Eagle-eyed viewers may notice a leaky capacitor in the power supply section. Rest assured that I'm very aware of it. I have seen the leaky capacitor in the power supply section, and I have made a note of it. Yes, in the power supply section, there is a leaky capacitor. To prolong the life of this compact disc player, perhaps I should address the leaky capacitor in the power supply section. You may also have noticed a leaky capacitor in the power supply section. Additionally, you might have also noticed that the circuit board is in fact held down with additional screws; those that secure the CD reader mechanism to the case. I have realized that. Plus, I completely forgot about the long screws that went through the power transformer near the power supply section, a section which is currently inhabited by a leaky capacitor. And did you see the leaky capacitor in the power supply section?
Dude, I could not love this channel more. I've been working with electronics and circuit boards and whatnot for years, and I've *never* seen someone 'trace' the... heh, traces... and explain the function of a piece of consumer technology in such a clear way to a generally-layman audience. Fascinating and well-done. Thanks for the effort.
And as a layman whose knowledge of electronics is sketchy at best but loves technology just the same, I really appreciate Alec's explanations and it's genuinely making me want to learn more about circuitry.
One thing that I think you missed that would be worth mentioning: the fact that CDs are constant linear velocity, and thus that the rotational speed of the disc changes over the course of the disc.
It's funny, you've uncovered a mental block of mine. On countless occasions I've told myself, "you need to explain that" but it's just such a hard-worn fact in my brain that I've kind of assumed it's common knowledge. But that's no excuse to not mention a thing!
Interesting fact is how long it took before they could be written in CAV (for higher average speeds) The wait was for CPU's (or FPGA's ) with high enough clock speeds to make a high enough resolution of the encoder clock possible.
The topic of the spindle control was completely avoided. I know because that CD player is tough to understand. The disc speed is controlled by the data slicer instead of the servo chip as other players do.
@@konatadesuka As I have commented earlier, there are basically two ways players have controlled the spindle. 1) As someone else suggested, the player would attempt to keep the buffer at fixed ratio. If buffer fills up faster than it depleted, spindle had to slow down and vice versa. (Looking at the chipset, this is how this player did it.) 2) As I know reference players do it: If slicer sees pits with lengths shorter than 3T (clock cycles) player needs to slow down. If slicer sees pits longer than 11T, player needs to speed up.
The explanation in this video is nothing short of excellent. I worked at Philips Optical Storage in Hasselt Belgium, where parts of the unit used were developed and made. The only possible recommendation I could make is that some mid-level block diagrams (higher level than circuit diagrams) can be helpful.
@@oscargr_ who says my grandfather is old? ;-) But I know what you mean :-P. But my grandfather did work there. He was also born and raised in Hasselt where the plant used to be.
Ah, so maybe you can explain how they managed all this witchcraft in 1982? I‘m still surprised how CD tech was so much more advanced than the rest of the contemporary consumer electronics field.
Then you can answer my question: Was het systeem van Philips hetzelfde als die van Technics? Het heeft dezelfde onderkant. (Hasselt speaks Belgisch, right?)
The actual power switch is "way back there" because it helps keep electrical noise away from the audio circuitry (such as the headphone jack directly below the power button). Also a plastic stick is cheaper and simpler to assemble than wires running back and forth across the chassis.
StraightOuttaJarhois Uh... a CD documentary, at this time of year at this time of day in this part of the country localized entirely within this video?
Look at that early SMD work! It’s always hilarious to see how they GLUED components to the board for wave soldering. This is excellent content. You deserve a raise. In fact, I’m raising my contribution now.
Honestly the early stages of evolution can indeed be hilarious regardless of their huge significance. Nowadays they keep doing this but with less glue at least.
Pick and place with newer solder pastes and reflow technologies doesn't usually use massive quantities of glue anymore unless they're doing double sided SMD work or are cheap parts from what I've seen. The hilarious part is the excess glue sitting on the unpopulated SMD pads where parts were dropped from the product but the glue mask wasn't altered, leaving red dots in spots where smd components once were. I've seen some of this old SMD work where they would use printed carbon resistors (which are pretty cool, actually) but they would switch AFTER the glue mask was put into production, so they'd print the carbon over a dab of glue. Janky, but it works. I'm not saying they had a better option, but it can still be funny even though it was the only cost effective means at the time. Have a sense of humor, electronics and mass production methods aren't some sacred cow. Wave soldering SMD components is a ridiculous way to solder them.
Back in the day I wrote code to take subcode streams off CD+G (Karaoke) disk images. I had to reverse engineer a lot of this stuff (there was no Wikipedia article to look at in those days.) This video has brought back some great memories. Thanks!
7:27 the Motorola IC is likely a Mask ROM microcontroller they usually have unique part numbers on them and no datasheet since the silicon die is custom made (programmed if you wish) for a specific manufacturer. It was more cost effective for high volume consumer products than programmable EPROM or OTP microcontrollers. As long as your code was thoroughly debugged before the chips were made of course.
Super common to use programmable EPROM or OTP in development models and even in the first runs. Then you find where the errors are in production, and once you’re solid on your code, now you can push out the custom die.
It's possible that the microcontroller has components or H/W blocks to deal with that information. I think a mask ROM MCU is the most likely explanation for the Motorola chip.
@@joshmbrown42 : Bitstream input is probably so that it can get the subcode bytes, which the decoder chip probably doesn't touch. The decoder chip likely provides a clock extracted from the bitstream input for precisely this purpose. All other connections from the processor to the other chips are likely for control purposes, since the processor is the most logical place to put "override logic" that modifies the "natural" behavior of the other chips.
It's simply amazing how well the early Philips CD players were made. I have a Philips CD824 from I believe 1989 that I still use daily. It never skipped a beat and still performs wonderfully.
yep, as long as all caps got changed, not only the bad looking ones. cap near 8808/5708 usually looks perfectly fine. Its responsible for smooth laser operation, so power is minuscule, but when it dries whole unit acts like it has a dirty lens - user starts "cleaning" the lens with alcohol screwing anti-reflex coating and still not making it work.
This takes me back to some of my university courses on digital communication. Parity, data frames, sectors. I'm so glad all that stuff is transparent to everyone except a few very specialized designers these days! (We were also told that it was physically impossible to get more than 2400 baud through a telephone line due to the inherent bandwidth of a twisted pair line. -- I guess they broke physics a few years later!)
In 2001 I was trying to set up a high speed WAN for my head office and 30+ stores across Canada. Shaw at the time (and still is in my opinion) was the leader in high speed internet (fiber trunk). Telus was trying to sell me "frames" at a ridiculous "business" price. Shaw was $100s for x MB Telus was $1000s for Kb and the "frame" rate was laughable at the time. I am so glad we have move on from those confusing days.
Actually, baud rate never got much beyond 2400, with 3200 being the highest we achieved over phone lines beginning with the 28,800bps and faster modems.. Rather what changed was the number of symbols encoded in each baud as “baud rate” is simply a reference to how many symbol windows/transitions occurred in a given second. In the early days there were only two symbols, 0 and 1, encoded per baud so the baud rate and bit rate were identical. Beginning with the 4800 bps modems, and beyond, we first increases the symbol table from two to four, then to eight, then to sixteen, and so forth. The absolute maximum possible bitrate over the old POTS phone lines was either 56 or 64 kbps (with 56K being the common denominator as even many 64K pipes reduced to 56K every fourth frame to relay backchannel and call setup data - the so called B and D telco channels) as that was the actual bitrate of the digitally converted audio stream that the phone companies used to transmit the audio data over their digital lines. To get speeds anywhere near 56K required exceptional timing and adaptive learning algorithms as the timing and signal levels had to be near perfect since you needed to break the Nyquist limit for 56kbps audio stream as you were effectively aligning the timing and signal levels with such precision that you were effectively hacking the individual bits that the downstream and upstream ADC and DAC were going to end up generating which is partly why such modems did not come into being until the 1990s with the availability of affordable digital signal processing chipsets.
There is a fundamental difference between the analogue modems using the severely band-limited POTS voice channel and the digital modems (terminal adapters in ISDN-speak) as used by DSL/ISDN. For analogue modems, it is all about ingenious modulation, e.g. TCM that was used to squeeze as much as possible over the voice channel, but ISDN/DSL is quite different. This is why you can run DSL in parallel with an analogue phone using a splitter. The B and D channels you refer to are ISDN channels.
A V.90 (up to 56k) modem connection actually had only one modem at one end. The other end (usually the ISP) fed a data stream directly into the digital phone network, which then got converted to audio via PCM at the linecard of the local exchange. All an 56k modem did was trying to reconstruct the data from the resulting audio signal. The other direction got modulated with regular V.34bis (up to 33.6k).
I enjoy the silliness, it's not childish and not self-induglent or attention seeking, just subtle and funny and works great in a very interesting video like all of yours are.
My very first CD player was an Emerson from the very same period as your Magnavox and I remember being so amazed by the thing that I played the same CD over and over again (because I only had one for a long while) in complete awe. I had an old Soundesign stereo and I simply couldn't believe that the old thing was capable of producing the kind of sound that was coming from that CD player. I've been in love with the CD ever since. Thanks for the great videos Alec, I absolutely love your work!
19:19 Mr TC, thanks for the fact, that you didn't explain it in a technical way, but gave to the audience some simple and basic explanation .I already started to have a severe head-ache.
I am an audiophile and love CDs, have tons of them. I still consider CDs the greatest leap in the audio industry (vinyl fans weeping here) and even though I have DACs and TIDAL streaming the sound of a good CD player is still fantastic. I also love my LPs and analog sound when done finely, but I have to admit I hate the noise, the lack of dynamic range, the expense and the excessive care they demand. Amazing explanation. My respects!
Being a vinyl record enjoyer (and compact cassette... and reel to reel when I finally get around to buying one!) doesn't mean I don't absolutely adore the crystal clarity of a well-mastered CD. That format offers a whole host of benefits, which were unfortunately mostly squandered by the cretins who engaged (and still engage) in the loudness wars. So many potentially excellent albums completely ruined 😢
@@duffman18 real to reel tape is very impressive. Where do you think most of your CDs originated from in the studio? Yep real real is awesome. Next time you see a CD that says AAD OR ADD The a probably came from real to reel as the analog component.
The components on the PCB are a wonderful mid-era mix of through-hole on the top, and surface-mount on the bottom. It's quite interesting to see - a few years earlier and none of that SMD would be on the bottom; a few years later and nobody would be doing through-hole and it would be all surface. Yet this board sits right there in the middle between the two.
I have recently started collecting CDs again and recently acquired a 90s Panasonic Discman. Sound quality and build quality is amazing. Such an interesting video as always! Love your humour too
Brilliant content as always. Mesmerizing to watch. The level of detail you put into your explanations is legendary, and your humor and enthusiasm keep the interest high!
The reason for the long plastic lever connecting the front panel power button to the actual power switch in the back is to maintain separation between the mains-powered part of the electronics (the power supply) and the low-voltage portion (the rest of the circuit). This is for two reasons: a) Safety. You want to maintain a good amount of separation between any high voltage and low voltage sections. Having a high voltage wire running from the power supply in the back to the front panel breaks this separation. b) Noise reduction. Having a long wire carrying AC voltage can easily introduce hum into the audio, so it's best to avoid it. Modern electronics often get around this by having the power switch not actually switch the AC voltage directly but instead control an electronic switch, but this can lead to "vampire" power usage since the circuitry to handle the electronic switch uses up a small amount of current even when the component is switched off.
Huh. Hadn't thought about the noise! I have a cassette player with the power switch actually at the front of the case, with incoming power routed through it. I didn't really see the harm in doing it that way and figured the plastic Fing-Longer was some sort of cost cutting measure.
Analog Oscilloscopes used to use these kinds of rods and mechanical parts to control most of the switches and pots mainly to have very short wires and PCB tracks which improves performance in general.
That actually reminded me of this: ua-cam.com/video/vZgem0EdCBQ/v-deo.html The whooping noise at ignition is sympathetic current in the cabin PA system. It's caused by the system that recharges the capacitors that provide current to the igniters. The two circuits are right next to each other, and run the length of the cabin.
Awesome! this is the best and most detailed explanation of how a CD player works I ever seen,on other videos they just open the unit up and make "educated" guesses, example "these large chips are probably the DACs" "these small ones are probably power supply regulators" just a bunch of nonsence.I'm impressed by how small the DAC chip really is since it's often given all the credit.And the good thing about old electronics is that they had discrete chips for discrete functions nowadays they just place everything in one small chip and call it a DAC even though it's a lot more than that.
Are you ever planning on a video about telephone switching technology, phreaking, etc? It would go nicely with the telephone video and you could explain the history of the Strowger switch, Panel switch, and/or Crossbar switch technologies. Perhaps you could even pay a visit to the Connections Museum in Seattle.
Oh indeed, he needs to do that, Evan Doorbell's tapes are an invaluable resource, but of he wants to do it, he better start listening to them now, the information isn't condensed in any way.
Yes please do this. My favorite is when Bill Gates found the telephone system manual in his university library and used it to phone freak. I can confirm that in my university library, there was some titles that shouldn't be there.
Yes indeed, the old BSP(Bell System Practices) documents, they explain everything from how a 4A toll switch works to the MF signaling system and how to sweep the floor of a central office.
I appreciate your willingness to go out on a limb with your educated guesses about the functions in front of such a huge audience, and also the large amount of work you must have put into this video, not just disassembling and reassembling the device, but actually going to the trouble to trace solder runs from pin to pin, which is no small headache considering you have to keep it all straight as you flip the board over from component side to trace side. Good on you.
Interesting that you decided to dive-in in that Philips player, although it is a "basic" player, it's more complicated than a Sony unit in my opinion because Philips made a set of chips to work together and they don't make much sense when viewed from outside or read from a schematic. Philips put a player together in a way so dissimilar to the Japanese manufacturers like Sony, Sanyo or Toshiba. What you got there in that Philips player is the two analog servo chips (TDA8808/9) for the tracking and focus control. They are connected to the Microcontroller (The Motorola chip, CPU from now on) because they output some error detection signals the CPU has to take into account when playing. The TDA7210 is the decoder that takes the raw analog data as read directly from the disc (The HF signal) and digitizes it when it goes through the "data slicer", then it reverses the EFM, does the CIRC error correction and masking then it spits out the audio data to the DAC(s) and the subcode information back to the CPU (to indicate play time, beginning and end of a track, remaining playtime, etc.). The reason why you didn't find any information on the CPU used there is because Philips asks for a bunch of such chips, already programmed, and relabeled from Motorola. They're HC08 microcontrollers and can be found in almost every Philips player based on the same "chipset", because they produce the chips and code to work with them on that architecture then they adapt the code to work in the application they require. For example I have a Philips FW26 system with a single tray player based on a new-ish implementation of the same design and it has a small QFP cpu with the Motorola logo and the label "CD SHORTL", it doesn't control the display like on your player but it takes commands through an I2C bus from the main CPU on the system. A note on the track seek and subcode, there's no standard way of accessing a particular track but the players can be as smart as the manufacturer wants them to and there are a few strategies. One is doing a raw calculation of how far the sled has to kick to position the laser "around" the track then do a slow seek, a "smart" move (observed in Sony players). Another is doing a 10 track skip seek and decoding whatever frame the CPU can catch under the beam until the beginning of the track is under the lens then doing a fine adjustment with 4-2-single track skips. This last one is the "dumb" one. Skip track by track looking at the P subcode. The P subcode has just zeros and ones to mark the beginning of each track, the CPU can just count how many "marks" it has skipped on the P channel and land exactly at the beginning of each track. Also, thank you very much for the credits, I genuinely love the CDM4 mechanism.
I have to say, your technical input is absolute gold. As a music lover, I seem to have unknowingly moved towards the cdm4 as it, or at least one of its close variants, is fitted to 4 from a total of six of my players. The alarming fact is that while cds are still being pressed in their millions, replacement cdms are nla. I still really like it as a music media. Thank you for your input.
I love how ridiculously deep you're going into this. And I love your narration and writing. Also the performance and editing. Long Live Technology Connections!
Your ability to explain all of this intricate technical detail in only a 20minute video is astounding! Good job! Your videos are so good, surprised you don't have more subs/views than this!
You are a freaking genius! I watched tons of your videos but sometimes I got, somehow, lost. But this time I completely got out of my mind... Ain't saying I'm stupid, on the contrary that I never have meet in my entire life, someone so committed to understand technology (besides people working in the field). You are definitely in my top 3, alongside Techno Moan and 8-bit guy! You are AWESOME and I hope you guys could cooperate together to make yt a sort of Wikipedia of audio/video systems. History and explanations of your knowledges combined!
Alex Root lol he’s been talking slower! back when i found his channel watching a video would last an hour cause i had to keep rewinding. i just couldn’t keep up!
I can tell that all of your research has taken TONS of time, and you have presented in a very concise, to-the-point way that makes it very easy to understand! Thank You for all your time and effort!
Thank you sir, for giving me information that I don't need but I watch every minute of every video. I just like to know what makes things tick. I'm a techie and always have been. I was a Crypto tech and a computer tech (not PCs) in the military. I really appreciate the level of detail!!
Makes me wonder what kind of computers the military uses, bet you can't say much but still wondering. Also heard about that time when the PS3 had just came out and there was rumors of the military buying a lot of them to work as a supercomputer clustered
How are you so smart! Not only do you research all this info but you are able to put it into a concise and a comprehensible video. AND edit that video, creating a great watch at a perfect length. Bravo ol' chap!
I can't decide whether I'm more impressed with the in-depth understanding of how all of that works, or the editing and effects, especially highlighting the board traces. Amazing work!
This was fantastic, as someone who used to pull things apart in an attempt to see how they work, and still does so on occasion, I really, really appreciate this. Thank you.
CD-ROMS were definitely possible in the early days of Audio CDs, but there was really no use for them. Home computers couldn't play videos, animations or even display still images with more than 16 colors. We didn't even need a full CD-ROM standard. Simple Audio CD with a way to send digital stream directly to a computer would be good enough. Much faster, convenient and reliable than audio cassettes used by ZX-Spectrum or C64. There was no need for such a capacity back then. Professional use could use any amount of storage, but only if it could read and write to it. It's a bit similar to Laser Disc vs VHS, where superior image quality wasn't that important, because VHS allowed for recording at home. Even a decade later first CD based consoles like Sega CD or Amiga CD32 weren't utilizing such big starage besides CD Audio or choppy video. At the same time CD-ROMs were gaining popularity in PCs. Software distribution was cheaper, since single CD was cheaper and more convenient than 10-20 floppies. And CD/DVD-ROM era started. CDs are brilliantly designed. Technology that is in use for almost 40 years. Of course DVD and Blurays took over in many aspects, but it's still the same technology, just with improvements.
Of course nothing then would have prevented text only, command line software from being stored on a CD. A single disc with 32x the capacity of my (upgraded!) IBM PC-XT would have been wild then.
The Atari 8-bit family is underrated in this regard. It was so forward thinking people didn't really realize the specs didn't change much until end of fabrication in 1992. 256 color, 384 lines, 1978.
@James Slick Yes it would have been wild but also pretty useless. It would be a bit like trying to distribute today a read-only storage format with 500TB of data.
soundspark, Are you sure? I heard that CD players are classified as Class 1 laser products *only* because it is enclosed and the light does not escape in *normal* use. I don't think it's safe to look at the diode directly. Please correct me if I'm wrong.
Bagana I am since I DID stare a LOT into a CD laser when I was a kid and have no vision problems at all :P Didn't know it was infrared back then. For burning it's a whole other story, but you'd be hard pressed to get that to activate without a disc being properly detected without serious hacking. What turns on at disc insertion is obviously the read one.
This is insanely good. Thank you. I feel like I have at least a big picture idea of what it actually takes to read the optical data and produce audio at the end.
that was fantastic. you mentioned before that tracking was done on the hardware level, i'm glad you've expanded on that and so much more here. please keep it up
I'll be honest, this mostly makes me even more impressed about the whole LaserDisc thingy - how did they get the "warp" and "off-center" cancellations (focus and track alignment) correctly while having a disc that spins so much faster?
I like my Sony CD player. The laser is fixed and it is the disc that moves back and forth. I like the madness of it. Yeah, lets move the rotating disc instead of the laser! It's supposed to be better than moving the laser, but it is also Sony telling you it is better.
I have watched the sound and video series all the way back to the beginning, and I am very appreciative of how much I have learned about both audio and video. I have always wondered, but found most explanations to get too complicated too quickly and never had the motivation to really study it. I now find myself giving brief explanations to people about things like how CRT monitors worked and how the laser reads a CD. Thank you for explaining throughly while keeping it simple! Always excited to see the next video.
8:22 This line seems to be to indicate data being sent from the TDA8808T to the SAA7210P. The data being sent goes over TDA8808T pin 3, labeled HFout, and goes to SAA7210P pin 25 labled HFI (High-Frequency Input). The lines between the processor and the decoder you marked send the information in the Q-channel to the processor. I found this information in the datasheet for the SAA7210 (non P) but it appears to be a compatible chip. Also the AGC is not a ADC. The TDA8808T still outputs an analog signal to the decoder.
Aww dang, why didn't I drop the P? That would've brought me where I wanted to go. Admittedly the leap from AGC to ADC was an assumption based on the fact that two devices were hooked into the focus chip. I figured its probably a binary stream by that point, but I stand corrected! In the next part I'm sure we can find some more details with the scope!
Typically the “P” suffix in chip part numbers will denote that it is a Plastic DIP package (versus, for example, ceramic, mil-spec, or other package style).
really love these videos; electrical engineering was something I seriously struggled with in college despite wishing I knew how all these devices worked. It's been really nice to get a systematic explanation of this stuff in a way that still keeps my attention long enough to absorb it.
Legit. Try getting your hands on one of the Sony YEDS test discs especially made to generate tones for use on an oscilloscope. I found rip of YEDS-18 a few years ago on some random forum. A bunch of the tracks are mastered with 'emphasis' for early players with 14-bit DACs.
Did I mis it, or did you forget to explain the Sony solution for keeping the laser on track and in focus. (fascinating episode, by the way. Keep this up! It is much appreciated!)
Hi, can you speak about the "flags" of the CD. I think they are in, or around, the TOC. They are also ver interesting and forward thinking. There are multiple flags, my favorite is the Emphasis flag. (A type of noise reduction). It alerts the player to engage its Emphasis circuitry. Since only some of the very early discs had (pre) emphasis, most players have never powered on their dedicated emphasis chips, yet compolsory for every CD player to comply with the red book standard. There are other flags too. Some never assigned and to be used for future planning (like how you mentioned the extra data storage, etc!)
The SP-DIF protocol includes a flag for pre-emphasis also, and outboard DACs are supposed to support it, thougn not all do. I was once debugging a Linux audio driver where this flag was being erroneously set, causing everything to sound rather muddy and flat, but only through one of the two DACs I had on hand.
That PLLOH signal could be used as a status that the CD isn’t spinning, or not locked. and either mute audio digitally, or simply stop the controller counting.
I just want to thank you for the amazing entertainment and wonderful explanation you always give and your dry humor is very much welcomed. And as a fellow IL resident (Central IL) I appreciate the little things here and there that are IL unique.
After wondering exactly how Audio gets from a little plastic disc to my ears for almost 30 years, i finally find someone who can explain it in terms I can understand. I love this video, I love this channel, and I can't wait for more.
I think you are wrong about the microcontroller doing EFM decoding. The chance of any microcontroller from the 80s being fast enough to handle EFM decoding is near-zero. The chance of a microcontroller being fast enough to also handle buttons and a display at the same time is zero. Besides, the datasheet for the SAA7210 shows it does EFM demodulation (and everything afterwards). Now your theory about decoding up the Q channel has some merit; An expert programmer could potentially design a cycle accurate state machine which pulled the correct bits out of the datastream. However, the SAA7210 datasheet shows it has built in subcode decoding, and even has a nice slow microcontroller interface to send that to a microcontroller when requested (those are the two pins you showed connecting between the two) No. I suspect the laser bitstream is going to the microcontroller for a much stupider reason: To detect if there is a disk in the tray or not. Even the slowest microcontroller can sample the bitstream a few times at random. If the random samples are all zero, or all one over 10+ samples then probability states that there is no disk (or some other disk reading error).
A fantastic video, as always! Unfortunately, I think there's a small mistake in there: The Motorola microcontroller likely doesn't have anything to do with EFM. The SAA7210's datasheet states clearly that it's responsible for EFM decoding, not some other chip in the system. The mysterious trace that's shared between the TDA8808, the SAA7210 and the uC is this likely *not* the EFM signal -- it looks like its connected to HFD (input) on the 7210 and PLLH (output) of the 8808. If so, it's just a status signal. 7210 pin 25 (HFI) is where the EFM money is at. Sure enough, it's connected to pin 3 (RFout) of the 8808.
Indeed Philips made all the hard working chips for the decoding. The SAA7210 datasheet shows it doing everything from RF decoding, CIRC, error concealment and interface to DAC. And sending subcode data somewhere (the Motorola). and the RAM is a buffer - this chip also adjusts motor speed if the buffer is too empty or full. It seems the CIRC decoder has its own RAM even. The data sheet has a really nice explanation of the workings, even which part of the RAM is used for what data and how the motor speed control operates. Or that a single value flagged as unreliable by error correction is interpolated linearly. It really seems that Motorola MPU is at the end of the pipeline - not somewhere in the middle.
Society owes you a tremendous debt for the great work you've been doing in preserving and demystifying this stuff. Your videos are part of my kids' curriculum.
Oh fuck that DAC chip in particular. I never got the stupid thing to work right. I was questioning my sanity many times playing around with it. Still don't know if I'm that bad at coding the I2S thing or I just got damaged parts (they are apparently 30+ years old lol).
Once I had the cover off a CD player because it was misbehaving. Set above the bedhead I realized I could see the reflection of the lens while spinning. I was blown away as I watched it maintain what seemed a perfect distance from the disc as it wobbled (as mentioned no disc is perfect) while it spun. I was mesmerized by the beauty of the control system doing its thing.
I did not mean to imply through saying "later on" that it would be later on _in the same video._ I need to get my hands on a cheap CD player that I can happily destroy.
Get a comparable player to this one (late 80s, early 90s). It doesn't make sense to get a new boombox for that as the chipset used in those is ridiculously "compact" to say a thing. It's just one piece of silicon that does everything.
@@TechnologyConnections if you could do it without destroying it I could offer up my first Sony single tray load CDP-670 I got from Crutchfield I believe in '89. Just replaced belt last year & works great. Just not sure it's worth the shipping vs just grabbing one at a thrift store.
Thank you very much for this more technical video about CD! I thought I knew everything about how CD worked and never realized it's possible to dive deeper. Looking forward to the next video about CD.
I thought it would be strange to split 8to14 modulation decoding from CIRC processing to 2 chips. And the SAA7210 datasheet concurs: it does demodulation, CIRC, and some more tasks, and outputs I2S data to DACs. The input data stream processing is definitely not split between the 2 chips. Oh, yeah, and it seems to extract the Q channel that is probably used by the Motorola chip.
Perhaps the Motorola listens in to the raw data to count how many tracks are skipped. Or just to check that signal is being detected (as part of disc detect).
I really appreciate this video. A great deal of thought was put into the design of the CD format - with a special emphasis on error-correction features - which were implemented in multiple ways. The convenience, robustness and fidelity of the CD format explains its popularity and longevity. The only thing they could have improved on, in my opinion, would be the inclusion of CD-TEXT and album artwork - from the get-go. Having the album title, track title, artist name, and album artwork included from the very beginning would have totally blown people away in 1982, and further accelerated the demise of the cassette and LP formats. Today, any CD-Rs I make always include CD-TEXT, but sadly, most of my legacy CD players can't display it. At least the CD engineers were smart enough to include unused data areas for this, and other features. Carry on!
How does Sony do the tracking? And: the reason for the switch in the back is simple: it cuts the mains voltage, so the transformer itself is turned off. And you don't want mains voltage near the front panel, it might interfere with the relative sensitiv ICs and with the audio. So it's good practise to separate the voltage and letting mains voltage only to the absolutely needed parts. Great video, BTW!
Sony does the tracking using 3 laser beams instead of one, and two more photodiodes in the detector to read back such extra laser beams. The main beam reads and it is centered on the main track and the two others, smaller ones, are stepping, off center of the adjacent tracks. The signal from those extra beams is monitored and when it is equal, the lens is on track, when there's a difference, more light on a diode than the other, a TE signal is generated, amplified and tracking is corrected by the servo chip.
I don't think I have learned anything that I didn't already know (cause I am a dork) , but it doesn't stop me from going OHHHHHHH whenever one of your video's show up in my subscriptions
William J. Some even incorporated a visual display of the buffer’s depth such that you could see the bar graph rise and fall depending upon how deep you dipped into the buffer. I seem to recall the buffer depth being around 8-10 seconds on some early units, but that was all a very long time ago.
@@pauldzim it was almost 40 years ago, which is an eternity when it comes to technology. It's really an astounding feat what they pulled off with CDs, given the technology constraints of the day.
Ever since I saw this video and this amazing Magnavox player with the Philips swing laser CDM-4 mech, which I didn’t even know existed (being a kid from the early 2000’s), I got into a life-death task of finding a player like this. I was already looking from a player from the 1980’s, because of their sound signature, that I quite enjoy, but this made me obsessed. Finally got me a Philips CD-610 a couple of months ago, which is, internally, the exact same machine. It sounds great too! Edit: I was one of the sneaky eyes that caught the leaky capacitor in the power supply. Mine has that exact same leaky capacitor, even tough I have no plans of replacing such leaky capacitor, as the leaky capacitor seems not to be affecting the board really bad the way some leaky capacitors do.
Better than just buying an album digitally. You get CD quality audio, a physical copy of your purchase and it's only a little more expensive. Plus you can make a MP3 on your PC.
Yeah, I've always loved CDs for the fact that you can get a physical copy of an album in (presumably) high-quality audio- higher-quality than vinyl records, even, if they're mastered properly- without having to worry about the disc wearing down with each passing play. Digital downloads are good for listening on the go, but when I'm sitting down and want to listen to some music, I'm gonna go for a CD 90% of the time. It's honestly a shame though that CDs could never recover their old momentum in the same way vinyl records and cassettes did; I blame record labels' continuing tendency to volume-boost and over-compress their digital masters because of the myth that louder music is somehow inherently better.
Eagle-eyed viewers may notice a leaky capacitor in the power supply section. Rest assured that I'm very aware of it. I have seen the leaky capacitor in the power supply section, and I have made a note of it. Yes, in the power supply section, there is a leaky capacitor. To prolong the life of this compact disc player, perhaps I should address the leaky capacitor in the power supply section.
You may also have noticed a leaky capacitor in the power supply section.
Additionally, you might have also noticed that the circuit board is in fact held down with additional screws; those that secure the CD reader mechanism to the case. I have realized that. Plus, I completely forgot about the long screws that went through the power transformer near the power supply section, a section which is currently inhabited by a leaky capacitor.
And did you see the leaky capacitor in the power supply section?
I didn't see the leaky capacitor, but I saw a "2" in the table of bits of the subcode channels :P
@@billyboh78 I think Alec had the same nightmare Bender had in Futurama
I have not yet seen the leaky capacitor, but when i do i'll let you know.
did you notice in the power supply section, there is a leaky capacitor? just thought I'd let you know, because I am a great guy. you're welcome.
Make sure that when you breath, you inhale THEN exhale.
Whatever you do, *don’t* cut back on the silliness - the world needs more dry humor, and yours is perfect.
It's an entirely different kind of humor, altogether!
Especially the Kenny G outro music
Amen! if not for the 'silliness', I'd never get all the way through these things.
kinda mushy, but you're exactly right.
I agree
Dude, I could not love this channel more. I've been working with electronics and circuit boards and whatnot for years, and I've *never* seen someone 'trace' the... heh, traces... and explain the function of a piece of consumer technology in such a clear way to a generally-layman audience. Fascinating and well-done. Thanks for the effort.
And as a layman whose knowledge of electronics is sketchy at best but loves technology just the same, I really appreciate Alec's explanations and it's genuinely making me want to learn more about circuitry.
Wait, how do you know his name is Alec?
@@wallykramer7566 He told it before? 🤷♂️
One thing that I think you missed that would be worth mentioning: the fact that CDs are constant linear velocity, and thus that the rotational speed of the disc changes over the course of the disc.
It's funny, you've uncovered a mental block of mine. On countless occasions I've told myself, "you need to explain that" but it's just such a hard-worn fact in my brain that I've kind of assumed it's common knowledge. But that's no excuse to not mention a thing!
@@TechnologyConnections it sounds so slowly at the end of the CD.
Interesting fact is how long it took before they could be written in CAV (for higher average speeds)
The wait was for CPU's (or FPGA's ) with high enough clock speeds to make a high enough resolution of the encoder clock possible.
The topic of the spindle control was completely avoided. I know because that CD player is tough to understand. The disc speed is controlled by the data slicer instead of the servo chip as other players do.
@@konatadesuka As I have commented earlier, there are basically two ways players have controlled the spindle.
1) As someone else suggested, the player would attempt to keep the buffer at fixed ratio. If buffer fills up faster than it depleted, spindle had to slow down and vice versa. (Looking at the chipset, this is how this player did it.)
2) As I know reference players do it: If slicer sees pits with lengths shorter than 3T (clock cycles) player needs to slow down. If slicer sees pits longer than 11T, player needs to speed up.
CDs still blow my mind. You can load music and data from a shiny, thin, plastic disc by shooting frikkin' LASERS at it? Such technical wizardry!
There's nothing old about lasers. The Heechee used them. (Frederick Pohl, The Heechee Saga) ;)
That's the power of mechanical engineering and years of calculus helped these engineers get to this point through massive trial and error.
The explanation in this video is nothing short of excellent. I worked at Philips Optical Storage in Hasselt Belgium, where parts of the unit used were developed and made. The only possible recommendation I could make is that some mid-level block diagrams (higher level than circuit diagrams) can be helpful.
Mijn grootvader heeft daar ook nog gewerkt! :-)
@@faithless8888 creative way of suggesting someone is really old. LoL
@@oscargr_ who says my grandfather is old? ;-) But I know what you mean :-P. But my grandfather did work there. He was also born and raised in Hasselt where the plant used to be.
Ah, so maybe you can explain how they managed all this witchcraft in 1982? I‘m still surprised how CD tech was so much more advanced than the rest of the contemporary consumer electronics field.
Then you can answer my question:
Was het systeem van Philips hetzelfde als die van Technics? Het heeft dezelfde onderkant.
(Hasselt speaks Belgisch, right?)
The actual power switch is "way back there" because it helps keep electrical noise away from the audio circuitry (such as the headphone jack directly below the power button). Also a plastic stick is cheaper and simpler to assemble than wires running back and forth across the chassis.
I keep finding your comments on audio related videos. Btw I like your videos
That tracking mechanism is devilishly simple.
That's delightfully devilish, CD!
It is quite similar to how hard drives did (and still do) tracking.
StraightOuttaJarhois Uh... a CD documentary, at this time of year at this time of day in this part of the country localized entirely within this video?
@@7EEVEE Yes!
...May I see it?
Look at that early SMD work! It’s always hilarious to see how they GLUED components to the board for wave soldering.
This is excellent content. You deserve a raise. In fact, I’m raising my contribution now.
Honestly the early stages of evolution can indeed be hilarious regardless of their huge significance.
Nowadays they keep doing this but with less glue at least.
They still do the same thing, nothing hilarious about it.
Pick and place with newer solder pastes and reflow technologies doesn't usually use massive quantities of glue anymore unless they're doing double sided SMD work or are cheap parts from what I've seen. The hilarious part is the excess glue sitting on the unpopulated SMD pads where parts were dropped from the product but the glue mask wasn't altered, leaving red dots in spots where smd components once were. I've seen some of this old SMD work where they would use printed carbon resistors (which are pretty cool, actually) but they would switch AFTER the glue mask was put into production, so they'd print the carbon over a dab of glue. Janky, but it works.
I'm not saying they had a better option, but it can still be funny even though it was the only cost effective means at the time. Have a sense of humor, electronics and mass production methods aren't some sacred cow. Wave soldering SMD components is a ridiculous way to solder them.
It is still done that way today.
Some cheap products like portable CD and MP3 players even have hand-soldered qfp components on them.
Plus the modern cleaner methods don't necessarily hold up as well over time.
That's the most in-depth explanation on how CD players work that I've ever seen.
Absolutely superb!
I love this channel, the presentation is excellent and content well researched. It's like a mini lecture.
Except it's about things that I actually find interesting. Lol
I tell everyone I know that has an interest in the tech and the history of that tech behind media about this channel.
Love this channel!!
"In the last video--" BINNNNNNNG
my thoughts exactly
AAAAAAAAAAAAAAAAAAH! It was almost as shocking as the "WOOSH" sounds in Austin Evans' benchmark sections.
Yeah what’s was that all about!?
Ooookkaaaayyy. I first watched this through Xbox UA-cam. In-vid cards don't pop up.
I knew I wasn’t the only one hearing this lmfao
There is a lot going on there. One tends to think digital is just programming, but it is quite a mechanical marvel too.
Back in the day I wrote code to take subcode streams off CD+G (Karaoke) disk images. I had to reverse engineer a lot of this stuff (there was no Wikipedia article to look at in those days.) This video has brought back some great memories. Thanks!
7:27 the Motorola IC is likely a Mask ROM microcontroller they usually have unique part numbers on them and no datasheet since the silicon die is custom made (programmed if you wish) for a specific manufacturer.
It was more cost effective for high volume consumer products than programmable EPROM or OTP microcontrollers. As long as your code was thoroughly debugged before the chips were made of course.
Super common to use programmable EPROM or OTP in development models and even in the first runs. Then you find where the errors are in production, and once you’re solid on your code, now you can push out the custom die.
It's possible that the microcontroller has components or H/W blocks to deal with that information. I think a mask ROM MCU is the most likely explanation for the Motorola chip.
@@joshmbrown42 : Bitstream input is probably so that it can get the subcode bytes, which the decoder chip probably doesn't touch. The decoder chip likely provides a clock extracted from the bitstream input for precisely this purpose. All other connections from the processor to the other chips are likely for control purposes, since the processor is the most logical place to put "override logic" that modifies the "natural" behavior of the other chips.
Probably 68HC05 or maybe 68HC11 micro. XC=Experimental Circuit. OTP, most likely, which is just a windowless EPROM. Mask ROM would have had MC prefix.
I'm no genius, but the descriptions are presented in such a way that even I can understand. I wish my college professors were more like this.
It's simply amazing how well the early Philips CD players were made. I have a Philips CD824 from I believe 1989 that I still use daily. It never skipped a beat and still performs wonderfully.
Actually this vintage CD drives based on TDA8808/5708 like to die due to silly Electrolytic capacitor drying up in LASER power supply (pin 17).
So I'm lucky I got it recapped completely about 5 years ago.
yep, as long as all caps got changed, not only the bad looking ones.
cap near 8808/5708 usually looks perfectly fine. Its responsible for smooth laser operation, so power is minuscule, but when it dries whole unit acts like it has a dirty lens - user starts "cleaning" the lens with alcohol screwing anti-reflex coating and still not making it work.
The repair shop recommended me to change them all after one of the caps in the PSU started leaking. So I guess I got good service there!
The best video explaining how a cd player works on UA-cam. Congratulations!
Wow, what an incredibly well documented video this is!! Thanks for bringing us this kind of content!
The more you know...
This takes me back to some of my university courses on digital communication. Parity, data frames, sectors. I'm so glad all that stuff is transparent to everyone except a few very specialized designers these days! (We were also told that it was physically impossible to get more than 2400 baud through a telephone line due to the inherent bandwidth of a twisted pair line. -- I guess they broke physics a few years later!)
Michael Steeves yes and no. A lot of the speed beyond 2400 was compression, etc.
In 2001 I was trying to set up a high speed WAN for my head office and 30+ stores across Canada. Shaw at the time (and still is in my opinion) was the leader in high speed internet (fiber trunk). Telus was trying to sell me "frames" at a ridiculous "business" price. Shaw was $100s for x MB Telus was $1000s for Kb and the "frame" rate was laughable at the time. I am so glad we have move on from those confusing days.
Actually, baud rate never got much beyond 2400, with 3200 being the highest we achieved over phone lines beginning with the 28,800bps and faster modems.. Rather what changed was the number of symbols encoded in each baud as “baud rate” is simply a reference to how many symbol windows/transitions occurred in a given second. In the early days there were only two symbols, 0 and 1, encoded per baud so the baud rate and bit rate were identical. Beginning with the 4800 bps modems, and beyond, we first increases the symbol table from two to four, then to eight, then to sixteen, and so forth. The absolute maximum possible bitrate over the old POTS phone lines was either 56 or 64 kbps (with 56K being the common denominator as even many 64K pipes reduced to 56K every fourth frame to relay backchannel and call setup data - the so called B and D telco channels) as that was the actual bitrate of the digitally converted audio stream that the phone companies used to transmit the audio data over their digital lines. To get speeds anywhere near 56K required exceptional timing and adaptive learning algorithms as the timing and signal levels had to be near perfect since you needed to break the Nyquist limit for 56kbps audio stream as you were effectively aligning the timing and signal levels with such precision that you were effectively hacking the individual bits that the downstream and upstream ADC and DAC were going to end up generating which is partly why such modems did not come into being until the 1990s with the availability of affordable digital signal processing chipsets.
There is a fundamental difference between the analogue modems using the severely band-limited POTS voice channel and the digital modems (terminal adapters in ISDN-speak) as used by DSL/ISDN. For analogue modems, it is all about ingenious modulation, e.g. TCM that was used to squeeze as much as possible over the voice channel, but ISDN/DSL is quite different. This is why you can run DSL in parallel with an analogue phone using a splitter. The B and D channels you refer to are ISDN channels.
A V.90 (up to 56k) modem connection actually had only one modem at one end. The other end (usually the ISP) fed a data stream directly into the digital phone network, which then got converted to audio via PCM at the linecard of the local exchange. All an 56k modem did was trying to reconstruct the data from the resulting audio signal. The other direction got modulated with regular V.34bis (up to 33.6k).
I enjoy the silliness, it's not childish and not self-induglent or attention seeking, just subtle and funny and works great in a very interesting video like all of yours are.
Those PCB transparencies are great. First time i see then in a video.
Nice video editing skills.
Ikr, he really has some huge passion on this
My very first CD player was an Emerson from the very same period as your Magnavox and I remember being so amazed by the thing that I played the same CD over and over again (because I only had one for a long while) in complete awe. I had an old Soundesign stereo and I simply couldn't believe that the old thing was capable of producing the kind of sound that was coming from that CD player. I've been in love with the CD ever since. Thanks for the great videos Alec, I absolutely love your work!
19:19 Mr TC, thanks for the fact, that you didn't explain it in a technical way, but gave to the audience some simple and basic explanation .I already started to have a severe head-ache.
I just love the sillyness. Your channel has evolved to be one of my absolute favorites!
It is spelt silliness
I am an audiophile and love CDs, have tons of them. I still consider CDs the greatest leap in the audio industry (vinyl fans weeping here) and even though I have DACs and TIDAL streaming the sound of a good CD player is still fantastic. I also love my LPs and analog sound when done finely, but I have to admit I hate the noise, the lack of dynamic range, the expense and the excessive care they demand. Amazing explanation. My respects!
Actually cd's are also fantastic quality and are also uncompressed
Real audiophiles never used vinyl. They used reel to reel tape. That's the audiophile format, not vinyl _OR_ CD
Being a vinyl record enjoyer (and compact cassette... and reel to reel when I finally get around to buying one!) doesn't mean I don't absolutely adore the crystal clarity of a well-mastered CD. That format offers a whole host of benefits, which were unfortunately mostly squandered by the cretins who engaged (and still engage) in the loudness wars. So many potentially excellent albums completely ruined 😢
@@duffman18 real to reel tape is very impressive. Where do you think most of your CDs originated from in the studio? Yep real real is awesome.
Next time you see a CD that says AAD OR ADD The a probably came from real to reel as the analog component.
The components on the PCB are a wonderful mid-era mix of through-hole on the top, and surface-mount on the bottom. It's quite interesting to see - a few years earlier and none of that SMD would be on the bottom; a few years later and nobody would be doing through-hole and it would be all surface. Yet this board sits right there in the middle between the two.
As you are very aware PCB is short for Printed Circuit Board
I have recently started collecting CDs again and recently acquired a 90s Panasonic Discman. Sound quality and build quality is amazing. Such an interesting video as always! Love your humour too
I can't get any more specific, sadly, but this channel brings back vague memories of educational programming on PBS back in the 80's - and I love it!
Brilliant content as always. Mesmerizing to watch. The level of detail you put into your explanations is legendary, and your humor and enthusiasm keep the interest high!
The reason for the long plastic lever connecting the front panel power button to the actual power switch in the back is to maintain separation between the mains-powered part of the electronics (the power supply) and the low-voltage portion (the rest of the circuit). This is for two reasons:
a) Safety. You want to maintain a good amount of separation between any high voltage and low voltage sections. Having a high voltage wire running from the power supply in the back to the front panel breaks this separation.
b) Noise reduction. Having a long wire carrying AC voltage can easily introduce hum into the audio, so it's best to avoid it.
Modern electronics often get around this by having the power switch not actually switch the AC voltage directly but instead control an electronic switch, but this can lead to "vampire" power usage since the circuitry to handle the electronic switch uses up a small amount of current even when the component is switched off.
Huh. Hadn't thought about the noise! I have a cassette player with the power switch actually at the front of the case, with incoming power routed through it. I didn't really see the harm in doing it that way and figured the plastic Fing-Longer was some sort of cost cutting measure.
@@TechnologyConnections so that's what would happen if I had invented the fing-longer! Ah a man can dream, a man can dream.
Analog Oscilloscopes used to use these kinds of rods and mechanical parts to control most of the switches and pots mainly to have very short wires and PCB tracks which improves performance in general.
That actually reminded me of this:
ua-cam.com/video/vZgem0EdCBQ/v-deo.html
The whooping noise at ignition is sympathetic current in the cabin PA system. It's caused by the system that recharges the capacitors that provide current to the igniters. The two circuits are right next to each other, and run the length of the cabin.
Main reason: cost
This was the first video you've done that melted my brain. But you explained it and I watched it twice (with ad block off) I got it!! Good work!
"This is a longer video." ...do I hear a dishwasher and heat pump laughing from the not-too-distant future?
Awesome! this is the best and most detailed explanation of how a CD player works I ever seen,on other videos they just open the unit up and make "educated" guesses, example "these large chips are probably the DACs" "these small ones are probably power supply regulators"
just a bunch of nonsence.I'm impressed by how small the DAC chip really is since it's often given all the credit.And the good thing about old electronics is that they had discrete chips for discrete functions nowadays they just place everything in one small chip and call it a DAC even though it's a lot more than that.
DAC-Digtial to Analog Converter
You mean nonsense
Also,a CD Player would not work without a DAC,witout a DAC CD Players would be totally useless
Are you ever planning on a video about telephone switching technology, phreaking, etc? It would go nicely with the telephone video and you could explain the history of the Strowger switch, Panel switch, and/or Crossbar switch technologies. Perhaps you could even pay a visit to the Connections Museum in Seattle.
Oh indeed, he needs to do that, Evan Doorbell's tapes are an invaluable resource, but of he wants to do it, he better start listening to them now, the information isn't condensed in any way.
Yes! THis would be awesome!!
Yes please do this. My favorite is when Bill Gates found the telephone system manual in his university library and used it to phone freak. I can confirm that in my university library, there was some titles that shouldn't be there.
Yes indeed, the old BSP(Bell System Practices) documents, they explain everything from how a 4A toll switch works to the MF signaling system and how to sweep the floor of a central office.
could begin a good segment talking about the evolution of circuit switched networks into packet switched networks
I appreciate your willingness to go out on a limb with your educated guesses about the functions in front of such a huge audience, and also the large amount of work you must have put into this video, not just disassembling and reassembling the device, but actually going to the trouble to trace solder runs from pin to pin, which is no small headache considering you have to keep it all straight as you flip the board over from component side to trace side. Good on you.
Interesting that you decided to dive-in in that Philips player, although it is a "basic" player, it's more complicated than a Sony unit in my opinion because Philips made a set of chips to work together and they don't make much sense when viewed from outside or read from a schematic.
Philips put a player together in a way so dissimilar to the Japanese manufacturers like Sony, Sanyo or Toshiba.
What you got there in that Philips player is the two analog servo chips (TDA8808/9) for the tracking and focus control. They are connected to the Microcontroller (The Motorola chip, CPU from now on) because they output some error detection signals the CPU has to take into account when playing.
The TDA7210 is the decoder that takes the raw analog data as read directly from the disc (The HF signal) and digitizes it when it goes through the "data slicer", then it reverses the EFM, does the CIRC error correction and masking then it spits out the audio data to the DAC(s) and the subcode information back to the CPU (to indicate play time, beginning and end of a track, remaining playtime, etc.).
The reason why you didn't find any information on the CPU used there is because Philips asks for a bunch of such chips, already programmed, and relabeled from Motorola. They're HC08 microcontrollers and can be found in almost every Philips player based on the same "chipset", because they produce the chips and code to work with them on that architecture then they adapt the code to work in the application they require. For example I have a Philips FW26 system with a single tray player based on a new-ish implementation of the same design and it has a small QFP cpu with the Motorola logo and the label "CD SHORTL", it doesn't control the display like on your player but it takes commands through an I2C bus from the main CPU on the system.
A note on the track seek and subcode, there's no standard way of accessing a particular track but the players can be as smart as the manufacturer wants them to and there are a few strategies.
One is doing a raw calculation of how far the sled has to kick to position the laser "around" the track then do a slow seek, a "smart" move (observed in Sony players).
Another is doing a 10 track skip seek and decoding whatever frame the CPU can catch under the beam until the beginning of the track is under the lens then doing a fine adjustment with 4-2-single track skips.
This last one is the "dumb" one. Skip track by track looking at the P subcode. The P subcode has just zeros and ones to mark the beginning of each track, the CPU can just count how many "marks" it has skipped on the P channel and land exactly at the beginning of each track.
Also, thank you very much for the credits, I genuinely love the CDM4 mechanism.
I have to say, your technical input is absolute gold. As a music lover, I seem to have unknowingly moved towards the cdm4 as it, or at least one of its close variants, is fitted to 4 from a total of six of my players. The alarming fact is that while cds are still being pressed in their millions, replacement cdms are nla. I still really like it as a music media.
Thank you for your input.
I love how ridiculously deep you're going into this. And I love your narration and writing. Also the performance and editing. Long Live Technology Connections!
Interesting and informative video. Well done.
Also... "It was just a dream Bender, there's no such thing as two" (11:47 if you're wondering)
Your ability to explain all of this intricate technical detail in only a 20minute video is astounding! Good job! Your videos are so good, surprised you don't have more subs/views than this!
This Video IS actually 21minutes21seconds in duration
You are a freaking genius! I watched tons of your videos but sometimes I got, somehow, lost. But this time I completely got out of my mind... Ain't saying I'm stupid, on the contrary that I never have meet in my entire life, someone so committed to understand technology (besides people working in the field).
You are definitely in my top 3, alongside Techno Moan and 8-bit guy! You are AWESOME and I hope you guys could cooperate together to make yt a sort of Wikipedia of audio/video systems. History and explanations of your knowledges combined!
Alex Root lol he’s been talking slower! back when i found his channel watching a video would last an hour cause i had to keep rewinding. i just couldn’t keep up!
I can tell that all of your research has taken TONS of time, and you have presented in a very concise, to-the-point way that makes it very easy to understand! Thank You for all your time and effort!
Thank you sir, for giving me information that I don't need but I watch every minute of every video. I just like to know what makes things tick. I'm a techie and always have been. I was a Crypto tech and a computer tech (not PCs) in the military. I really appreciate the level of detail!!
Makes me wonder what kind of computers the military uses, bet you can't say much but still wondering.
Also heard about that time when the PS3 had just came out and there was rumors of the military buying a lot of them to work as a supercomputer clustered
How are you so smart! Not only do you research all this info but you are able to put it into a concise and a comprehensible video. AND edit that video, creating a great watch at a perfect length. Bravo ol' chap!
First three words: "This exploded machine..."
YOU HAVE MY ATTENTION.
I can't decide whether I'm more impressed with the in-depth understanding of how all of that works, or the editing and effects, especially highlighting the board traces. Amazing work!
I barely understand what he is saying because of my poor knowledge of English, but this channel is awesome, and I like watching his video
This was fantastic, as someone who used to pull things apart in an attempt to see how they work, and still does so on occasion, I really, really appreciate this. Thank you.
CD-ROMS were definitely possible in the early days of Audio CDs, but there was really no use for them. Home computers couldn't play videos, animations or even display still images with more than 16 colors.
We didn't even need a full CD-ROM standard. Simple Audio CD with a way to send digital stream directly to a computer would be good enough. Much faster, convenient and reliable than audio cassettes used by ZX-Spectrum or C64. There was no need for such a capacity back then.
Professional use could use any amount of storage, but only if it could read and write to it. It's a bit similar to Laser Disc vs VHS, where superior image quality wasn't that important, because VHS allowed for recording at home.
Even a decade later first CD based consoles like Sega CD or Amiga CD32 weren't utilizing such big starage besides CD Audio or choppy video.
At the same time CD-ROMs were gaining popularity in PCs. Software distribution was cheaper, since single CD was cheaper and more convenient than 10-20 floppies. And CD/DVD-ROM era started.
CDs are brilliantly designed. Technology that is in use for almost 40 years. Of course DVD and Blurays took over in many aspects, but it's still the same technology, just with improvements.
Of course nothing then would have prevented text only, command line software from being stored on a CD. A single disc with 32x the capacity of my (upgraded!) IBM PC-XT would have been wild then.
The Atari 8-bit family is underrated in this regard. It was so forward thinking people didn't really realize the specs didn't change much until end of fabrication in 1992. 256 color, 384 lines, 1978.
@James Slick
Yes it would have been wild but also pretty useless.
It would be a bit like trying to distribute today a read-only storage format with 500TB of data.
@@1337Jogi I know it wouldn't have been practical, just possible. :)
Still crazy to think about that.
No one has ever explained CD working to such great detail, and so simply. Still my favourite channel.
Do not stare into the beam with your remaining good eye
e4r haha
if he failed this advise he won't know cause he has no eye left to be able to read this
soundspark, Are you sure? I heard that CD players are classified as Class 1 laser products *only* because it is enclosed and the light does not escape in *normal* use. I don't think it's safe to look at the diode directly. Please correct me if I'm wrong.
Bagana I am since I DID stare a LOT into a CD laser when I was a kid and have no vision problems at all :P Didn't know it was infrared back then. For burning it's a whole other story, but you'd be hard pressed to get that to activate without a disc being properly detected without serious hacking. What turns on at disc insertion is obviously the read one.
Nor should you look directly at the ghost trap.
This is insanely good. Thank you. I feel like I have at least a big picture idea of what it actually takes to read the optical data and produce audio at the end.
In the last video *SUDDEN HEARING TEST* we talked about
The audience is now deaf.
@@CaveyMoth WHAT? :D
@@mclatchyt [Types in sign language] urdef
that was fantastic. you mentioned before that tracking was done on the hardware level, i'm glad you've expanded on that and so much more here. please keep it up
I'll be honest, this mostly makes me even more impressed about the whole LaserDisc thingy - how did they get the "warp" and "off-center" cancellations (focus and track alignment) correctly while having a disc that spins so much faster?
not to mention laserdisc was analogue so it didn't have the luxury of digital error correction
Your channel blows my mind! Your research is insane! Keep this up! This stuff will be fascinating 1,000 years from now!
I like my Sony CD player. The laser is fixed and it is the disc that moves back and forth. I like the madness of it. Yeah, lets move the rotating disc instead of the laser! It's supposed to be better than moving the laser, but it is also Sony telling you it is better.
How is that better? The disc and motor must be much heavier than the laser assembly. Gosh, sony at it again.
@@autogolazzojr7950 I don't know that it is, but it is MUCH easier to clean the laser lens. I use a 20es if you want to see the mechanism.
I have watched the sound and video series all the way back to the beginning, and I am very appreciative of how much I have learned about both audio and video. I have always wondered, but found most explanations to get too complicated too quickly and never had the motivation to really study it. I now find myself giving brief explanations to people about things like how CRT monitors worked and how the laser reads a CD. Thank you for explaining throughly while keeping it simple! Always excited to see the next video.
8:22 This line seems to be to indicate data being sent from the TDA8808T to the SAA7210P. The data being sent goes over TDA8808T pin 3, labeled HFout, and goes to SAA7210P pin 25 labled HFI (High-Frequency Input).
The lines between the processor and the decoder you marked send the information in the Q-channel to the processor.
I found this information in the datasheet for the SAA7210 (non P) but it appears to be a compatible chip.
Also the AGC is not a ADC. The TDA8808T still outputs an analog signal to the decoder.
Aww dang, why didn't I drop the P? That would've brought me where I wanted to go.
Admittedly the leap from AGC to ADC was an assumption based on the fact that two devices were hooked into the focus chip. I figured its probably a binary stream by that point, but I stand corrected! In the next part I'm sure we can find some more details with the scope!
Typically the “P” suffix in chip part numbers will denote that it is a Plastic DIP package (versus, for example, ceramic, mil-spec, or other package style).
really love these videos; electrical engineering was something I seriously struggled with in college despite wishing I knew how all these devices worked. It's been really nice to get a systematic explanation of this stuff in a way that still keeps my attention long enough to absorb it.
Did you ever talk about your background education regarding electronics?
Just saying how you learned about might be as entertaining as this show.
And maybe Mr Technology Connections can say what his name is too? :-)
Just be glad he shows his face and does not use a robotic generated voice to speak his script.
ua-cam.com/users/edisonphono1
Thanks!
Oh god. You found that.
Fantastic episode. I really appreciated the main board breakdown as I've never seen that before.
Legit. Try getting your hands on one of the Sony YEDS test discs especially made to generate tones for use on an oscilloscope. I found rip of YEDS-18 a few years ago on some random forum. A bunch of the tracks are mastered with 'emphasis' for early players with 14-bit DACs.
I saw a picture which made me think of how a cd player works, specifically the "anti-skip" technology. Glad to find your video!
Did I mis it, or did you forget to explain the Sony solution for keeping the laser on track and in focus. (fascinating episode, by the way. Keep this up! It is much appreciated!)
Coming soon to a screen near you. Freakishly near you. In fact, quite probably it's the *very same screen* you're gazing at now!
_X-files theme plays_
13:57 This is the replacement for vinyl, people. We. Have. Got. To. Get. This. Right!
Hi, can you speak about the "flags" of the CD. I think they are in, or around, the TOC. They are also ver interesting and forward thinking. There are multiple flags, my favorite is the Emphasis flag. (A type of noise reduction). It alerts the player to engage its Emphasis circuitry. Since only some of the very early discs had (pre) emphasis, most players have never powered on their dedicated emphasis chips, yet compolsory for every CD player to comply with the red book standard. There are other flags too. Some never assigned and to be used for future planning (like how you mentioned the extra data storage, etc!)
I'm giving away part 2 but.......the extra sub-code tracks were used for CD-Text and CD+G data storage.
oh you mean sonys root kit?
The SP-DIF protocol includes a flag for pre-emphasis also, and outboard DACs are supposed to support it, thougn not all do. I was once debugging a Linux audio driver where this flag was being erroneously set, causing everything to sound rather muddy and flat, but only through one of the two DACs I had on hand.
One of the few channel's I get giddy over when I see a new video in my feed. Thanks :)
That PLLOH signal could be used as a status that the CD isn’t spinning, or not locked. and either mute audio digitally, or simply stop the controller counting.
I just want to thank you for the amazing entertainment and wonderful explanation you always give and your dry humor is very much welcomed. And as a fellow IL resident (Central IL) I appreciate the little things here and there that are IL unique.
I see you, hiding a 2 in the bitstream!
I was wondering if that was an easter egg.
After wondering exactly how Audio gets from a little plastic disc to my ears for almost 30 years, i finally find someone who can explain it in terms I can understand. I love this video, I love this channel, and I can't wait for more.
You will agree that CDs are fantastic and they are also known to extremely durable and even with heavy use will last over 40+ years
I think you are wrong about the microcontroller doing EFM decoding.
The chance of any microcontroller from the 80s being fast enough to handle EFM decoding is near-zero. The chance of a microcontroller being fast enough to also handle buttons and a display at the same time is zero. Besides, the datasheet for the SAA7210 shows it does EFM demodulation (and everything afterwards).
Now your theory about decoding up the Q channel has some merit; An expert programmer could potentially design a cycle accurate state machine which pulled the correct bits out of the datastream. However, the SAA7210 datasheet shows it has built in subcode decoding, and even has a nice slow microcontroller interface to send that to a microcontroller when requested (those are the two pins you showed connecting between the two)
No. I suspect the laser bitstream is going to the microcontroller for a much stupider reason: To detect if there is a disk in the tray or not.
Even the slowest microcontroller can sample the bitstream a few times at random. If the random samples are all zero, or all one over 10+ samples then probability states that there is no disk (or some other disk reading error).
Wow, really well done! This is probably the easiest to follow visualization of how a CD player works. You sir, have a knack for this.
This video is pure nerdgasm :x:x:x
This is an excellent explanation of how a cd player works.
A fantastic video, as always! Unfortunately, I think there's a small mistake in there:
The Motorola microcontroller likely doesn't have anything to do with EFM. The SAA7210's datasheet states clearly that it's responsible for EFM decoding, not some other chip in the system.
The mysterious trace that's shared between the TDA8808, the SAA7210 and the uC is this likely *not* the EFM signal -- it looks like its connected to HFD (input) on the 7210 and PLLH (output) of the 8808. If so, it's just a status signal.
7210 pin 25 (HFI) is where the EFM money is at. Sure enough, it's connected to pin 3 (RFout) of the 8808.
damn, made same comment 30 minutes ago :) should of read all of the comments first
Indeed Philips made all the hard working chips for the decoding. The SAA7210 datasheet shows it doing everything from RF decoding, CIRC, error concealment and interface to DAC. And sending subcode data somewhere (the Motorola).
and the RAM is a buffer - this chip also adjusts motor speed if the buffer is too empty or full.
It seems the CIRC decoder has its own RAM even. The data sheet has a really nice explanation of the workings, even which part of the RAM is used for what data and how the motor speed control operates. Or that a single value flagged as unreliable by error correction is interpolated linearly.
It really seems that Motorola MPU is at the end of the pipeline - not somewhere in the middle.
Society owes you a tremendous debt for the great work you've been doing in preserving and demystifying this stuff.
Your videos are part of my kids' curriculum.
Oh fuck that DAC chip in particular. I never got the stupid thing to work right. I was questioning my sanity many times playing around with it. Still don't know if I'm that bad at coding the I2S thing or I just got damaged parts (they are apparently 30+ years old lol).
Once I had the cover off a CD player because it was misbehaving. Set above the bedhead I realized I could see the reflection of the lens while spinning. I was blown away as I watched it maintain what seemed a perfect distance from the disc as it wobbled (as mentioned no disc is perfect) while it spun. I was mesmerized by the beauty of the control system doing its thing.
Where's the comparison between the Philips and Sony tracking mechanisms? Or did I miss that?
Its on side B of the video.
I think the Sony is linear.
I did not mean to imply through saying "later on" that it would be later on _in the same video._ I need to get my hands on a cheap CD player that I can happily destroy.
Get a comparable player to this one (late 80s, early 90s). It doesn't make sense to get a new boombox for that as the chipset used in those is ridiculously "compact" to say a thing. It's just one piece of silicon that does everything.
@@TechnologyConnections if you could do it without destroying it I could offer up my first Sony single tray load CDP-670 I got from Crutchfield I believe in '89. Just replaced belt last year & works great. Just not sure it's worth the shipping vs just grabbing one at a thrift store.
Thank you very much for this more technical video about CD! I thought I knew everything about how CD worked and never realized it's possible to dive deeper. Looking forward to the next video about CD.
I thought it would be strange to split 8to14 modulation decoding from CIRC processing to 2 chips. And the SAA7210 datasheet concurs: it does demodulation, CIRC, and some more tasks, and outputs I2S data to DACs. The input data stream processing is definitely not split between the 2 chips.
Oh, yeah, and it seems to extract the Q channel that is probably used by the Motorola chip.
Perhaps the Motorola listens in to the raw data to count how many tracks are skipped. Or just to check that signal is being detected (as part of disc detect).
I really appreciate this video. A great deal of thought was put into the design of the CD format - with a special emphasis on error-correction features - which were implemented in multiple ways. The convenience, robustness and fidelity of the CD format explains its popularity and longevity. The only thing they could have improved on, in my opinion, would be the inclusion of CD-TEXT and album artwork - from the get-go. Having the album title, track title, artist name, and album artwork included from the very beginning would have totally blown people away in 1982, and further accelerated the demise of the cassette and LP formats. Today, any CD-Rs I make always include CD-TEXT, but sadly, most of my legacy CD players can't display it. At least the CD engineers were smart enough to include unused data areas for this, and other features. Carry on!
How does Sony do the tracking?
And: the reason for the switch in the back is simple: it cuts the mains voltage, so the transformer itself is turned off. And you don't want mains voltage near the front panel, it might interfere with the relative sensitiv ICs and with the audio. So it's good practise to separate the voltage and letting mains voltage only to the absolutely needed parts.
Great video, BTW!
Sony does the tracking using 3 laser beams instead of one, and two more photodiodes in the detector to read back such extra laser beams.
The main beam reads and it is centered on the main track and the two others, smaller ones, are stepping, off center of the adjacent tracks.
The signal from those extra beams is monitored and when it is equal, the lens is on track, when there's a difference, more light on a diode than the other, a TE signal is generated, amplified and tracking is corrected by the servo chip.
I loved the amount of dense, technical data in this video. I found it very accessable.
As a branch off of this video, I'd love to see a video on CD-Text and CD-G (CD-Graphics) tech, since it uses the subcode on the cd.
Right down into the subcode. Such detail, yet clearly understandable presentation, thanks.
In the next video could you explain how do the CD burners and the CD-R and CD-RW work? Please.
I don't think I have learned anything that I didn't already know (cause I am a dork) , but it doesn't stop me from going OHHHHHHH whenever one of your video's show up in my subscriptions
Can you make sure you go over the Anti Skip CD players that you could bump really hard and they'd never lose track of the audio?
They just have a bigger RAM chip.
William J. Some even incorporated a visual display of the buffer’s depth such that you could see the bar graph rise and fall depending upon how deep you dipped into the buffer. I seem to recall the buffer depth being around 8-10 seconds on some early units, but that was all a very long time ago.
They were double speed players that read music information into a RAM chip, asynchronously from the DAC output.
I have to watch all your vids at half speed to understand fully , but I appreciate all the stuff you teach me, thank you 👍
It's amazing to me that this technology was successfully developed as long ago as it was.
It wasn't that long ago, you young whipper-snapper. Now get off my lawn!
@@pauldzim it was almost 40 years ago, which is an eternity when it comes to technology. It's really an astounding feat what they pulled off with CDs, given the technology constraints of the day.
Ever since I saw this video and this amazing Magnavox player with the Philips swing laser CDM-4 mech, which I didn’t even know existed (being a kid from the early 2000’s), I got into a life-death task of finding a player like this. I was already looking from a player from the 1980’s, because of their sound signature, that I quite enjoy, but this made me obsessed. Finally got me a Philips CD-610 a couple of months ago, which is, internally, the exact same machine. It sounds great too!
Edit: I was one of the sneaky eyes that caught the leaky capacitor in the power supply. Mine has that exact same leaky capacitor, even tough I have no plans of replacing such leaky capacitor, as the leaky capacitor seems not to be affecting the board really bad the way some leaky capacitors do.
If I had your videos as a Kid... man this is nerd porn for me. Keep doing and hailings from a fan from mexico.
The information density on this one is amazing. I'm gonna have to come back to this video when I'm more lucid. :)
I still use CDs this day and I still make CDs to
Better than just buying an album digitally. You get CD quality audio, a physical copy of your purchase and it's only a little more expensive. Plus you can make a MP3 on your PC.
Clay3613 you are right about that
Same here, loading and watching a CD spin is marvelous.
Yeah, I've always loved CDs for the fact that you can get a physical copy of an album in (presumably) high-quality audio- higher-quality than vinyl records, even, if they're mastered properly- without having to worry about the disc wearing down with each passing play. Digital downloads are good for listening on the go, but when I'm sitting down and want to listen to some music, I'm gonna go for a CD 90% of the time.
It's honestly a shame though that CDs could never recover their old momentum in the same way vinyl records and cassettes did; I blame record labels' continuing tendency to volume-boost and over-compress their digital masters because of the myth that louder music is somehow inherently better.
game4brains 123 I again with you man
This channel keeps getting better and better.
*_i think there is a leaky capacitor in the power supply section_*
This is so amazing. CD technology seems so old these days and yet is so far beyond my ability to understand. Thank you for all your hard work!