefully controlling the intensity of the beam over time, we can
create a complete image. If you look really closely at a black and
white television, you won’t find pixels. Rather, you’ll find lines. See, the image
is made of lines, in fact there are roughly 525 lines that make up the NTSC signal, and
about 480 are visible on the screen. The deflection yoke is making a pattern on the screen called
a raster, and in NTSC countries, it’s drawn on the screen 60 times a second. There’s
a bit of a trick, though, because the screen is only COMPLETELY redrawn 30 times a second.
See, as the raster is drawn, only every other line is filled in. This is called a field,
and it’s the principle behind interlaced video. That’s the i in 1080i. Because not-a-lot
of bandwidth is available, the whole screen can only reasonably be filled in 30 times
a second, but this would be noticeable as flicker and could give many people headaches.
By skipping every other line and then repeating the scan to fill in the rest, the screen is
drawn from top to bottom 60 times a second, which was too fast for most people to perceive
flickering. Also, it allowed for smoother motion, with the caveat that fast-moving objects
would have less detail as only every-other line is filled in with each field (however
that never proved to be a huge concern as it’s hard to see detail in fast moving objects,
anyway.) Side-note: It’s no coincidence that the
60 hz refresh rate matches the frequency of the AC electricity sent into homes, as the
60 hz sine wave coming from the socket powering the TV made for a convenient timing reference
for vertical scanning. PAL countries, which have 50 hz electricity, have a television
frame rate of 25 frames per second interlaced, with a scanning refresh rate of 50 hz. So,
tv framerates are what they are because convenience. So now that we have the means to generate
this raster, well how does that make a picture? Well, it’s just like the POV effect from
the toy, only it’s a helluva lot faster and the light moves in two dimensions. Let’s
slow down time and see how the TV builds an image. Let’s say we want to show this on
the screen. At the start of a field, the deflection yoke is pointing the electron beam at the
top left of the screen. As it moves towards the right, the beam changes its intensity
along with how bright the image should be, so at a point along the line that’s bright,
it produces a lot of electrons, and thus that point on the screen glows brightly. Dark parts
send little to no electrons. When the beam gets to the end of the line, the deflection
yoke almost instantly pulls it back to the left-hand side and starts the next line. But
remember, it skipped a line. This process repeats until it reaches the bottom of the
screen. Then the yoke flings the beam back to the top, and we start again filling in
the alternate lines. This happens way too fast for us to notice it, so it appears like
a fully illuminated screen. One thing to note is that the vertical deflection
isn’t happening in steps. Rather it’s a constant downward motion. This means that
the horizontal lines are actually slightly slanted downward to the right. To counteract
this, the deflection yoke is mounted to the tube ever so slightly crooked, so the lines
drawn on the screen are actually level. The constant downward travel is also how the interlacing
is accomplished. The next line will start at the same height as the end of the first,
which creates a gap. You may remember an extremely high-pitched
noise coming from a TV set whenever it was turned on. This noise actually came from the
deflection yoke and the electronics that drive it. In NTSC televisions, the horizontal deflection
occurred 525 times per frame, and there are 30 frames in a second, which means the electron
beam is being deflected left-and-right 15,750 times per second. In PAL countries, the framerate
is only 25 frames per second, but 625 lines are drawn with each frame, which works out
to 15,625 deflection per second. The yoke and the flyback transformer, along with some
other components, actually vibrate at this frequency ever so slightly, which produces
audible noise. This is what it sounds like. Adults over the age of 25 or so can’t hear
this sound, as it’s at the upper limit of human hearing range, which gradually diminishes
with age. So for those viewers, sorry. When it comes to actually producing an image,
the trickiest part is matching that raster to an incoming television signal. To help
with this, the TV signal contains triggers which assist the TV in grabbing hold of the
image and keeping it in one place. This Sony TV is correctly tuned to channel
3, which is currently displaying this video that you’re hearing right now. But there’s
absolute nonsense on the screen. What gives? Well, the TV is generating its own raster,
and right now it’s not synchronized with the raster coming into the TV. You’re seeing
all of the image, but each part is in the wrong place because it’s not lined up. Here,
to show you what the TV’s looking for, let’s fade to white. You’ll notice that there
are a ton of black gaps swirling around what should be an entirely white screen. These
gaps are the horizontal blanking intervals between individual scanlines. When horizontal
hold is properly adjusted, electronics in the TV can see these gaps and line them up. Hold up, how can the set tell the difference
between the blanking intervals and a black spot on the screen? Well, it can tell them
apart because the blanking intervals are actually BLACKER THAN BLACK. No, really. Here’s a
one line of a television signal drawn on a graph. These parts at the ends are the blanking
intervals between scan lines. They are the lowest parts of the graph because their amplitude
is near zero. Here is the actual start of the scan line. The higher the line goes, the
brighter that part of the scan line will be drawn on the screen. Makes sense, but black
is all the way up here. Television sets are calibrated to not fire the electron gun at
amplitudes at or below this amount, so to they eye, any amplitude below this point won’t
be visible, but the electronics can clearly tell blanking intervals from signals. The
blanking interval isn’t there just to provide a reference for the beginning and end of a
scan line, it’s also there prevent anything from being drawn on the screen as the deflection
yoke sweeps the electron beam back to the left-hand side before the start of the next
line. The TV just has to line these low points up by catching them at the beginning of each
scan line, and then they’ll fall into the TV’s own raster. Everything is hunky dory. So then, when I adjust the horizontal hold,
you can see that this moves the blanking intervals closer to each other, and eventually, the
image snaps into place. well, sort of. Now the image is rolling, it’s continually moving
downwards. Ah, see, we have only synchronized the television's raster with the horizontal
components of the signal. Without a reference as to what starts a field scan, the pictures
just gonna roll around like this. See that hunk of black between my head and my waist?
That’s the vertical blanking interval, which is little more than a bunch of empty scan
lines. Just like the horizontal intervals, it allows the deflection yoke time to get
back to the top of the field. Again, this is BLACKER THAN BLACK, and it allows the television
to hold onto the start of each field and keep them in one place. The vertical blanking interval
also contains some special pulses to differentiate between the odd and even numbered fields. So, i’ll adjust the vertical hold, and eventually,
the frame snaps into place, and you get a truly stable image. Very intentionally, the
CRT is scanning outside the borders of the face of the tube. This is called overscan,
and it’s done to hide the blanking intervals as well as just ensure the whole screen is
being used. On this set, you can see how the scan extends beyond the tube itself when looking
from behind. This unseen overscan area was used later to add closed captioning into television
broadcasts. On one of the lines that make up the VBI, alternating white-black bits created
a barcode of sorts that contained digital text information. A decoder inside the television
set could read this data from that line, and when enabled place text graphics on top of
the image. I think that’s pretty friggin nifty. As far as audio, well that’s really simple.
That’s nothing more than simple FM radio built into the TV, and each channel has an
audio signal being transmitted at a set offset frequency from the video source. Since the
signals are transmitted together, they are always in sync. So, that’s how these old things work. But
there’s a lot more to explore. For one, how did television cameras actually create
the signal that drives this TV? And who were the people responsible for inventing it? What about color? We’ll
explore that in a later episode, along with the precursor to CRT television, mechanical
television, so be sure to subscribe to Technology Connections. If you liked this video, I humbly
ask that you hit that like button and maybe leave a comment. I’m doing my best to keep
videos like this headed your way. Thanks for watching!