CDM-4 troubleshooting help

Hi,

I've got a CDM-4/11 that I'm trying to troubleshoot and repair. When it came to me it wasn't working at all but I've managed to adjust the laser current to 50mV and the focus offset to 400mV and now it will proceed to service mode 3. It will eventually read the TOC in normal operation but this takes a considerable time and sometimes results in an error. However when I try and scope the eye pattern on the HF pin of the TDA5708 the swing arm slams against the other end of its travel and the CD stops playing. Does anyone have any suggestions why it's doing this ? I must have it adjusted more or less correctly for it to reach service mode 3 but it clearly isn't spot on because of the time it takes to read the TOC and the scope probe is clearly upsetting the HF amplifier for some reason. Without being able to check the eye pattern I feel I've hit a bit of a wall with it. Any and all help greatfully received :)

Jon
 
In an update to my earlier message I've now measured the eye pattern and it looks like the example in the service manual BUT has an amplitude of 2.3V pk-pk rather than the 1.2-1.5V that it should be. Does anyone have any suggestions why it's doing this and how I correct the amplitude ?
 
Reduce the laser current until the eye pattern has the correct peak-to-peak amplitude. This may require only a small rotation of the trimpot. The P-P amplitude of the eye pattern is proportional to the amount of IR light generated by the laser diode.

Also:

Is your oscilloscope reasonably accurate measuring the amplitude of RF signals of similar frequency (1 to 4 MHz)?

Are you using a 10:1 scope probe which is properly calibrated to your oscilloscope?
(The 10:1 probe prevents excessive loading of the eye signal. A direct 1:1 input to the scope may have too much load capacitance.)

Finally the PC board that drives the optical pickup has a 33uF electrolytic capacitor which often fails from old age. I recommend checking the ESR of all electrolytics in items this old.

-EB
 
Thanks EB,

The probe amplitude calibration is around 9% out with respect to the 'scope test point but I've taken this into account. I'm using a 100MHz 'scope bandwidth limited to 20MHz.

I've tweaked the pot up to give an eye pattern of 1.5V pk-pk but the laser current is now only 14mV and no longer as sharp. It won't go into service mode 1 and in normal operation it won't read the TOC.

I'd wondered if setting the laser current to 50mV should equate an output of 1.5V pk-pk so is there an external component to the focus servo that controls the gain ? Maybe something around the amplitude detector ?

I've replaced the four electrolytics (2x100uF and 2x220uF). I've also replaced the 1.5uF bi-polar in the radial motor snubber network with a 1.5uF film capacitor. The blue axial philips on the laser supply seems to raise some discussion. Some people insist that it will only work if replaced like for like but I've used a 47uF/16V Oscon here on a Micromega Solo years ago with no problems and I've fitted a 47uF Panasonic FM to this mechanism as suggested elsewhere.

I've re-adjusted the laser current to 32mV which gives an eye pattern of 2.12V pk-pk. It still takes several seconds to read the TOC and will now skip through to track 3 or 4 before throwing an error.

I still don't think I've got the focus servo running correctly as it should read the TOC in around a second, I believe, but does the stopping at track 3 or 4 suggest I also have an issue with radial servo too or is this still likely to be a focus servo issue ? I guess I need to start by looking at the RE1 / RE2 signals but it's a real pain to probe around the servo board with the mechanism up on its end. I need four pairs of hands.

Jon
 
I also have a second mechanism which I've got as far as service mode 2. This also has a higher than 150mV pk-pk amplitude HFout signal with 50mV laser current. This one has some hash on the HFout pin at around 10MHz (even when the laser is off) which is superimposed on the eye pattern when the CD is spinning and creates a messy eye pattern. This mechanism won't read the TOC. I don't see this on the supply rails so I assume it is being generated in or around the TDA5508.
 
I've gone back and read the TDA5708 datasheet again.

I assume the 150nF on GChf (pin 4) provides a poll and thus roll-off of the HF amplifier gain ?

Other things I'll check / replace are the 10nF on the DEC bias current decoupling (pin 1) and the resistor values on Bgc (pin 23) and Beq (pin 22), also the 2k2 / 100pF between HFout and DET.

I assume the best replacement for the HFin coupling cap is a 1% C0G/NPO as I'm going to struggle to get a surface mount polystyrene, looks like it's a 1206 ?

It will probably take a few days for components to arrive though.
 
I should do a bit of research about the design of the CDM-4 optical pickups. I can provide a general overview however:

All optical pickups contain a dedicated photodiode light detector which directly samples IR light emitted by the laser diode and thereby controls the laser current. In other words there is an internal feedback and control system for IR light intensity. This is used because laser diodes are quite nonlinear in terms of light emitted vs. laser diode current.

Adjusting the “laser” trimpot controls the amplitude of IR light emitted by the laser via this internal feedback control system.

Some optical pickup servo systems may also sample the amount of IR light being received by the ABCD signal receiver photodiodes which recover the RF eye pattern signal while playing a disc. But keep in mind there won’t be any IR light received by these detectors until the IR beam has successfully been focused on the data layer of the CD disc.

When an optical pickup is incapable of “initializing” the disc there might be a variety of different faults at work. We cannot assume that failure of the laser diode itself is ***always*** the root cause of the optical pickup failing to function.

My own personal method for diagnosing optical pickups is to substitute “a known-to-be-good” optical pickup and then observe the differences. But I must confess to being a packrat and gear hoarder, which means that I have an ample supply of donor units and spare “used but still good” component parts on hand for experimentation.

I wish you good luck with this project and I look forward to your success with it.

-EB
 
The general flowchart for initializing a CD is as follows:

1) The sled is put into the reset position with the optical pickup located at the inside track of the disc. There is frequently a sensor switch to signify that the sled is in the reset position.

2) The laser diode is switched on.

3) At this time a “focus search” is performed. A ramp voltage is applied to the focus coil. This causes the lens to move up and down over its full range.

4) The combined signal from the ABCD photodetector diodes is monitored. There should be a brief pulse of signal as the lens passes through the focus point. This, of course, will only occur if a disc is in the tray. It should be possible to observe this “pulse” with an oscilloscope connected to the “RF eye pattern” test point.

5) If the peak signal level doesn’t exceed a certain threshold, then the ramp voltage applied to the focus coil will repeat several times. Many servo systems will also reset the sled position a few times. Some (but not all) CD players may also apply power to the spindle motor at this time. But eventually the player will indicate “no disc” and stop.

6) When the peak signal exceeds the detection threshold, then the focus servo will be activated. This should result in a steady level of signal recovered by the ABCD detector diodes.

7) Next the spindle motor will be switched on and the servo system will attempt to detect the RF eye pattern signal.

8) Then the tracking servo will be switched on and the servo system will attempt to read the table of contents (TOC) of the disc. The sled may move several times during this operation.

9) When all is well the number of tracks on the disc will be displayed and the “play” button will become functional.


I find it useful to compare “how long it takes to initialize a disc” among several similar players. When the optical pickup is good and the servo system is properly adjusted, then disc initialization should be rapid and reliable.

CD players of similar design and architecture usually follow the same sequence of events during “disc initialization. I expect that most CD players with the CDM-4 optical pickup will have similar behavior during disc initialization.

When the behavior differs this may help identify the failure cause. For example I’ve encountered faulty optical pickups which fail at step 8 (reading the TOC). At this point the disc is rotating and the RF eye pattern is visible with the oscilloscope. Yet the player still fails to complete the initialization process. I suspect this might be caused by excessive “noise” or “jitter” which interferes with recovery of data from the RF eye pattern signal. In some cases a partially faulty spindle motor has turned out to be the cause of this type of fault. Evidently it was unable to steadily rotate the disc at precisely the correct RPM. In this case spindle motor replacement corrected all problems.


-EB