CD Player repair

[knudsenuk]

Hello All
I am repairing a CD player. The spindle motor drive circuit has some faulty parts. I managed to download a manual and it identifies two resistors as follows:
R147 6E7
R146 6E7

Both are marked as critical parts to be replaced by the same value. There are also lines in parallel with the body of the resistor on each side. I am a bit confused with the notation. Can anyone give me a steer please. One resistor measures 7R6 the other nothing. Both have some heat damage and T103 on the schematic is blown.
Any help much appreciated.

[knudsenuk]

Actually the value is 4E7 apologies

[Gyro]

They stand for 4.7 ohms, (they are 1/4W in this case). Note that the symbol used (the line each side of the rectangle) indicates a fusible resistor on early Philips reference design based players - Philips, Marantz, Rotel, B&O, Quad etc. They are intended to prevent smoke risk in the case of a shorted transistor.


P.S. I'm guessing it's an RCD855 / 865. From what I've seen, they do appear to blow spindle motor driver transistors occasionally for no readily aparent reason.

[knudsenuk]

Thanks for your very informed response. It is indeed the 855.
Now working, although I didn't replace with fused resistors.

Is this 4E7 normal? I have always used 4R7.

Once again, many thanks. Really appreciate the help.

[Gyro]

You're welcome. By the way, if you download the service manual for the RCD-955 you will find the print quality a bit clearer (that was the giveaway!). The internals are identical, it was just a styling change, they decided not to mess with a good thing.

Philips stuffed in fusible resistors everywhere (which Rotel followed, even though they designed their own PCB, unlike the other manufacturers), you'll find them scattered over all the IC supplies even though the regulators all have current limit. It wouldn't happen today, but at least it saves the transformer from overheating and other collateral damage. If you do want to put in fusibles, you can get them from CPC, order number RE07242 - you'll have to buy a bunch of other stuff to get free postage though.

Regarding the 'E' instead of 'R', it was a Philips thing back then. I don't know if it was a more general Dutch / Belgian usage though.

[knudsenuk]

AHH, I see. Never seen that before. It does have a lovely sound.
I changed the opamps for lm4562. Replaced the PSU electrolytics.
I will track down the later manual. Many thanks

[Gyro]

You're right, I've been playing around with TDA1541A based players over the years, including my own DAC designs using a slave clocked transport and the 855 is the most musical and relaxing player I've ever found - I'm no Audiophool by any stretch, but it's a perfect example of taking a pretty much standard design and, with an optimal PCB layout, taking it to the next level, something that Philips themselves couldn't seem to manage. An example of how sensible supply layout and signal routing can make all the difference in a mixed signal design.

Again, desperately avoiding the audiophool label, the mods I've found most effective...

- A separate +5V regulator for the TDA1541A (just a basic 7805 with 220nF output cap feeding the lifted end of the fusible resistor). The main +5V feeds pretty much everything, including the decoder, micro, front panel, and rather noisy SAA7220 digital filter. It has quite a lot of hash on it, even with local decoupling.

- Swapping the fusible resistor on the -15V to the TDA1541A for a 33uH inductor. Inproved decoupling on a sensitive rail.

- Inserting series resistors in the I2S lines from the SAA7220 to the TDA1541A, 120R on BCK, 390R on WS and Data. The 7220 is a MOS ic with pretty much 5V output swing whereas the 1541a uses ECL logic and only needs a minimal voltage swing. The resistors help to reduce noise injection into the 1541a substrate (it is possible to do much better). As I mentioned, the SAA7220 is a very noisy digital filter IC, it pulls nearly 200mA @ 5V.

Basically, it's all about keeping noise out of the DAC, nothing mystic.

It's possible to go further - the 11.2896MHz master clock is a simple Xtal on-chip inverter based oscillator on the SAA7220. The level of noise on the chip causes a lot of jitter. This then gets piped back to the 7210 decoder without any particular care or terminations (which shouldn't actually matter as the 7220 re-clocks the I2S at it's inputs) and is also divided by 2 and used as the BCK to the DAC. External clocking seems desirable. The 855 is a good platform with a good solid chassis and plenty of free space just where you'd want it.

As with all consumer products, models are designed to fit their allotted space in the price range. Cheaper products are supposed to sound worse than more expensive ones. Rotel rather messed that up, the multibit 855 / 955 sounded better than the more expensive sigma-delta 865 / 965 (and achieved slightly better specs for THD+N and dynamic range too). An example of 'doing it right' can be seen in the Arcam Alpha 5 / 5 plus / Delta 70. If you look in the service manual you can see the differences between variants - minor supply decoupling improvements, 7220 on chip clock vs external, adding I2S data reclocking before the DAC etc.

Sorry, I'm blathering at this point. 


P.S. I picked up my 855 for less than £30 incl. last year, listed as intermittently skipping. No, not the laser, the top of the CD clamp had worn a divot in the pressure plate and was causing enough lateral vibration to affect tracking. An easy fix!

[knudsenuk]

I did notice the noise on the digital filter and thought it must be an anomaly.
Did many mods to the arcam. The best result was isolating every ground return as it's pretty much a bus on the PCB. It then became a raspberry pi streamer via I2S and then I did my own DAC too, again with a pi.
I just like the hands on with CD.
I have no belief in so called hi res so the tda DAC chip is perfect. Exactly as you say relaxed and musical.
The naim cd2 is very good but the lasers seem doomed on naim equipment.
Interesting chatting. I was a  hifi dealer but just got bored of all the bollocks . Like you I now trust my ears rather than my wallet.
Regards Richard

[Gyro]

Interesting chatting. I was a  hifi dealer but just got bored of all the bollocks . Like you I now trust my ears rather than my wallet.

To you too, I didn't realise I was talking to an ex-peddler. At least you saw the error of your ways.   A great opportunity to play with the toys though.

Yes, the SAA7220 is a noisy beast (I think a few people have sucessfully substiututed other filter ICs). Some people give it it's own regulator but it seems more sensible to provide a clean separate supply to the victim rather than the offender, it avoids dissipation issues and extra heatsinks too. Beefing up the filter supply decoupling and, again, substituting an inductor (as Arcam did on the 70.2) for the supply resistor seems a better bet.

The pi-1541 streamer sounds like a fun project, but like you, I prefer media that I can see on the shelf, anything else is just too 'cloudy' at my age!

Interesting about the Naim, they used the CDM9 irrc so should have been reliable, It shouldn't have been a major challenge not to glitch a laser, especially at those prices, maybe mounting it in a slam door wasn't the brightest move.

I plan to keep tweaking, at least this one is in a case, unlike my standalone DAC. I'll try to update you on any further significant improvements. As long as I don't open it up one day and let the 'magic' out.

Regards,
Chris

[Greybeard]

Regarding the 'E' instead of 'R', it was a Philips thing back then. I don't know if it was a more general Dutch / Belgian usage though.

I've seen that before in a schematic of a Bugera 5W guitar amp (G5):

[knudsenuk]

Hi Chris
Swapping the resistor R301 for an inductor?
I remember buying an 855 when I first got into audio. The rest of the system was not great but I decided to "upgrade" to the 865 and was so disappointed. I bought this one for the DAC chip but when I listened to it I thought "she's a keeper".
We used to buy and sell second hand Naim and Linn equipment and have had some very expensive and surprisingly average esoteric components. There is a very definable quality to the TDA DAC and it seems to be evident across many manufacturers.
I will buy another one I think (Rotel).
This is the rather basic DAC I built with slightly over the top power supplies. I need to finalise the filters and gain. Bass seemed a bit light compared to the Rotel. The pi based streamer runs Volumio which I find pretty good and can interface with Tidal too.

Rich

[Gyro]

Hi Rich,

No, R304, the -15V rail to the TDA. I actually used a Murata SMD inductor (patched on topside).

I fed the dedicated +5V regulator into the lifted end of R301. As I say, I just used a 7805, patched on to ground point near R904 footprint (part of the non-populated headphone amplifier circuit - not shown on the 855 schematic, but you can see it on the 955). I didn't want to start scratching up the solder resist and it's a convenient point. The +10V feed for the regulator can be picked up on the anode of D701 without removing the board.

The I2S resistors are fitted to the underside between the SAA7220 filter and the DAC. I haven't beefed up the decoupling on the filter yet, it's a pain to keep removing the board!
I've attached a photo for reference. Ignore the muting relay stuff, some of the higher end players used relays rather than muting transistors but on listening, I'm really not convinced.


Thanks for the DAC photo. Nice attention to power supply segregation, I'm certainly not going to fault you on that. It looks like you did everything right, that I can think of anyway. Interesting info on the pi streaming, I have a model B hanging around so may investigate that.

I've attached a photo of my DAC (don't laugh!). As you can see, a rather messy and evolving approach, not to mention a couple of hot cathodes - it turned out to be a nice way of getting plenty of gain in a single stage from low value passive I/V. Filterning is passive too. The digital stack on the left keeps growing vertically. Everything is reclocked to the master clock, and slave clocking of the transport - a CD210 (it was cheap and disposable) with S/PDIF back to the DAC. Not quite as many supplies (although the toroidal is from a Datron meter with a second screen between the winding feeding the two logic supplies and the one feeding the DAC, additional local shunt regulation of the DAC supplies too. It sounds very nice but ironically, a little loose on the bass compared to the 855.

Chris

[knudsenuk]

We have much in common. Amplifiers are more my thing. If you need any cases I had a bunch made and will never use them all. Proper CNC jobs.
This is the pi. The default clock multiples are not a fit with 44.1 so I used a kali reclocker, but you could remove the oscillator off the pi and recompile the kernel. But I hate this kind of thing!

[Gyro]

It does seem that way, doesn't it.

Yes, I'm into designing amplifiers (and speakers too, Tannoy DC based), starting with MOSFET (not happy) and graduating to hot cathodes (again) based on ears - athough the distortion measures very well too, mine uses former soviet NOS valves rather than the silly money ones. I evan have a few LM3886s sitting in a drawer, just in case the chipamp hype turns out to be true!

Again a healthy supply on the Pi I see. Yes I think I'd want to reclock too. The cases certainly sound interesting, my creations are more nuts and bolts and extrusions!

Oh, btw I modded my 855 display board to look a bit more minimalist, the default looks a bit cheesy. You lose the annunciators but I can now read the track number across the room.

(Amp photo linked from another thread...)




P.S. The display colour is thrown off by the camera, it's still the same whitish tint, not green!

[knudsenuk]

Hi Chris

Sorry for the long period of silence. We had a raft of things 2 weddings and a funeral. The amp looks great. Not seen those valves before. I did use Sovtek in my TL12+ about 20 years ago and they seemed comparable with anything esoteric.

Have you had any issues with the laser in the CD player? I was reading the manual about calibration and the pot they mention is (of course) under the transport.

The cases I have would be fine with solid state but not valves, you are welcome to a few if you want them.

Regards

Richard

[Gyro]

Hi Richard,

That's ok, you could almost use something like that as a film title (sorry)! We've had 1 wedding and 1 funeral in the same time, not immediate family though, thankfully.

Ah, the valves are still a bit of a secret, they're old Soviet NOS GU50 RF PA tubes. As standard they come with an Aluminium cap and hot-swap knob, designed to be yanked out and replaced on the battlefield. This easilly snips away (together with a whole bunch of shims to standardise the length), revealing a surprisingly modern looking valve with thick, hard glass, envelope and no less than 4 big getters. They were designed to run very hot (intermittent red plate) in the field, radio hams like them too for that reason. They last forever in audio use. The only drawbacks are 12.6V heater (but at usefully low current compared to KT66/88/6550) and low maximum screen grid voltage making them unsuitable for UL operation - they're happy in Triode mode past 400V though. The above, as well as different socket makes them cheap, I paid about £6 each but they've gone up a bit since then, especially with the Ukraine situation. Like most very cold warehouse stored NOS they need cooking in the oven for a while to de-gas before applying HT for the first time. They have an interesting history (Russian's capturing German WW2 radios etc. The Chinese still manufacture them as FU50s probably using the same old tooling... https://en.wikipedia.org/wiki/GU-50.

No, I've never had a laser problem, in fact one of the service manuals (not sure if it's Rotel or Philips) cautions that the adjustment is for new lasers only. Trying to bring up the output could kill an older one, although they're extremely reliable on the CDM4. As long as the player will read a CDR then there's no cause to worry. Even then, suspect check the lens is clean and check for the pressure plate divot that I mentioned in Reply #6 first. Luckily the Rotels don't suffer from the philips axial cap problem on the laser supply that plagues Philips and Marantz players, it has a decent Japanese radial lead one instead. As you say the pots are difficult to access - particularly the focus offset one which really needs to be adjusted (best not to) when the transport is horizontal!

Very kind of you, thanks. Maybe you could PM me a photo.

Regards,
Chris.


P.S. I almost forgot, I recently found that the Dual CD5150RC is a re-brand of the RCD-855. The scan of the service manual has very good PCB layout drawings, showing the PCB traces really clearly. You can find it on elektroanya.

[mzacharias]

If the spindle motor is a permanent magnet type (I'm betting it is) those burned resistors strongly suggest shorted spindle motor brushes.

[knudsenuk]

Hi Chris

I haven't forgotten about the boxes. CD player now has dedicated PSU for the DAC.
The boxes have a 10mm face plate in aluminium, a cover and a tray. Will pm you photos.
Was this the sort of thing you meant?

Regards

Richard