Philips CD720 (CDM12.1) beginning to skip

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Our secondary player which used to play all CD-Rs, even those that the car player refuses due to low reflectivity, is now beginning to skip occasionally even on high-reflectance CD-Rs. I don't think I have observed this on original CDs yet, but I may be wrong there.

The eye pattern after preamplification looks fine (it's the variety with a discrete preamp, half of which sits on the flexible pcb next to the pickup).

What should I do next?
- crank up the laser (which direction?)
- put automobile engine oil on the steel rod that the sledge glides on?

As a last resort, I still have a Grundig drive & electronics that I purchased as industry surplus. I noticed there are different CDM12.1 drives, some of which have no transistors on the pickup unit and a TDA1302 preamp either on the pickup or on the main pcb instead. The Grundig still has the discrete amp. Can I count on it to be compatible?

Eric
 
capslock said:


What should I do next?
- crank up the laser (which direction?)
- put automobile engine oil on the steel rod that the sledge glides on?

Eric

Did you clean the lens? (Distilled water & cotton stick)
Don't cranck up the laser. Won't help you in the long run.
Don't apply "Engine Oil". Clean the rod(s) with alcohol or similar and apply some very thin oil. Just a tiny drop.

/Hugo
 
capslock said:
Lens looks fine (no smoker ever came near), so I decided not to fool with that. Besides, eye pattern looks fine to me.

Mrfeedback even recomments automatic transmission oil...

Regards,

Eric

You should clean it; don't be afraid, if you are gentle with it nothing will happen.
Check this too: http://www.diyaudio.com/forums/showthread.php?s=&threadid=11797

I think Mrfeedback must have been joking? ;)

/Hugo
 
Hi

Since the eye pattern is fine, cleaning the lens would not realy be required

You may verify the radial tracking mechanism, its' grease can become slightly "rigid" after a while

I do not have a clue which grease is allowed, but be careful because some grease interacts at a nasty way wich plastics......

Also, this great link may help

http://www.repairfaq.org/REPAIR/F_cdfaq.html

succes
 
mr feedback doesn't recommend engine oil, but automatic transmission fluid. this is fine as it has additives for bronze based bushes. just don't use too much, only rub it on. in fact, use very little. incidently, you can also use motorcycle fork oil as this is the same, but comes in much thinner grades. personally i use watchmakers oil. super thin and lasts nearly forever.
 
jean-paul said:
If you want info from Mr Feedback it is better to send a private mail. You don't seem to accept the answers you got so opening a thread is a bit useless.


Well, the answers here were really basic (this isn't exactly the first player I repair) and not really in response to my questions. MrFeedback has given good advice on repairing my CDM-9, and with the exchange having gone on in a forum here, it is all there for everybody's reference.
 
Disabled Account
Joined 2002
Well, the answers here were really basic (this isn't exactly the first player I repair) and not really in response to my questions. MrFeedback has given good advice on repairing my CDM-9, and with the exchange having gone one in a forum here, it is all there for everybody's reference.

OK, the suggestion to crank up the laser ( and asking which direction ) together with the engine oil gave the impression you were repairing a player for the first time.

The basic answers of cleaning the lens and lubricating with the right grease are the solution in at least 50 % of failures. You could have excluded basic answers by telling you already looked into that. Besides that you asked yourself if lubrication was a solution.
 
little update

I have cleaned the lens (with ethanol which I know from past experience does not hurt CDM12-lenses) and greased the mechanism -> no change.

I have replaced the original CDM12.1/15 with an unused CDM12.1/05 from surplus Grundig repair kit. The pickup-sledge unit looks exactly the same, even to the discrete amp sitting on the flex pcb. The only difference I can detect is the length of the feet of the tray assembly (mounting on top of pcb vs. on bottom of case. The 05 worked from start, but skips in exactly the same positions of the same CD-Rs as with the 15.

I have cranked up the laser power on the 15 by turning the pot clockwise so that the wiper moved by about 0.5 mm to the right (WARNING: overdoing this may cause the laser or your eyes to fail prematurely). Result was improved playability of one CD-R (one skip instead of half a dozen) but no change on another CD-R. Moving the wiper by another half mm did not improve things.
Connecting a 100 nF stacked film on top of the app. 220 nF SMD device on the flex board (be careful to solder quickly and with a temperature of only 280°C) may have improved things a little.

Grounding and decoupling on the main pcb is not as bad as in the more expensive CD-931. On first sight, it was clear that the servo driver ICs had little really local decoupling. I added 100 nF each over the supplies of the focus/tracking and spindle/slegde driver ICs, did the same thing for the TDA1301 and the analog supply of the SAA7345 => no change.

It may be possible to introduce more improvements, but this would require careful analysis of the whole ground network and is not likely to be the root cause here. I may decide to change a few electrolytics that look under-dimensioned to me, but apart from this, I can't see anything that would have aged on the main pcb.

I am really beginning to wonder whether these CD-Rs are the problem, knowing that a brand new CDM12.1 has the same problems.

The player used to play CD-Rs from the same brands that are now skipping (or actually, unlike some other players, all brands I tried). I am pretty sure it used to play even some of the very CD-Rs that cause dropouts or sometimes skipping now. All of these play beautifully on a cheapo Kenwood player, as well as on a very old Technics player that uses a Philips CDM-4 mechanism.
 
skipping

i have had a few cd player repairs that where skipping. and turned out to be the motor for the platter or the tracking. the brushes ware out. you could play with the lasers wavelength but it just makes em run hotter and thay dont like heat[]:scratch:
 
Hi Karma,

the pickup unit comes with spindle and sledge motors, so it can't be the spindle motor. Besides, I now after all attempts, now get more dropouts than skips.

The pot changes the laser output power, not the wavelength. When the diode gets hotter, it may emit a slightly different wavelength, but this is really second order (couldn't resist commenting, being a laser physicist).

Eric
 
Disabled Account
Joined 2002
Grounding and decoupling on the main pcb is not as bad as in the more expensive CD-931

Please explain this if you like. What is exactly wrong with grounding in this one ? I know that there are a lot of decoupling caps omitted while there is space for them and even soldering pads.

It also has a nice ground plane as far as I can remember.
 
jean-paul said:


Please explain this if you like. What is exactly wrong with grounding in this one ? I know that there are a lot of decoupling caps omitted while there is space for them and even soldering pads.

It also has a nice ground plane as far as I can remember.


Most of the details should be in the old thread about the 931. It does have a double-sided pcb, but no contact holes. Contact to the bottom side is made through wire bridges which are randomly scattered over the board and as much as 10 cm spaced apart. When you trace where the signal really goes, there are things like ground loops and shared ground impedance. My favorite one was something like this: the photodiode signal itself is referenced to a reasonably clean ground, but the chip that processes this low-level analog signal (it was 8809 or something like that) is on a ground line on which also the main user processor is. This means all the ground spikes that the processor induces are seen by the analog processor because it uses its ground as a reference, and this ground is quite different from the reference ground of the photodiode. Another interesting thing was that the TCA0372 that controls the tray loader had excellent decoupling but decoupling on the TCA that controls focus and tracking was close to absent.

Cutting up ground traces, hand wiring grounds so that analog HF, digital control and DAC had local star grounds each which all went back to the same super-star-ground improved signal quality a lot. In the end, though, it turned out to be the greasing of the single ball that the swing arm pivots on...

Regards,

Eric
 
Netlist said:
1) There should be a green wire from the spindle motor, going to the main PCB.
Check for loose contacts on that wire.

2) Check if there is a resistor 5k6 at pin 14 of IC 7100. (SAA7345/M5)

/Hugo ;)


Two pickup units had the same problem, so if there is a problem, it must be on the pcb side of the connector. I will check.

Pin 14 is crystal resonator out. I looked at that circuit before, and I don't think there is any extra component there (where should it go when no other part uses the signal?). I will check again. What is the reasoning here? Does the output need a load other than the resonator for the oscillator to work properly? According to the data sheet, there should even be a 2k2 to analog ground.

Pin 29 outputs 4.2336 MHz - which might be used to reclock the TDA1301. Maybe I will try that.
 
capslock said:



Two pickup units had the same problem, so if there is a problem, it must be on the pcb side of the connector. I will check.

Of course when the entire wire has a fixed connection to the unit, and you changed that wire, it is almost impossible that this could be the problem.

There should be a 5k6 to ground. In early productions it wasn't there. It is introduced in production week 9510. You can find the production date in the serial number. e.g.: (x)xx129510 xxxxxx

Here's another hint:
Problem: Sometimes focus gets lost.
Cure: Add two 3,9Mohm (Mega) resistors between VRH (pin4) and pin 5 and 6 of DSIC2.
(Digital Servo Processor) This mod has also been introduced in production time but I can't tell you when.
Don't ask me more about those cures, I never did them but they come from Philips notes.

/Hugo – hopes this helps. ;)
 
Hey Hugo,

you're a treasure trove!

Well, I checked. The connection belonging to the green wire on the pcb looks and measures well, all the way to the output of the TCA op amp.

The load resistor on clock out is there, and it is 2k2. In parallel there is an axial ceramic cap, probably 100p (anybody know what direction the code starts?).

Clocking the TDA1301 from pin 29 won't work: it has an external 8.46 resonator, and the data sheet says the divider is set by software. I wonder why they went through the extra cost of using another resonator rather than a resistor...

The 3M9 resistors are not there on the 1301, there isn't even a space on the pcb. I will try this....
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.