1
00:00:00,080 --> 00:00:04,280
This exploded machine is capable of a marvelous
feat.

2
00:00:04,280 --> 00:00:09,500
These parts, when assembled correctly, are
able to interact with this silver plastic thing

3
00:00:09,500 --> 00:00:13,580
and make melodious sounds for you to
enjoy in the highest of fidelities.

4
00:00:14,500 --> 00:00:15,820
Am I being vague?

5
00:00:15,830 --> 00:00:16,830
Maybe.

6
00:00:16,830 --> 00:00:19,779
Well, more specifically, this is a compact
disc player!

7
00:00:19,779 --> 00:00:25,100
The innards of a Magnavox CDB 260, made in
August of 1988.

8
00:00:25,100 --> 00:00:28,790
From what I can gather this is a pretty basic
CD player, even for the time.

9
00:00:28,790 --> 00:00:31,010
But that’s fine, in fact it might even be
better.

10
00:00:31,010 --> 00:00:32,300
So, in the last video,

11
00:00:32,300 --> 00:00:32,800
*ding*

12
00:00:32,800 --> 00:00:36,200
we talked about
how the data is stored on a compact disc.

13
00:00:36,210 --> 00:00:40,920
Just below the top surface of the disc, millions
of little bumps are found in the polycarbonate,

14
00:00:40,920 --> 00:00:46,050
and the aluminum coating will reflect light
like a mirror back from whence it came.

15
00:00:46,050 --> 00:00:51,480
These bumps, from the bottom side, form pits
and lands and they are what encodes the data.

16
00:00:51,480 --> 00:00:55,920
So the first thing any CD player will need
to do is to shine a light on the CD, a light

17
00:00:55,920 --> 00:01:01,719
so finely focused that it’s shining mostly
on only one track, and it needs a way to detect

18
00:01:01,719 --> 00:01:04,170
the change in brightness caused by the pits.

19
00:01:04,170 --> 00:01:06,259
That’s what this guy does, here.

20
00:01:06,259 --> 00:01:08,939
This is the laser and lens assembly.

21
00:01:08,939 --> 00:01:12,689
To understand what it does, it’s best to
take a look at it from this angle.

22
00:01:12,689 --> 00:01:13,689
Excellent.

23
00:01:13,689 --> 00:01:17,869
OK, so notice that there are two circuit boards
at right angles to each other.

24
00:01:17,869 --> 00:01:21,319
This one to the side is where you will find
the laser diode.

25
00:01:21,319 --> 00:01:24,359
This is what creates the beam of light that
will read the disc.

26
00:01:24,359 --> 00:01:28,880
The light from the laser diode gets shot straight
to the left, and a prism, right about here,

27
00:01:28,880 --> 00:01:32,780
will cause the light to make a 90 degree turn,
and end up going...

28
00:01:32,780 --> 00:01:34,189
up through the lens, and

29
00:01:34,189 --> 00:01:35,689
onto the disc.

30
00:01:35,689 --> 00:01:39,749
Because the disc has a mirrored surface, that
light is gonna come straight back down through

31
00:01:39,749 --> 00:01:40,899
the lens.

32
00:01:40,899 --> 00:01:43,469
And here’s where the other circuit board
comes in.

33
00:01:43,469 --> 00:01:46,959
Some of the light being reflected from the
CD will go straight through the prism and

34
00:01:46,960 --> 00:01:48,320
end up at the bottom.

35
00:01:48,320 --> 00:01:52,319
Now you can’t see them because this is sealed
and I don’t really want to break it but

36
00:01:52,319 --> 00:01:57,939
inside, on top of the circuit board, there
are 4 photodiodes arranged in a 2 x 2 grid.

37
00:01:57,939 --> 00:01:59,749
Why are there 4, you ask?

38
00:01:59,749 --> 00:02:03,789
Well, not only do the photodiodes need to
react to the changes in brightness of the

39
00:02:03,789 --> 00:02:08,479
reflected light in order to, for lack of a
better word, detect the zeros and ones, but

40
00:02:08,480 --> 00:02:11,420
they also forms the basis of the focus and
tracking system.

41
00:02:11,760 --> 00:02:12,460
A ha!

42
00:02:12,470 --> 00:02:13,969
A new thing to unpack.

43
00:02:13,969 --> 00:02:15,829
No CD is perfect.

44
00:02:15,829 --> 00:02:20,390
All of them are going to have some deviations
from perfectly flat, and they’re also probably

45
00:02:20,390 --> 00:02:25,120
not going to be held exactly perfectly centered,
so the track of pits will waver like this

46
00:02:25,120 --> 00:02:26,420
off-center record.

47
00:02:26,420 --> 00:02:30,890
To stay on the same track of pits and lands,
some mechanism will need to be able to move

48
00:02:30,890 --> 00:02:36,049
the lens closer to and farther away from the
disc, in other words stay focused on it, and

49
00:02:36,049 --> 00:02:40,730
there will also need to be a way to move the
lens back and forth as the spiral track wobbles.

50
00:02:40,730 --> 00:02:42,689
In other words, it needs to track it.

51
00:02:42,689 --> 00:02:46,440
Philips’ solution to the problem of focus
and tracking is the less common one, but it

52
00:02:46,440 --> 00:02:49,530
is so much more elegant than Sony’s in my
opinion.

53
00:02:49,530 --> 00:02:53,260
I’ll show you Sony’s system later on,
but if you didn’t see my last video, I showed

54
00:02:53,260 --> 00:02:56,940
you the disc tray of this machine and suggested
it hinted at the difference.

55
00:02:56,940 --> 00:02:58,939
So, first the easy bit.

56
00:02:58,939 --> 00:03:03,480
To maintain focus on the disc, the lens is
floating, and you can kinda see that there’s

57
00:03:03,480 --> 00:03:06,520
a coil of wire that can push it up or pull
it down,

58
00:03:06,520 --> 00:03:09,379
pretty much exactly like a loudspeaker driver.

59
00:03:09,379 --> 00:03:13,680
The low mass of the lens allows it to react
practically instantly to changes in the distance

60
00:03:13,680 --> 00:03:18,980
between the CD and the lens, so even a badly
warped CD won’t fool the CD player.

61
00:03:19,320 --> 00:03:21,280
Alright, we have the Y axis accounted for.

62
00:03:21,290 --> 00:03:25,420
But how do we move the lens left and right,
in order to account for a wibbly wobbly track?

63
00:03:25,420 --> 00:03:29,609
Well, notice that the lens is on an arm that’s
surrounding a metal disc.

64
00:03:29,609 --> 00:03:33,299
This allows the lens to rotate around the
disc, and you’ll notice that this disc is

65
00:03:33,299 --> 00:03:34,749
a magnet.

66
00:03:34,749 --> 00:03:39,019
Flip this upside down and you'll find another
coil of wire, surrounding the magnet.

67
00:03:39,019 --> 00:03:43,590
This swing-arm design is remarkably similar
to how hard drives move their read/write heads

68
00:03:43,590 --> 00:03:49,239
with a voice coil, and it’s this that sets
the Philips design far apart from Sony’s.

69
00:03:49,239 --> 00:03:53,650
Because the lens can move vertically, and
its platform can move horizontally, constant

70
00:03:53,650 --> 00:03:55,870
focus and tracking can be maintained.

71
00:03:55,870 --> 00:03:59,810
Now, I’ll reattach the laser mechanism to
the spindle motor assembly.

72
00:03:59,810 --> 00:04:03,810
It has ball bearings on the top and bottom,
and when this bracket is in place the arm

73
00:04:03,810 --> 00:04:07,099
essentially floats and moves nearly effortlessly.

74
00:04:07,099 --> 00:04:11,779
This mechanism is fastened to this assembly,
which also operates the disc tray.

75
00:04:11,779 --> 00:04:15,420
The disc tray is made in two pieces, and it’s
a rather unique design.

76
00:04:15,420 --> 00:04:19,810
The bottom half sort of drops down when it
is inserted into the player, and you can sort

77
00:04:19,810 --> 00:04:21,780
of move it even while it’s open.

78
00:04:21,780 --> 00:04:26,130
When it is holding a CD to be played, it is
brought into the player and as it reaches

79
00:04:26,130 --> 00:04:31,389
the spindle, the lower half drops down and
this spring-loaded clamp will rest on it.

80
00:04:31,389 --> 00:04:35,110
The reason why I said the tray might give
a hint to the Philips pickup system, is that

81
00:04:35,110 --> 00:04:39,300
it has a curved cut out to accommodate the
path of the laser’s swing.

82
00:04:39,300 --> 00:04:43,870
That’s a tell-tale sign that the CD player
has a swing-arm laser mechanism.

83
00:04:43,870 --> 00:04:46,920
Now with the disc held in place, the spindle
motor spins the disc,

84
00:04:46,920 --> 00:04:48,820
and the laser will attempt to read it.

85
00:04:48,820 --> 00:04:53,380
And now we go back to the photodiodes, because
this is the really genius bit.

86
00:04:53,389 --> 00:04:58,330
If the disc is focused correctly and it’s
being tracked correctly, all four photodiodes

87
00:04:58,330 --> 00:05:01,449
will receive the same relative intensity of
light.

88
00:05:01,449 --> 00:05:06,180
This is because the spot of light that’s
hitting the array is a perfect circle.

89
00:05:06,180 --> 00:05:11,490
Due to the optics of the lens and prism, an
out-of-focus beam will morph into an ellipse,

90
00:05:11,490 --> 00:05:15,849
and depending on which way it’s oriented,
the focus is either near or far.

91
00:05:15,849 --> 00:05:20,380
Additionally, tracking errors either left
or right will push the spot...

92
00:05:20,380 --> 00:05:21,789
left or right.

93
00:05:21,789 --> 00:05:26,699
By comparing the relative intensities between
these four photodiodes, the player can determine

94
00:05:26,699 --> 00:05:29,490
what corrective action it needs to take.

95
00:05:29,490 --> 00:05:31,110
So what does the comparing?

96
00:05:31,110 --> 00:05:35,410
Well, when assembled, the ribbon cable which
carries power to the laser unit and returns

97
00:05:35,410 --> 00:05:39,039
signals from the diodes plugs in here on the
circuit board.

98
00:05:39,039 --> 00:05:43,069
Now if we flip it over, we find two rather
large chips.

99
00:05:43,069 --> 00:05:47,069
If we follow the traces from the plug where
the ribbon cable went in, there are four that

100
00:05:47,069 --> 00:05:51,020
go right to this chip, labeled TDA8808T.

101
00:05:51,020 --> 00:05:52,240
What does this chip do?

102
00:05:52,250 --> 00:05:56,440
Well, it’s a photodiode signal processor
for compact disc players!

103
00:05:56,440 --> 00:06:00,419
Its data sheet tells us that it is “a bipolar
integrated circuit designed for use in compact

104
00:06:00,419 --> 00:06:03,030
disc players with a single spot read-out system.

105
00:06:03,030 --> 00:06:07,710
It amplifies the photo-diode signals and processes
the error signals for the focus- and radial

106
00:06:07,710 --> 00:06:09,140
control network.”

107
00:06:09,140 --> 00:06:10,180
Well ain’t that handy!

108
00:06:10,560 --> 00:06:14,880
So this chip not only handles the focus, but
it also amplifies the actual signal coming

109
00:06:14,880 --> 00:06:16,360
from the photodiodes.

110
00:06:16,360 --> 00:06:17,320
Neat.

111
00:06:17,320 --> 00:06:19,060
But it doesn’t do tracking.

112
00:06:19,060 --> 00:06:21,560
But if you noticed this strikingly similar
chip,

113
00:06:21,560 --> 00:06:24,540
this one a TDA8809T,

114
00:06:24,540 --> 00:06:27,000
you might have guessed that this guy does that.

115
00:06:27,009 --> 00:06:31,639
But it only needs two inputs because it just
wants the sums of these two diodes, which

116
00:06:31,639 --> 00:06:34,800
handily the other chip has generated for it.

117
00:06:34,800 --> 00:06:39,110
These two chips feed their calculated corrections
up to this chip, which actually handles the

118
00:06:39,110 --> 00:06:43,770
voltage to drive the deflection coils for
the lens’s vertical movement and the arm’s

119
00:06:43,770 --> 00:06:45,100
horizontal movement.

120
00:06:45,100 --> 00:06:47,880
OK, there’s one last thing that the first
chip handles.

121
00:06:47,880 --> 00:06:52,220
It has some automatic gain control circuitry
onboard to actually generate a binary stream

122
00:06:52,220 --> 00:06:55,930
from what is the analog signal coming from
the photodiodes.

123
00:06:55,930 --> 00:07:00,410
Remember, the pits and lands represent ones
and zeroes, but we’re still in the analog

124
00:07:00,410 --> 00:07:03,710
realm, reading voltage from the four photodiodes.

125
00:07:03,710 --> 00:07:08,370
So in essence, this chip is an analog-to-digital
converter, creating an actual bit-stream for

126
00:07:08,370 --> 00:07:10,760
the following components to utilize.

127
00:07:10,760 --> 00:07:15,250
If we follow this trace, from the pin which
is labeled “TO DECODER” on the datasheet,

128
00:07:15,250 --> 00:07:18,500
we’ll find it heading over to this Motorola
chip.

129
00:07:18,500 --> 00:07:22,130
Now this one was a bit of a mystery for a
while, in fact it tripped me up so much that

130
00:07:22,130 --> 00:07:25,600
what you’re looking at now happened a day
after what happened before.

131
00:07:25,600 --> 00:07:30,080
Eh, what exactly the motorola chip does is
not clear, as I couldn’t find a datasheet

132
00:07:30,080 --> 00:07:34,030
for it, but it appears to be the main processor
for the machine.

133
00:07:34,030 --> 00:07:37,939
If you look at these pins here, you’ll see
that they head to these connectors which connect

134
00:07:37,939 --> 00:07:40,289
the main board to the front panel.

135
00:07:40,289 --> 00:07:44,050
These are likely inputs for the buttons on
the front of the machine, as well as the outputs

136
00:07:44,050 --> 00:07:48,220
for driving the LEDs on the display, though
there is probably some communication protocol

137
00:07:48,220 --> 00:07:52,409
for that task, given there are definitely
not enough pins to be driving each segment

138
00:07:52,409 --> 00:07:54,370
and each indicator directly.

139
00:07:54,370 --> 00:07:58,560
What is clear is that some of these traces
head from the Motorola chip up to the tracking

140
00:07:58,560 --> 00:08:03,060
chip, which would make sense because if the
processor handles button inputs, it probably

141
00:08:03,060 --> 00:08:06,940
also talks to the tracking chip for navigating
around the disc.

142
00:08:06,940 --> 00:08:13,310
But what’s odd is that the datasheet for
the TDA8808T said that pin 10, which is definitely

143
00:08:13,310 --> 00:08:17,370
this trace, I’ve checked with an ohmmeter,
goes to the decoder.

144
00:08:17,370 --> 00:08:21,440
And this other big chip, from Philips, is
called the decoder chip.

145
00:08:21,880 --> 00:08:23,860
Hmm, what’s going on?

146
00:08:23,870 --> 00:08:25,659
To be honest, I’m not entirely sure.

147
00:08:25,659 --> 00:08:28,220
I can make a few educated guesses, though.

148
00:08:28,220 --> 00:08:32,010
First, if you haven’t already spotted it,
you’ll see that the trace from pin 10 on

149
00:08:32,010 --> 00:08:36,840
the focus chip branched off here, and it also
connects to the decoder chip through a resistor

150
00:08:36,840 --> 00:08:38,400
on the top side of the board.

151
00:08:38,400 --> 00:08:42,920
So the decoder chip does have a connection
to the laser diode’s output, but whether

152
00:08:42,920 --> 00:08:46,930
or not this is where it’s getting the data
it needs is uncertain.

153
00:08:46,930 --> 00:08:51,450
There are direct connections between the processor
chip and the decoder chip, so it could be

154
00:08:51,450 --> 00:08:55,890
that this trace here is providing the decoder
with some sort of redundancy or perhaps even

155
00:08:55,890 --> 00:08:58,190
something as simple as a clock signal.

156
00:08:58,190 --> 00:09:02,270
One thing that is certain is that the decoder
chip has a number of connections to this chip

157
00:09:02,270 --> 00:09:03,600
right next to it.

158
00:09:03,600 --> 00:09:08,170
This chip is a RAM chip, which the player
likely uses to undo the cross-interleaved

159
00:09:08,170 --> 00:09:10,500
reed-solomon coding used in the disc.

160
00:09:10,500 --> 00:09:15,400
Remember, the data is spread about on the
disc in addition to having parity bytes mixed in,

161
00:09:15,400 --> 00:09:19,250
so the player needs somewhere to dump
the raw datastream in the short term for some

162
00:09:19,250 --> 00:09:21,280
data reorganization and decoding.

163
00:09:22,040 --> 00:09:25,420
Ah right, that’s probably why it’s called
the decoder chip.

164
00:09:25,430 --> 00:09:30,410
If I were to make a guess, which I am doing
right now, I’d say that the processor contains

165
00:09:30,410 --> 00:09:35,520
the lookup table and is demodulating the eight-to-fourteen modulation used on the disc.

166
00:09:35,520 --> 00:09:38,630
Then it generates the raw stream of 8 bit
words.

167
00:09:38,630 --> 00:09:43,350
From that point, it sends the demodulated,
but still CIRC encoded, data to the Philips

168
00:09:43,350 --> 00:09:44,600
decoder chip.

169
00:09:44,600 --> 00:09:49,460
The Philips chip dumps the raw data into RAM
so it can reorganize and decode the CIRC error

170
00:09:49,460 --> 00:09:53,680
correction and recover the actual audio samples
from the datastream.

171
00:09:53,680 --> 00:09:55,620
Then it can be sent to the DAC.

172
00:09:55,620 --> 00:09:59,580
But the processor is probably also handling
the logic of the disc.

173
00:09:59,580 --> 00:10:04,180
There’s some low-level stuff in the datastream
that can be seen before CIRC error correction

174
00:10:04,180 --> 00:10:05,290
is decoded.

175
00:10:05,290 --> 00:10:06,920
For example, the subcode.

176
00:10:07,340 --> 00:10:08,140
Subcode?

177
00:10:08,640 --> 00:10:10,710
Hmm, have we overlooked something?

178
00:10:10,710 --> 00:10:12,100
Well kinda.

179
00:10:12,100 --> 00:10:16,720
We’ve talked in the past about the table
of contents and the timecode, but where are

180
00:10:16,720 --> 00:10:18,490
these along this disc?

181
00:10:18,490 --> 00:10:23,640
If the CD was just a raw stream of audio samples,
how is the timecode being stored?

182
00:10:23,640 --> 00:10:28,580
Well, earlier I had said that 192 bits of
audio samples are stored in a frame,

183
00:10:28,580 --> 00:10:30,780
consisting of 24 bytes.

184
00:10:30,780 --> 00:10:33,200
That’s not strictly true, though.

185
00:10:33,200 --> 00:10:35,420
The frames are actually 33 bytes.

186
00:10:35,420 --> 00:10:41,170
Now, if you’ve been keeping track, reed-solomon
code produces a parity byte for every 3 audio bytes,

187
00:10:41,170 --> 00:10:45,880
which would mean that in 24 bytes of
audio, there should be 8 redundant parity bytes.

188
00:10:45,880 --> 00:10:48,280
That would bring the frame to 32 total bytes.

189
00:10:48,280 --> 00:10:49,910
But there’s an extra one.

190
00:10:49,910 --> 00:10:51,720
Ooh, the plot thickens!

191
00:10:51,720 --> 00:10:55,910
That is the subcode byte, of which there is
one in each frame.

192
00:10:55,910 --> 00:11:01,370
98 frames comprise the timecode frame, sometimes
also called a sector, though the use of the

193
00:11:01,370 --> 00:11:06,920
term sector was added in the CD-ROM days and
is not part of the original Red Book standard.

194
00:11:06,920 --> 00:11:09,380
However I’m going to continue to use the
word Sector

195
00:11:09,380 --> 00:11:13,200
because “Frame” and “Timecode Frame” are too similar.

196
00:11:13,200 --> 00:11:19,940
These combine to create 1/75th of a second
of audio, so each second contains 75 timecode

197
00:11:19,940 --> 00:11:25,340
frames, or sectors, which comprise the most
specific time references there are on the CD.

198
00:11:25,340 --> 00:11:28,280
The way the subcode is handled is pretty ingenious.

199
00:11:28,280 --> 00:11:34,380
Though there is only one byte per frame, a
total of 96 frames have a usable byte per sector.

200
00:11:34,380 --> 00:11:38,240
The first two in a 98 frame sector are used
as a synchronization word.

201
00:11:38,240 --> 00:11:42,350
Now, you might have realized that 96 is divisible
by 8.

202
00:11:42,350 --> 00:11:46,730
So rather than just create one bitstream,
each subcode byte actually represents one

203
00:11:46,730 --> 00:11:49,670
bit of 8 subcode channels.

204
00:11:49,670 --> 00:11:53,460
Depending on which channel the player wants
to read, it will be looking for the nth bit

205
00:11:53,460 --> 00:11:56,000
in each subcode byte, and stringing them together.

206
00:12:01,420 --> 00:12:04,680
And the crucial one for the timecode is the
Q bit.

207
00:12:04,690 --> 00:12:08,820
And even within just that datastream, there’s
some extra specificity.

208
00:12:08,820 --> 00:12:13,620
The first bits of the Q channel are actually
signifying specific information about the

209
00:12:13,620 --> 00:12:19,870
disc, and it’s the 72 bits before the last
16 that we need to look into.

210
00:12:19,870 --> 00:12:24,120
But even then, there are different modes that
can be selected within the datastream of the

211
00:12:24,120 --> 00:12:27,690
Q channel, so there was a lot of thinking
going into this.

212
00:12:27,690 --> 00:12:32,060
I’ve linked the Wiki article on compact
disc subcode as it goes over this much better

213
00:12:32,060 --> 00:12:33,060
than I can.

214
00:12:33,060 --> 00:12:37,710
This explains the rather elegant way in which
a CD player can find tracks on a disc.

215
00:12:37,710 --> 00:12:41,990
The lead-in contained the table-of-contents
(which incidentally is sent in the Q channel

216
00:12:41,990 --> 00:12:45,520
subcode as well) and the CD player stored
in memory.

217
00:12:45,520 --> 00:12:49,170
It now knows where each of the tracks are
by their time code.

218
00:12:49,170 --> 00:12:53,260
So when you select Track 10, which the Table
of Contents told the player can be found

219
00:12:53,260 --> 00:12:59,220
at this time code, it will first send its
laser to around where that timecode probably is

220
00:12:59,220 --> 00:13:03,220
(after all the physical characteristics are nearly
identical from CD to CD)

221
00:13:03,220 --> 00:13:06,460
and then it simply mutes the audio while it looks.

222
00:13:06,460 --> 00:13:10,700
By reading the Q channel datastream it knows
exactly what it’s looking at, and it will

223
00:13:10,700 --> 00:13:16,500
move inward or outward in fine steps until
it sees exactly the timecode that it’s looking for.

224
00:13:16,500 --> 00:13:19,160
Then it will unmute the audio, and play the
song.

225
00:13:19,160 --> 00:13:23,600
In a rather smart move, Sony and Philips minded
their P’s and Q’s by making these the

226
00:13:23,600 --> 00:13:27,360
only two subcode channels used in the Red
Book standard.

227
00:13:27,360 --> 00:13:31,720
6 entire data streams were left unused for
future extensions of the standard.

228
00:13:31,720 --> 00:13:36,650
And each of these streams, when discounting
the two synchronization bits, can produce

229
00:13:36,650 --> 00:13:38,990
7.2 kilobits per second.

230
00:13:38,990 --> 00:13:43,360
Now of course that’s not a lot, but that’s
900 characters of text per second.

231
00:13:43,360 --> 00:13:48,380
Multiply that by 6 channels and you have
5.4 kilobytes per second of untapped data

232
00:13:48,380 --> 00:13:52,400
available, or about 24 megabytes total
for a 74 minute CD.

233
00:13:52,400 --> 00:13:55,800
In later years this would be used to produce
some nifty things.

234
00:13:55,800 --> 00:13:57,530
Which we’ll get to, don’t worry.

235
00:13:57,530 --> 00:14:01,480
But just imagine for a moment how crazy forward
thinking that was.

236
00:14:01,480 --> 00:14:06,800
In any CD, there were more than 16 3.5 inch
floppy diskettes worth of data

237
00:14:06,800 --> 00:14:09,120
sitting completely unused.

238
00:14:09,120 --> 00:14:13,340
And in 1982, these things weren’t even in
commercial production yet, plus they wouldn’t

239
00:14:13,340 --> 00:14:17,080
hit 1.44 megabytes in size until 1986.

240
00:14:17,080 --> 00:14:21,150
In a time when hard drives for computers were
around 10 megabytes, if you even had one in

241
00:14:21,150 --> 00:14:27,390
your PC, the idea of having 24 megabytes in
one place, let alone 24 megabytes of what

242
00:14:27,390 --> 00:14:33,410
amounted to unneeded and unused data on a
sound storage medium, was quite remarkable.

243
00:14:33,410 --> 00:14:36,760
But remember, no one was thinking of that
at all.

244
00:14:36,760 --> 00:14:41,290
The CD-ROM was some years away, though it
does strike me as a bit of a waste that no

245
00:14:41,290 --> 00:14:45,620
one developed computer peripherals that could
interface with the subcode-out jacks found

246
00:14:45,620 --> 00:14:50,500
in some CD players to get programs or data
off of audio CDs.

247
00:14:50,500 --> 00:14:51,960
That would have been kinda nifty.

248
00:14:51,960 --> 00:14:56,090
Now of course, this just an educated guess,
but the decoder chip probably isn’t even

249
00:14:56,090 --> 00:14:57,500
looking at the subcode.

250
00:14:57,500 --> 00:15:01,430
The motorola chip is probably the only thing
paying attention to it, and in fact it might

251
00:15:01,430 --> 00:15:06,650
even strip that out before sending the CIRC
encoded audio data to the decoder chip.

252
00:15:06,650 --> 00:15:10,670
But don’t assume the decoder chip has an
easy job; it’s got a lot of math to work

253
00:15:10,670 --> 00:15:13,160
out to undo the reed-solomon code.

254
00:15:13,160 --> 00:15:18,440
Plus, keep in mind that there are 16 bit audio
samples here, but we’re dealing with 8 bit

255
00:15:18,440 --> 00:15:23,230
words, so it needs to decode what exactly
each sample is and send this and this alone

256
00:15:23,230 --> 00:15:24,570
to the DAC.

257
00:15:24,570 --> 00:15:27,150
Which is definitely the role of this chip.

258
00:15:27,150 --> 00:15:30,900
Flipping the board back over, if we follow
the traces we find we’ve got a couple of

259
00:15:30,900 --> 00:15:33,840
connections to this little 8 pin chip.

260
00:15:33,840 --> 00:15:36,610
And believe it or not, that is the DAC.

261
00:15:36,610 --> 00:15:42,660
This diminutive TDA1543 is all that we need
to turn ones and zeros into music.

262
00:15:42,660 --> 00:15:46,480
Just provide it with a clock signal, your
data stream, a power and ground, and you’ll

263
00:15:46,480 --> 00:15:49,770
get left and right channel outputs on pins
6 and 8.

264
00:15:49,770 --> 00:15:54,630
Now the datasheet says that pin 3 is the data
input, and if we follow the trace back, sure

265
00:15:54,630 --> 00:15:56,810
enough this heads to the Philips chip.

266
00:15:56,810 --> 00:15:59,320
So let’s put this entirely back together.

267
00:15:59,320 --> 00:16:03,660
The circuit board is held down with only one
screw, as it also gets support from screws

268
00:16:03,660 --> 00:16:04,830
on the rear.

269
00:16:04,830 --> 00:16:09,090
The CD mechanism rests above the circuit board
here, and we need to plug in the ribbon cable

270
00:16:09,090 --> 00:16:13,300
for the laser mechanism, as well as the spindle
motor and the disc tray motor.

271
00:16:13,300 --> 00:16:17,650
The disc tray slides back in place like this,
though you have to kind of hold up the moveable

272
00:16:17,650 --> 00:16:22,070
piece as you do so, and now we’re ready
to put the faceplate back on.

273
00:16:22,070 --> 00:16:25,890
Simply plug in the connectors to its circuit
board for the display and buttons, and don’t

274
00:16:25,890 --> 00:16:29,640
forget the headphone jack, and now we just
have to put on this plastic thing which

275
00:16:29,640 --> 00:16:33,980
connects the power button to the actual power
switch, which is way back here for some reason,

276
00:16:33,980 --> 00:16:36,080
and now it’s more or less assembled.

277
00:16:36,080 --> 00:16:38,150
Well, except for the cover.

278
00:16:38,150 --> 00:16:40,051
So now let’s play a CD.

279
00:16:40,051 --> 00:16:43,380
I turn the player on, and you’ll see that
it performs a check of whether or not a CD

280
00:16:43,380 --> 00:16:44,590
is in it.

281
00:16:44,590 --> 00:16:47,050
It’s empty, so it enters a standby mode.

282
00:16:47,050 --> 00:16:51,190
I press the open/close button and the motor
pushes the tray open.

283
00:16:51,190 --> 00:16:55,230
Now I deposit my disc of choice, and press
the open/close button again.

284
00:16:55,230 --> 00:16:59,880
Once the tray has closed, the focus chip performs
its initial focus routine, where it sweeps

285
00:16:59,880 --> 00:17:05,120
the lens from its lowest to highest position
in order to detect whether or not a disc is there.

286
00:17:05,120 --> 00:17:10,029
Once it sees the disc it locks focus on it,
and as the disc is spun by the spindle motor,

287
00:17:10,029 --> 00:17:14,730
it begins analyzing the ratio of stimulation
between the two pairs of photodiodes to maintain

288
00:17:14,730 --> 00:17:17,100
focus even on badly warped discs.

289
00:17:17,100 --> 00:17:21,470
Now, the tracking chip kicks in, nudging the
laser left and right at the request of the

290
00:17:21,470 --> 00:17:24,009
Motorola chip, which is working for the Philips
chip.

291
00:17:24,009 --> 00:17:27,480
The tracking chip is helping to look for the
table of contents.

292
00:17:27,480 --> 00:17:32,559
Once it’s seen it, it locks onto it and
the processing chips load the TOC into memory.

293
00:17:32,559 --> 00:17:35,430
At this point the player re-enters a standby
mode.

294
00:17:35,430 --> 00:17:39,580
If I had selected play, or pushed the tray
in rather than hitting the close button, it

295
00:17:39,580 --> 00:17:43,049
would have started playing immediately after
loading the TOC.

296
00:17:43,049 --> 00:17:45,269
In any case, it’s ready to play the disc.

297
00:17:45,269 --> 00:17:50,080
I hit PLAY, and the disc begins spinning while
the focus chip maintains focus and the tracking

298
00:17:50,080 --> 00:17:53,279
chip helps to look for the beginning of the
program area.

299
00:17:53,279 --> 00:17:57,370
After finding it, the tracking chip simply
locks on to the datastream, and the raw data

300
00:17:57,370 --> 00:17:59,490
is processed by these two chips.

301
00:17:59,490 --> 00:18:02,539
Again, exact specifics unknown here.

302
00:18:02,539 --> 00:18:06,919
After undoing the eight-to-fourteen modulation,
as well as reordering the data and filtering

303
00:18:06,919 --> 00:18:12,169
out the parity bytes from the CIRC error correction,
the player has now extracted the raw datastream

304
00:18:12,169 --> 00:18:14,070
of audio samples.

305
00:18:14,070 --> 00:18:17,960
These are sent to the DAC, which recreates
the original analog sound wave based on the

306
00:18:17,960 --> 00:18:20,730
principles of the Nyquist-Shannon Sampling
Theorem.

307
00:18:20,730 --> 00:18:25,049
If I want to select a later track using the
NEXT button, the Motorola chip will take over

308
00:18:25,049 --> 00:18:30,080
the tracking chip once more, nudging the laser
farther out while it looks in the Q channel subcode

309
00:18:30,080 --> 00:18:32,610
for the timecode of the next track.

310
00:18:32,610 --> 00:18:37,640
Once it sees it, it locks onto the datastream
again, and the audio is processed as normal.

311
00:18:37,640 --> 00:18:41,889
If I select a track that’s far from the
start, the laser flings itself out there,

312
00:18:41,889 --> 00:18:44,700
still looking in the Q channel, for the appropriate
track.

313
00:18:44,700 --> 00:18:49,860
The CD player will happily decode the audio
for you until it reaches the end of the last track.

314
00:18:49,860 --> 00:18:53,509
Now, it’s reading the lead-out, in which
the subcode will tell the player it’s reached

315
00:18:53,509 --> 00:18:55,820
the end of the disc, and it stops.

316
00:18:55,820 --> 00:18:59,160
Unless of course you’ve got repeat enabled,
in which case it’s back to the beginning

317
00:18:59,160 --> 00:19:01,080
to start the whole thing over again!

318
00:19:01,080 --> 00:19:03,660
Hooft, that was intense.

319
00:19:03,860 --> 00:19:04,960
Like the circus!

320
00:19:05,940 --> 00:19:09,059
Oh no, you’re seeing the CD player all disassembled
again!

321
00:19:09,059 --> 00:19:10,539
The continuity errors!

322
00:19:10,540 --> 00:19:11,539
The horrors!

323
00:19:11,780 --> 00:19:13,880
OK, I’ll cut back on the silliness.

324
00:19:13,890 --> 00:19:18,460
This was my best attempt at explaining how
this CD player works in the most comprehensive

325
00:19:18,460 --> 00:19:22,000
way possible, while trying not to get too
technical.

326
00:19:22,000 --> 00:19:26,500
Already I can tell that this is a longer-than-normal
video, so if you made it here, thanks!

327
00:19:26,500 --> 00:19:27,980
And congratulations!

328
00:19:27,980 --> 00:19:30,960
Soon, I’ll be taking this back apart--

329
00:19:30,960 --> 00:19:32,500
after
I reassemble it--

330
00:19:32,500 --> 00:19:33,780
and adding some test wires

331
00:19:33,780 --> 00:19:37,480
to the board so that we can take a look at
the waveforms produced by these components

332
00:19:37,480 --> 00:19:39,080
with an oscilloscope.

333
00:19:39,080 --> 00:19:42,720
That will probably be the next video on the
channel, so be sure to subscribe so you won’t

334
00:19:42,720 --> 00:19:43,720
miss it!

335
00:19:43,720 --> 00:19:46,649
Thanks for watching, and I really hope you
enjoyed this video, it was a lot of fun to

336
00:19:46,649 --> 00:19:50,830
make, especially going through the circuit
board, finding the datasheets, and all that jazz.

337
00:19:50,830 --> 00:19:54,679
As always, a great big thank you to everyone
who supports the channel on Patreon.

338
00:19:54,679 --> 00:19:58,740
One of the greatest things about the support
I have from Patreon is that it allows me to

339
00:19:58,740 --> 00:20:02,080
keep the channel going without resorting to
outside sponsorships.

340
00:20:02,080 --> 00:20:06,610
I’d much rather this channel stay entirely
community driven, and thanks to people like

341
00:20:06,610 --> 00:20:08,680
you that’s remained the case!

342
00:20:08,680 --> 00:20:12,360
If you’re interested in pledging some support
to the channel to help it grow, please check

343
00:20:12,360 --> 00:20:14,039
out my Patreon page!

344
00:20:14,040 --> 00:20:16,860
Thank you for your consideration, and I’ll
see you next time!

345
00:20:16,860 --> 00:20:20,560
♫ uncomfortably smooth jazz ♫

346
00:20:21,820 --> 00:20:24,460
...binary stream from what is the analog

347
00:20:24,460 --> 00:20:28,080
(long
pause)

348
00:20:28,080 --> 00:20:29,180
No, that was right.

349
00:20:29,190 --> 00:20:30,190
OK.

350
00:20:30,190 --> 00:20:32,340
That it Reepresent…Reepresents.

351
00:20:32,340 --> 00:20:33,080
Reepresents?

352
00:20:33,320 --> 00:20:34,060
Reepresents.

353
00:20:34,060 --> 00:20:36,160
These continue to create 1/75th....

354
00:20:37,200 --> 00:20:37,920
Oh man…

355
00:20:40,220 --> 00:20:40,960
c’mon!

356
00:20:40,960 --> 00:20:43,220
But even then, there are different nnmmmmmMMMMMMmmm.

357
00:20:45,180 --> 00:20:46,340
“modes!”

358
00:20:46,340 --> 00:20:48,850
...which the CD player stored in memory.

359
00:20:48,850 --> 00:20:50,280
It now knows whe…

360
00:20:50,280 --> 00:20:53,300
(glances at mechanism
he just bumped up against)

361
00:20:55,400 --> 00:20:56,280
or that

362
00:20:56,700 --> 00:21:01,320
...plus they wouldn’t be the 1.44 megabytes
in size until 1986…

363
00:21:01,320 --> 00:21:02,820
they wouldn’t… hit…

364
00:21:04,520 --> 00:21:05,320
dangit.