-
Hello world,
Among the 20 old cheap broken CD players I bought the other day.. is this little thing. I bought it for 10 Euros, a Radiola CD1371. Looked like a cheap random player to me, but bought it just because somehow I found it cute, and wanted to diversify my experience, and also it had an uncommon fault : seller said it was working/playing CD's fine, but the display was not displaying anything, never. So works, but going blind...
Then when I popped the hood, was surprises to see a swing arm in it, a CDM 4/11, along with a TDA 1541A DAC !
That's when Gyro told me that this little thing was actually nothing less than an old Philips CD 371, pure and simple, simply wearing a "Radiola" badge... because, I learned soon after, Radiola was actually a brand owned by Philips. So makes sense. This old Philips is worth a lot more than the 10 Euros I paid for it.. even found one on Ebay for 500 USD / 400 Euros, yikes !
Could really do with some cash at the moment but... I am not going to sell it for a profit. I am now very fond of this thing, especially know I know it's a technically interesting player. So, am keeping it for sure !
So... ater testing, here is what I found :
1) Display does not work indeed, completely kaput. Gyro already told me it was a common problem on this player. The couple 7 segments LED displays, are driven by decoder which sports a serial interface to the CPU on the main board. Decoder goes south but the displays are fine. IIRC, Gyro said the faulty decoder chip can be replaced by an off the shelf decoder that's still available today ? Is that right ?
2) Unlike what the seller said, the player won't read CD's. Seller is surprised, said it was working fine "last time" he tested it... now "last time" can be ages ago I guess... but regardless of how long that may be, I guess at least we can assume that if the pickup was good "last time".. it should still be good now, since it has seen no use/wear at all during that period of rest. The guy really was very decent, I will give him the benefit of the doubt, at least until proven otherwise, of course...
So for now, we will consider that the problem could be due to
A) The storage period causing trouble : electrolytic caps gone bad ?
B) Physical damage during the transport.
Here is what happens. Initial conditions : no CD present, tray closed, player at rest, and CD clamp/top part removed so I can see what the pickup is doing.
Then I press the PLAY button and here is the sequence of events that that I witness, i chronological order
1) Something strange from the get go : the tray motor runs for one second, in the direction where it wants to close the tray, which is already closed.....
2) Laser turns on
3) Spindle turns for one second or so, a CD revolution or so, while the focus servo goes up and down full range, once. Trying to focus then, I assume...
4) Spindle stops for a split second
5) Spindle turns again for one second or so, while the focus servo goes up and down full range, once... I guess performing a second attempt at focusing.
6) This fails so spindle stop, and laser is turned off.
Then if I press PLAY again, the whole sequence is carried out again.
In the video below, I press PLAY a few times. A disc is present so you can't see the pickup, and the disc does not stop when spindle is de-energized, due to the inertia of the CD of course, but if you pay attention you can see the CD slow down a little bit then speed up again when the player does its second attempt at focusing.
You can also hear that the CD is rubbing a bit.. don't know where it comes from.. the clamping arm I guess ? Not sure it would be enough to cause the player no to read the disc ? I mean the disc looks like it's turning quite effortlessly no ?...
Still, would be cool if it didn't rub at all of course, but I don't know what I can do to fix that, in practice...
Should I try to fix this before the fixing the display, or the other way around ?
I have a Toshiba with a CDM 4/11 transport too, so I guess plug and play.. could swap transports to at least be 100% certain the problem comes from this, rather than some problem on the main board.
So here we go... let's try to save this cute little Philips shall we....
-
Yay, the impatient little bugger gets his wish at last!
I'd investigate the power rails first. The failure mode of the LED display is usually to lose the odd segment, not blank, which makes me suspect an issue with the +5V rail.
The disc seems to be starting and stopping fine. You occasionally get a scraping sound from the disc clamp if the white nylon cage catches the edge of the clamp disc. Nothing serious, it just needs a little grease. If the clamp is badly worn then it can cause the transport to vibrate, but I don't see / hear that. It's a good idea to lubricate the hinge points and clamp with a little plastic grease.
It might be worth a gentle clean of the lens too (at least this one has one!).
P.S. The case looks in pristine condition, always a good sign (the PCB too). -
Another thought. That Axial 47uF electrolytic (2546) located underneath the transport is notorious for failure, causing laser power stability issues. I would replace that immediately (as you have the transport out). A radial part is fine as long as you lay it flat.
This could easily cause initial focus lock issues. Never attempt to change the laser adjustment without replacing this capacitor first.
P.S. You can quickly replace the electrolytic without removing the PCB, just snip the existing leads at the body and tack on a radial one. Just don't apply heat for long enough to loosen the PCB solder joints. -
Yep Gyro knows his classics!
To me it sounded like a slipping sound in your video, I'd check that the clamping is correct. I can't remember for sure but no doubt Gyro will chime in, I think there is a metal ball pivot between the white plastic and the black clamping disc.
Maybe check that isn't lost or seized in old grease.
Other than that usual checks... Swing arm swings freely, state of flex at the back plastic clamp on the lower right of the transport. -
Haha, thanks shakalnokturn.
Yes, there is a metal ball on the top of the black disc clamp shaft which bears on the underside of the main sprung flap. The white nylon bit is there to hold and centre the disc clamp when it is released and when it is dropping down onto the spindle hub and CD. If the two pivot points between the white nylon carrier and the main flap are a bit tight, it won't sit level and will slightly rub on the disc clamp. Everything just unclips.
Earlier transports used a slightly rounded blunt plastic end instead of a metal ball. These tended to wear badly and cause the whole transport to shake.
EDIT: Just reading back to the first post - Don't swap in the Toshiba transport at this stage, best to break one at a time.
Another Edit (I miss a lot of questions):Quote1) Display does not work indeed, completely kaput. Gyro already told me it was a common problem on this player. The couple 7 segments LED displays, are driven by decoder which sports a serial interface to the CPU on the main board. Decoder goes south but the displays are fine. IIRC, Gyro said the faulty decoder chip can be replaced by an off the shelf decoder that's still available today ? Is that right ?
The shift register/decoder is built into the display. It is easy to replace with a decode chip and standard common anode LED display though. The decoder is a (NS/Microchip MM5450 / ST M5450). -
Thanks guys for your input ! I knew I could count on Gyro's love of this little thing, and Shakal's total inability to go bed before 4AM !
Gyro, it's too late : I have ALREADY swap transport with the Toshiba, see picture !
At first, before doing that, I tried easing the pressure on the disc by removing the spring that holds the clap down. Not much difference...
Once I had swapped the transport...the Philips worked just fine with the Toshiba guts....
So... bad news. I was already so much in love with that beautiful piece of art that is this white cast resin CDM 11/4, and now it's dead, probably
Well I guess there is still a faint ray of hope for the white one.. maybe ?
Shakal did say to make sure the swing arm was moving freely and.... actually I did compare the two transports and did notice a difference. The black one, which works, has a significant " springiness " to it, feels healthy, iif I shall say ? I mean if I pull it with my finger to its outer stop, fully outward, then release my finger all of a sudden... the pickup races to the inner stop and makes quite a loud "bang" as it impacts (sorry for brutalizing it, though). Also, if I hold the pick-up by hand to the inner stop, then release my finger.. the pickup will stay in place, but the white/defective one will bounce a little, move back 10mm or so, and stay there, strange. The white one is more sluggish, feels tired. Not as responsive or bouncy. I can feel the black one has good force to it, whereas the white one is very weak, like it was 100 years old. Nothing against 100 year olds, but the black one in comparison feels like it's 20 year old..
I see no actual spring in the CDM, so t's just the magnetic magic that gives it its springiness, even when powered off ??
So how does one proceed to "clean" a swing arm ? Don't want to mess with the thing and ruin it, out of ignorance....
Can it be safely dismantled, then clean old grease if I see some, and add some new grease.. .which again I don't have so will have to do with machine oil for now, at least for debugging purposes. Worked fine on the Pioneer...
Clamp : no metal ball anywhere. Both the Toshiba and Philips are constructed 100% the same way. Below pics of the Toshiba clamp as an illustration.
There is just that white plastic thingy that keeps the spindle clamp in place while allowing to move, and also has some spring action to keep it centered.
Both clamps look in the same condition / state, both move very freely. Both have remains of very dry and crusty grease, for sure, but still appears to move perfectly freely none the less.
Condition of the Philips : yep, all the white plastic parts are like new, bright white, really like new. The Toshiba OTOH, all the parts are significantly yellowed. Maybe the owner was a smoker...
I see that there is a small PCB under the swing arm, so there is some hope that the problem comes from there rather than the pick-up itself...
But first, I need to give the swing arm its springiness back ! Tell me how I should proceed to do a good job of it, what tragic mistakes not to make if any.
Making progress anyhoo !
Here is a video where I compare the "health" of the two swing arms, one after the other. First the black one, that works, then the defective one.
-
Another Edit (I miss a lot of questions):
The shift register/decoder is built into the display. It is easy to replace with a decode chip and standard common anode LED display though. The decoder is a (NS/Microchip MM5450 / ST M5450).
Thanks a lot for the details. So the displays do need to be replaced, was mistaken. But they zare nothing special as you said so should not be difficult to find an appropriate replacement... come to think of it, I do have a bunch of brand new green 7 segments LED displays in my drawers... 25 year old now, didn't think I would get to use them one day ! Will check dimensions and polarity...
BTW, when you edit your past messages I don't get notified, so I miss the nice new content... spotted it purely out of luck, as I glanced over previous messages just out of... I don't know what, but I did....
-
Forgot : did clean the lens on both transports, at the very beginning of the troubleshooting. Made zero difference. But, at least I know I can rule this out of the equation, which is always nice. The less variables there are, the happier I am.
OK almost 11PM, going to bed. See you tomorrow evening for more swing arm action... -
The white CDM isn't necessarily dead, certainly give it the benefit of the doubt. The swing arm does look a tiny bit stiff, not too bad though. The springiness actually comes from the flexi-print ribbon, there is no physical spring. The white movement is closer to what I am familiar with actually.
No don't dismantle the swing arm suspension. If you release the clamping leaf springs (with the ball in the middle, it will affect the vertical angle of the swing arm (it will be at different height at each end of the swing. The service manual does have the alignment procedure, but is involves 'eyeballing' it with a transparent disc and a linear light source (like a florescent tube). I've never done it, and never quite understood it from the picture.
It is normally sufficient to lubricate the ball to swing-arm pivot points with a tiny drop of silicone oil as Philips recommend. A small amount on the the tip of a darning needle for access.
As I mentioned before, it might just be a dirty lens, or the laser drive is marginal due to that electrolytic capacitor.
Most CDM4s are black by the way, the earlier ones were just using up stock of the CDM2/10 white mouldings I think. -
Forgot : did clean the lens on both transports, at the very beginning of the troubleshooting. Made zero difference. But, at least I know I can rule this out of the equation, which is always nice. The less variables there are, the happier I am.
OK almost 11PM, going to bed. See you tomorrow evening for more swing arm action...
Haha, good progress for one evening. Good night.
-
Clamp : no metal ball anywhere. Both the Toshiba and Philips are constructed 100% the same way. Below pics of the Toshiba clamp as an illustration.
There is just that white plastic thingy that keeps the spindle clamp in place while allowing to move, and also has some spring action to keep it centered.
Both clamps look in the same condition / state, both move very freely. Both have remains of very dry and crusty grease, for sure, but still appears to move perfectly freely none the less.
The metal ball is on the top end of the spindle clamp shaft (you're showing the bottom). Yes, the three springy arms are to centre the clamp so that it goes down smoothly onto the hub and disc. Dry crusty grease is normal at that age, it's enough to just make things a little sticky. You can gently pop the white carrier out of the main flap. You'll then see the ball and be able to clean up better. -
Not in bed just yet ! Within 5 minutes though !
Wanted to add a quick little something / erratum before I call it a day : I realized I fucked up my springiness test ! The black one, plugged into the Philips player.. was actually powered UP !! If I remove power, it behaves just like the white one, but still feels much stronger/healthier...
So the springiness is not magic after all , just the flat flex ?!
Feeling stupid......................... 
Lens : I just wrote above that I cleaned them both long ago, with no improvement at all, don't you read ?!
Laser blue cap.. hmm yes good point, might be good enough to get a healthy laser going, but maybe not good enough to get a weaker laser going... will definitely look into that indeed, I promise ! I don't have a new cap to replace it with though.. only 30 year old salvage caps, which are likely just as bad or worse... no ESR meter at hand to actually check for the condition of the cap. Yes, I am looking into getting a tester in the short term, I am fed up not being able to know what I am doing with old caps, I am sick of it. Will try one of my old caps anyway. Won't be new but it will probably be a different state of bad, so might manifest as a slight difference in overall behaviour of the drive, which would therefore confirm the problem and motivate me to spend 2 hours riving to town center and back, to go get a new cap at my local electronics shop...
Or maybe if the value of the cap is not that high, I could cobble an equivalent cap with a bunch of FILM caps.. there tend to still be good even after decades, they don't degrade, so old film caps might still be plenty good enough to get the laser going. I don't know.. trying to do my best to make progress !
Maybe I could plug the white one into the Toshiba player, hence with a cap in a different state of bad, might spot a difference in behaviour this way.
Cleaning swing arm : thanks fr the big warning and glad I asked !!! OK, won't take ANYTHING apart.. .just doing my best to clean and lubricate it IN SITU !
Clamp : OK my bad, will take that stuff apart to try and see that bloody metal ball....
REALLY going to bed now, see ya !
-
Just to get a head start on you for tomorrow!!!
Wanted to add a quick little something / erratum before I call it a day : I realized I fucked up my springiness test ! The black one, plugged into the Philips player.. was actually powered UP !!
If I remove power, it behaves just like the white one, but still feels much stronger/healthier...
So the springiness is not magic after all , just the flat flex ?! Feeling stupid.........................
Ha, yes, that would explain it!QuoteLaser blue cap.. hmm yes good point, might be good enough to get a healthy laser going, but maybe not good enough to get a weaker laser going... will definitely look into that indeed, I promise ! I don't have a new cap to replace it with though.. only 30 year old salvage caps, which are likely just as bad or worse... no ESR meter at hand to actually check for the condition of the cap. Yes, I am looking into getting a tester in the short term, I am fed up not being able to know what I am doing with old caps, I am sick of it. Will try one of my old caps anyway. Won't be new but it will probably be a different state of bad, so might manifest as a slight difference in overall behaviour of the drive, which would therefore confirm the problem and motivate me to spend 2 hours riving to town center and back, to go get a new cap at my local electronics shop...
Or maybe if the value of the cap is not that high, I could cobble an equivalent cap with a bunch of FILM caps.. there tend to still be good even after decades, they don't degrade, so old film caps might still be plenty good enough to get the laser going. I don't know.. trying to do my best to make progress !
There isn't anything particularly special about that cap, it's not low ESR or anything. It's just that that particular range of axial Philips ones have a reputation for drying out and losing capacitance. A salvage one will probably be ok for testing. If your DMM has a capacitance range, a value check would be good enough.QuoteMaybe I could plug the white one into the Toshiba player, hence with a cap in a different state of bad, might spot a difference in behaviour this way.
I would be tempted not to swap the transports too much at this stage. Firstly there is the risk of ESD damage, and secondly, the possibility that there might be a significant difference in laser current setting. -
If you have the CD371 basically working (producing audio?) using the CDM4 from the Toshiba, it's probably a good time to turn your attention to the display problem.
Note: For future search purposes the Toshiba model is XR-9128 - It took me ages to track it down in your previous repair threads last night.
The main board microcontroller is clearly receiving +5V and working as you have CD operation and key scan, so the problem is limited to a front panel or interconnects. Time to remove the front panel PCB and check for +5V. I think there's a fusible resistor on the +5V rail (the CD371 has fusible resistors on the individual supplies of every IC!). If no supply is present, then it could be a dry joint or ribbon cable problem. If it is present at the display, then it's either a completely dead display or problem with the 3-wire serial interface (data, clock and data enable) again, dry joint or ribbon cable. It's easy to continuity check back to the microcontroller pins.
As I mentioned before, the display data interface is very simple, the display decoder is a 35 bit shift register. You basically set the first bit high, followed by 34 bits of display segment states. Once the 35th bit reaches the end of the shift register, it latches the following 34 bits and outputs to the relevant pins. There is no multiplexing, it's a straight one to one match with the common anode display segments. The MM5450 datasheet gives better detail.
This page gives full details of the correspondence between the data bit, and display segment (or LED), together with an example of replacing a broken display with separate MM5450, ordinary displays and some stripboard...
http://audio.etata.hu/2010/12/17/nsm4202/
I've also attached a copy of the pin mapping table for reference (just in case the page ever goes away). Note that I have edited this - the original shows Bit 34 as not used. From looking at the CD472 front panel circuit, I think Bit 34 is the IR remote indicator of IR equipped players (CD372 / CD472). Note that the CD371 front panel PCB also has an unpopulated footprint for an IR receiver.
-
OK it's late and I am super tired so I will have to keep it short...
Spent a few hours on the thing. Did as Gyro suggested , didn't even touch the CDM, spent all my time on the display/front panel board.
Took it out of the front panel, power up.. and go figure, by some miracle it DID display some stuff !!! So it's not dead...
To cut a long story short, I fiddled with the thing for hours, tried everything a brain can think of. Tried all the easy stuff, nothing worked.
No bad connection no bad solder joint, no broken traces.. only found a couple 22k resistors that went open circuit, but there role is not vital, they are in parallel with the two decoupling caps (yes blue Philips..) , bleeder resistors to make sure the display goes off instantly when the user powers the unit off, I presume ?
The blue Philips I pulled one leg off, to test them reliably. Indeed you wee right Gyro, these bloody caps do lose a lot of capacitance of time ! In this case almost 50% !!! Supposed to be 47uF, bot measured about 26uF... still, power suply looks good on the scope, no excessive ripple or voltage drops, nothing that I would worry about.
Then scope the serial interface, pics attached.
What I found interesting is that the CPU refreshed/overwrites the display constantly, whether or not there is an LED that needs its status changed. IT just spits out the same data frame over and over again, go figure.
Scopes the Chip Enable pin (active low it appears). It refreshes the display at about 30Hz. Active for about 4ms out of a period of 32+ ms or so.
The clock signal is gated by the MCU accordingly, so it spits out 4ms bursts of pulses. Clock runs at 14kHz. A pulse lasts about 70us, so... yes, I counted the pulses one by one, there are 56 of them ! 56... for a 35 bit register ?! How can that work ? What am I missing ? What's the point ? The MCU is well aware that the register is only 35 bits long, so why bother shift it 56 times ?! Eh ?
Then the data frame. Perfectly stable and consistent, nothing wrong there. if I press the repeat or pause button, which have their dedicated discrete LED, I can see a bit in the frame go on and off. If I switch to "time" mode, then I can see the last bits of the frame change every second, in a counter like pattern, as expected.
So the CPU is definitely sending proper data, no worries. It's the display module that's bad. More precisely, I would say the decoder (chip on board, big black blob behind the LED digits) is fine, since I managed, for 30 seconds to have it behave perfectly and I had alllll the LEDs and digits working perfectly. Didn't last long though. Most of the time it displays partial garbage and partial sense. often it won't display anything at all. It's completely random.
Point is. I know the module is getting good data, all the time... 30 times a second. I also know the LED displays are good. I also know the decoder itself and shift register are good, since it worked once, and works partially most of the time. So... can only be a bad bond between the LED digits and the chip output pins. When the display is all dark and "frozen", unresponsive... it comes back to life when I press hard with my finger on the LED displays.
So, bad luck... I can fix the bad joints of course, no access. So will need to do like the guy you linked the website of, and replace the chip. Looking at his pictures (he doesn't mention this particular point in the text), it looks like he managed to reused the existing LED display.
For the rest of his project, it looks so messy and horrible, there is no way, no way in hell that I am inflicting this to my lovely player, or to my eyes really, as it hurts so much. At least it works, I get that... but still.
So I think it should be possible to make replacement module using a surface mount chip...stuck behing the display, just like the original chip on board.. except replacing chip on board by a surface mount chip. Sadly the MM5450 according to the datasheet I pulled, is not available in surface mount packages. There is a PLCC package which could be surface mounted, but I guess it would be too thick to fit, couldn't clear the front panel board I fear.
So will have to search see if I can fiund a comparable chip in a surface mount package.
Goal is to make a replica of the existing module, ie NO ugly wiring... zero wiring. Just the LED displays soldered to the custom PCB, and the decoder chip soldered directly to the PCB, with traces linking the two. zero ugly wire. If I can't find a suitable chip, I don't see why we couldn't just daisy chain a few bog standard 8 bits shift registers ?! The advantage of the M5450 though, being it allow to control the intensity/current of the display with a single pin...
Anyway, this sounds like good fun, but it's time consuming and I have not made a PCB in 15 years... times travels faster than light eh....
So it will have to wait... house work is the priority.
So now that the display problem has been diagnosed.. tomorrow I will get back to the CDM.. see if I can get it working...
Midnight, exhausted, good night and see you tomorrow for the next and probably last (for now) episode....
-
https://www.diyaudio.com/forums/digital-source/259790-nsm4202a-led-display-module-replacement-philips-sony-cdm-transports.html
Built modules can be found for 35€, maybe less, on Aliexpress.
Regarding the 56/35 bit enigma, possible that the CPU is ready for 6 digit displays (time + track N°) just the excess data is dismissed on the 4 digit models. Besides isn't the MM5450 a workaround with hardware similar but not identical to the original design?
Maybe you could get away with a couple of these:
https://www.st.com/resource/en/datasheet/stp16dpps05.pdf
Or
https://www.njr.com/electronic_device/PDF/NJU3716A_E.pdf
Or four 74 family 8 bit SIPO shift registers...
I'd probably go for a PIC, some code and a multiplexed LED drive to simplify the PCB layout...
-
Sorry to hear that it is the display module, I feared that it might be but (because it was totally blank) I hoped not. Responding to physical pressure is a sure sign of a bond wire failure on the blob-top IC, which i think was the general issue with them.
Yes, looking again at the page I linked, the guy did re-use the existing display LEDs by tacking onto the rear of the display module. Personally I would have used finer wires, which might have made it neater! Yes the smallest package for the MM5450 is PLCC.
I found another site where the author has really gone to town on designing pin compatible replacement display modules...
https://www.bramjacobse.nl/wordpress/?p=171
I notice that some of the ebay replacements (I didn't look at Aliexpress) are populated PCBs which re-use the plastic moulding from your original display - still cost more than you paid for the player though. I think these designs are lifted from the above site.
@shakalnokturn: I suspect that NS did use the same die for both the display and the IC, they originated both, and the display does bring out the relevant external LED drives on pins. No way to be sure of course. I can't find the display datasheet anywhere on the web (I will try my old NS databook collection) but one of the block diagrams on the above mentioned site looks identical.
By the way, the CD471/CD472 use the MM5848 which has the same functionality, but with high voltage outputs to drive their VFD displays, making them interchangeable on the same main board (only the transformer is different, having extra windings).
Very interesting findings on the serial data stream. I had never thought to actually scope it, but was expecting it to only update the track data on track change (obviously it would need to be more often on time display). The extra data bits are curious. As shakalnokturn suggests, it may be for extra digits - although the player defaults to track number display and this includes Index so is already 4 digits long. It does explain why the display enable signal (which just internally gates to clock inside the IC), otherwise the 35th bit latch scheme would only need data and clock to operate (the CD47x front panels use a diode-OR scheme to gate the clock as there is no data enable pin).
One of the reasons that I like these players is that they offer opportunities to be used in a home project. The CD tray and loader mechanism are separate from the case moulding - if you look at later models (the CD610 for instance), things like the loading motor mount are part of a molded 'web' which is part of the case moulding (the PCB smd parts are accessed from below, but it is a pain to get at top side parts).
The separate mechanics makes it very easy to integrate the whole assembly into your own constructed case, with brushed Aluminium, Wood etc. (take your pick!) front panel. There is plenty of front overhang on the tray so all that is needed is a slot in the front panel. The front of the tray is even removable (Rotel substituted a metal one). The tray mechanism is also the simplest design of all the players, no cams or transport movement, just one belt and gear wheel. You could then implement just the front panel controls that you wish - for instance, a minimalist 2 digit display to just display track number, minimal controls for eject, play, pause, stop, and maybe track selection (although you could just make these functions available on the IR remote).
Such a case would make it easier for quality modifications, move the DAC to a separate PCB with separate analogue supplies and bigger DEM circuit decoupling capacitors. You will find that the higher-end, much more expensive versions of the player (from Philips, Marantz, Arcam, and others) made such changes, but the actual circuit changes are minimal, mostly additional supply regulators and minor component quality changes. S/PDIF is also available as a output on the SAA7220 (parts just not populated on the PCB), making it suitable as a Transport.
The reason that so many manufacturers' models were based on the CD373 layout is that it uses I2C interface. This made it much easier for them to customise lots of differentiating features and variations - very handy if you have your own front panel tooling facilities, custom multiplexed VFD sourcing capabilities etc. but much harder from the DIY perspective. This PCB also has a very strange placement of the TDA1541A DAC - right at the front of the PCB, well away from the analogue opamp I/V converter and filters. It's not ideal to have long, noise susceptible analogue tracks when the DAC has a settling time of 1uS and feedback is involved! The CD371 layout is much more logical (and accessible).
This is all stuff that I have every intention of doing - I just haven't got around to it yet!
Sorry, that was a bit of a wall of text. -
Just to round off, I though I would document the changes that I have made to my CD472 as they currently stand. These have had a major impact on sound quality (as I said, the differentiation between higher end and lower end models (as long as they used the same transport and DAC was very small in component changes).
Firstly, circuit and ground noise is a big factor. The SAA7220 is a power hungry and noisy device for instance. The board layout does benefit from the entire top side being ground plane but it has quite a lot of slots due to components.
1. Clock. I replaced the simple crystal on the SAA7220 with a canned oscillator. Clock stability is key, and the on-chip oscillator isn't good due to it being a noisy IC. Clock improvements were made by Arcam and others.
2.Replacement of IC electrolytic decoupling capacitors on the main power rails, DAC and opamp local supplies. The ones on the CD371 are the minimum that Philips could easily use. In particular, I used a large OSCON on the SAA7220 sypply for minimum ESR.
3. Replacement of the DAC DEM smd ceramic filter capacitors with PPS smd ones. Filter capacitor values have an effect on sound quality. Phillips and Marantz went from 100nF ceramic to 220nF film capacitors on their high-end models. Arcam used 220nF with 470nF for the two most significant bits.
4. Replacement of the opamps. Basic- mid range players use LM833, higher end use NE5532. Both are pretty old designs now. I inserted socketed OPA2134s, not optimal, but they were what I had at the time. There are some ADI opamps which are faster and probably a better speed match for the I/V conversion stage. I Removed the headphone opamp - it was fixed level anyway.
5. Removal of the output muting transistors and replacement with a muting relay. While the muting transistors shouldn't affect anything on paper, they do have an effect. Higher end players used muting relays (N/C contacts to ground the outputs).
6. Bypassing the output coupling bipolar electrolytic capacitors. Possibly age is a factor, but bypassing does improve things. BTW, it is helpful to use 'Vero' style push fit PCB pins (intended for stripboard) inserted from the underside of PCB for access to signals. This minimises the risk of track damage and minimises the number of times that the board needs to be removed.
EDIT:
7. I forgot, replaced the In4002 rectifier diodes with fast, soft recovery types - another noise source.
8. Arcam inserted 390R series resistors in the I2C signal between the SAA7220 and TDA1541A, this reduces noise injection into the TDA1541A die (I have seen other implementations where attenuators are used to reduce the signal swing to TTL levels). I haven't tried this yet.
I've attached some photos. Sorry, I had the flash on. The mods are rather messy, but functional.
-
Walls of text don't scare Vince!
-
He is rather good at them!
-
Just to show how overboard I have gone previously with the TDA1541A - this one is standalone, with slave clocking of the CD player (as transport) and data re-syncronisation. The result is very detailed, but lacks the bass quality of the CD472.
Please bear this aberration in mind when listening my advice!
-
SRPP output stage? I had that as a project for some (or no) day.
-
No, they're standard cathode (LED) biased triode gain stages using former Soviet 6S45Ps. Plenty of gain for passive resistively loaded TDA1541A outputs (20R) and nice low output impedance.
Taking the photo jogged me into waking it up again. On listening again, it is pretty magical, (even the bass).
It makes me realise that I still have some work to do on the analogue supplies on the CD472! -
Wowwwwwww..... little did I know what I would find on here once back home !
So much info in there... no way I can reply to everything so... just a big thanks for all the links info and pictures, a gold mine.
Looks like my 10 Euros CD player is now the star of the show, and will be turning into a long term mod project.... that will end up costing me 300 Euros at the end. Definitely can't afford that at the moment, nor the time to do all that. But, this thread will now be my bible and I can read your messages later on when I have time and money to tinker with this player.
You both went very wild, deep into mods ! I like to make fun of audiophools, and I would hate to become one... but you are both electronic guys, rational and tehcnical people, so I trust you to draw the line between what is actual facts, science.. .and just plain paranoia or "religious" audio faith....
So might mod my player here and there at some point, just to see if I can actually hear a difference. You do, so maybe my brain will too...
However, my brain is not trained to appreciate quality sounds. Am 43 and all I have had for now are a pair of shitty Altec Lansing small amplied speakers connected to the on-board sound chip of my 15 year old desktop computer motherboard.....
Only 4 months ago did my dad give me these shitty Toshiba speakers to go with the Philips CDC486 I fixed for him, now mine. And the amp that goes with it, cheap low end Sony amp, that I bought defective for 10 Euros. So I guess it must output crap, but I wouldn't even know, not having heard any other amp / speakers...
I guess brain needs to make side by side comparisons to train itself and spot the good from the bad...
Might happen sooner than I expected : last week, an old grand-pa on his electric bicycle, as he went past my house, noticed my lab and collection of old vacuum Tek scopes through the window. Parked his bikes, and sticked his head onto the window, trying to get a better view. I didn't know if I should call the police, or grab a hammer to get him out of the way.... I opted for the friendly approach. Glad I did. Guy is an old electronic guy, studied electronics in the early '80s. He was a bit into audio stuff himself. He scrapped a shit load of old electronics stuff just the other day, wished we had met a few weeks earlier so he could donate all that to me ! What a waste, a tragedy. But, he still have a little bit of stuff left. He showed me pics, will bring the stuff with his car this week-end. He will donate to me 3 pairs of loudspeakers. Some crappy/ obscure ones branded "Boston" he said, but also decent 3 way stuff a pair of Technics, and a pair of Marantz, in nice wooden cabinets. It's bound to be much better than my crappy small 2 way low end Toshiba speakers so... maybe I will a difference with his speakers. He did say however that speakers don't age well so they might degraded and not perform well ?! Well, it's all free so who cares. Will plug them in and see what happens !
Also some modern junk, jsut so I can salvage the electronic board from them and ditch the rest : a big 5.1 "home theater" 2x100 Watts IIRC, and also a couple DVD players, and also a small old Sony CRT monitor, like old CRT TV/monitors, always a bunch of cool power resistors and high voltages caps to salvage. Love CRTs....
Guy came back to my house a minute ago, and gave a little something already, see pics : an antic " QUAD " pre-amp, model 33.
I gather QUAD was high-end audio, like NAD ROTEL etc ?
Look at that thing, with DIN input connectors at the back not RCA's ?! Took a pic of the inside just because I love you so much
Huuuuuu... just checked the interweb... this antic QUAD 33 preamp is actually known to the world ? Found one for sale on Ebay for.... 1200+ Euros ?!
https://www.ebay.fr/itm/Vintage-Quad-33-Controle-Pre-ampli-Tres-Bon-Etat-Entierement-Revise/373522379328
.. because it's in good nick and serviced. "as is" units still sell apparently for 275/350 Euros, still a fortune !
I will happily "service" it and give a good clean, for the extra 1,000 Euros !
But mine is missing a button at the front, so the audio people will probably deem that my unit is therefore worth nothing anymore, or maybe 10 Euros for parts !
For 1250 Euros though, worth the trouble of 3D printing one and going to te trouble of trying to give it some patina to make it look legit !
Anyway, back to the little Philips.
As for the display, l now understand it was a third-party off the shelf unit used in a gazillion old players, and that they all go bad 30 35 years later. Hence why apparently so many people are trying hard to find a work around, and the "market" is so large that some sell ready made units on the interweb.
So, in retrospect... it was given that my module would be bad !
I still think I would go through the trouble of designing my own module though, though that will be sped up by all the work that these people have already done. Mainly, it looks like the first guy, who managed to salvage and reuse the original LED display, had horrific DIY electronic skills. On the othe hand, the guys who designed their own PCB, and surface mounted the 5450 PLCC like I was thinking I could do.. did not reuse the existing display. There replacements don't please me. Segments don't have the same shape, and color matching with the surround discrete LED's could be a nightmare.
So, what I retain from all the info/links you gave me on that subject is :
1) Yes it's possible to make a replacement module, so no need for a extra board with tons of ugly wires. It can be made to look as clean tidy and compact as the original.
2) it IS possible to reuse the original display, since that guy managed to do it...
3) Surface mounting a PLCC 5450 can indeed do it, it fits, just.
So if I ever get the motivation to make my own module, then I would target a surface mounted PLCC 5450 + reusing the original display.
If I can't, well might as well buy a ready made module, but it sucks.. that would be a defeat...
But at least it would get the player working again, would be a clean technical solution, quick solution, and not cost a mega fortune, and not cost me any R&D time. So whatever I end up doing, at least I can sleep on both ears : this particular problem has already been tackled and solutioned, great....
So now I am left with the CDM not working. Let's see if some cleaning and a replacement laser capacitor will improve things.... stay tuned !
No Gyro, as you have gathered by now, I am not at all afraid of text walls. I am rather afraid of people who are incapable of erecting one !
How the hell could you and Shakal have conveyed all the useful information you gave me.... in a one liner tweeter style ? Information needs SPACE !!!
Bandwidth these days is CHEAP ! There is no good reason to restrain your thoughts...99,9% of the internet bandwidth is porn, spam and useless crap on youtube and social media... let's use the remaining 0.1% to write useful stuff !!!
-
I though it might be a nice little surprise for you.
Your CD371 has the makings of a nice little project. It has all the same relevant parts as some of the really-expensive-on-ebay high end players (ok a few of them use the selected S1 grade part, but not all, and there isn't much difference.
I still think you could make up a replacement display using an MM5450, and a bit of stripboard - as an interim step to your final metal cased monster! Enough to be useable anyway, and you only really need to connect up two digits (only 15 wires)... or none at all if you just want music out of it!
Yes, there is a fine line to be drawn on this forum between good solid engineering and Audiophool accusations. I stick with the former, but I am sometimes surprised by what my ears tell me and the sound suddenly becomes 'right' or 'special', it makes the effort worthwhile. A good differentiator in approach is not buying ridiculously expensive components!
As I mentioned before, it is surprising how much grading in sound quality and product positioning that the likes of Philips, Marantz etc. could achieve with really very minor component changes, it's really a happy accident that they put such a nice transport into the relatively cheap 371/471 series. The TDA1541A is so good that they must have actually had to 'engineer down' the sound quality on cheaper units! It all changed later with the lower quality and cost TDA1543 and then sigma delta (Bitstream in Philips language) DACs that they designed for lower production cost and better yield - sound quality was a secondary consideration. If you compare the THD+N and dynamic range specs with the earlier DAC, it is clear. Cost was the main driver. Likewise with Burr-Brown and their transition from Multi-bit to sigma delta.
Look at some of the high cost Philips based players on ebay and download their service manuals - you will be surprised.
It sounds as if you have come into contact with a nice... and generous, person. Yes, Quad is a long established (probably around 70-80 years old by now) quality British HiFi manufacturer, one of the best and far higher than Rotel and NAD (NAD were the 'good value for money, affordable' new kids back in the '80s) - if your new friend mentions electrostatic speakers, or Quad II valve amps, be very humble! Yes the Quad 33 preamp can go for serious money, particularly if accompanied by the matching 303 power amp. Enough to finance personal project costs, or to enjoy. Yes 'service' in this context means replace the electrolytics and clean the contacts, leaving as much original as possible.
I'm glad the information from both of us has been helpful [EDIT: and you contributed some useful data too!]. It's been quite therapeutic to actually write it down and re-discover some of it - I purchased an MM5450 today in preparation for the master plan implementation!