Sony X5000 troubleshooting

Hello PlasicIsGood
#298 is my favourite.

It has become warmer outside. So putting the player at an open window @0 degrees centigrade instead of -10 degrees centigrade did probably have more effect than switching caps. (Time was not shortened significantly, no condensation btw)
10 seconds with a hairdryer (through opened lid) are enough to make the effect disappear and the laser unit is directly exposed to the heat. I can even keep the CD inserted while warming up.

I can see that the tacking error signal coming from KSS-273A / KSS-274A is split for tracking error input and sled error input (Pins 27/28) at CXD-2515. Maybe changing the value of R102 could make the sled work in a tighter margin? (R102 would have probaly been a variable resistor in an old player)
All the best,
Salar
 
After a lot of music playing and testing of the startup of the player when cold, it has not failed one single time. I will therefore leave the modifications in place as I am satisfied with the result so far. Today I modified the bias of zener D976 by placing a 1K resistor across R973 (3,3K) giving a total R of around 750 ohm. The marking on the Zener says 3.9 V, but I measure 4,4 volt across it. Before the mod I measured 4,2 volt across it. This tells me that it has been biased well over the zener point. I will leave it as it is with the mod. The zener does not even get warm.
 
Hello Dacen!
Nice coincidence, I would have posted about my mods today as well.
Just to make things clear:
You had startup problems when the player was cold, but no playback problems, correct?

What I tried the last two days was aimed to tweak the sled error / tracking error signals.
Yesterday I replaced resistors R101/15k and R102/100k, carrying the sled error (R102) and tracking error (R101) signals
with multiturn trimmers, 20k and 200k.
Took an evening to glue in a small pcb and cut the old traces and rewire.
Today I played with the trimmers but surprisingly, it did not show any effect at all.
I turned the trimmers up and down in values of more than 50% no effect,
even on the scope where I monitored the tracking error signal passed through R101.
Interestingly the glitches were not visible on the scope.
Also no changing dc offset visible on the scope. This could have helped to determine the coil/sled ratio.

So putting in parallel to 3,3k 1k was your last fix to stabilize power?
The other big tweak was to give the laser its own power supply, but I do not understand why you added another 74HC logic...?
All the best,
Salar
 
Last edited:
"You had startup problems when the player was cold, but no playback problems, correct?"
Yes that is correct !
"So putting in parallel to 3,3k 1k was your last fix to stabilize power?"
Yes, but I doubt that this does any difference at all.
"The other big tweak was to give the laser it`s own power supply, but I do not understand why you added another 74HC logic...?"
This is just to give the player a double security equal to what is the case with players with a conventional drawer. As you can see the control of the laser (on-off)is handled by the output of pin 62 IC201 as a result of input from lid switch going into pin 45 IC201. I have combined this input from lid switch to cut the laser by controlling Q601 as well. This is the power feeding the laserdiode. As you can see there is also a disable function of the laser coming from the Power On Reset IC. This is just a protection during power up in the case that the lid is open and no CD is in the player. You could get a short
pulse from the laser into your eyes if this is the case. This signal is the other one I have combined in the 74HC00. Both these inputs are ending up in one output controlling Laser Power (Q601).
 
Just a little info to clarify. There is a connection from the lid switch through the 1 Mohm resistor JW601 originally. But the low resistor (JW600) 10 or 330 ohm to ground on the input of Q601 is preventing the base of Q601 to go high through the 1Mohm. So from what I can see this has never been working the way they intended. This must be a design flaw.
 
Last edited:
On the coils themselves, or connections to same, I guess.

Looking at an older Sony, the E and F signals go first to a pair of r.f. amps, one of which has variable gain for the adjustment of EF balance. From these amps the signals go to the two inputs of an opamp, the output of which is the tracking error. One of the inputs of that opamp is offset by means of a pot across the + and - 12V reg. supplies, and a high value resistor, to make a variable current source. The balance should be set to centre the signal about zero, and the offset set to zero, according to the manual. The tracking error then goes via a pot for tracking gain to the tracking error amp/filter. The output from this is then simply split to the tracking coil servo and sled servo, seemingly leaving the necessary sled delay largely to the inertia of the motor. There is no provision on this machine for altering the relative signal amplitude to sled and tracking coils.

Watching this machine running upside down, I see that the sled moves in very small steps with perhaps 2 to 3 seconds between jogs at the beginning of a CD, and about 6 seconds at the end. That may fit in with the frequency of your glitches.

You may find that some adjustments don't show up in obvious ways because of all the feedback. If your machine has no suitable testpoints or adjustments on the board or in the manual, then a look at the manual for an older machine might be helpful. The Sony I'm staring blankly at now is unfortunately the only one that doesn't have the same design of head. However, the principles are the same, and test procedures for focus and tracking balance and offset are well covered in many Sony manuals, and always the same AFAIK with the exception of this still-maddening CDP101 that sometimes gets lost.
 
Hello PlasticIsGood,
thanks for the reply. Yes I remember those things from servicing my Nakamichi OMS-5II CD which is Sony-based. Bringing it into perfect alignment, adding a digital output and later even building in a BuffaloII DAC (The Nak now serves as player and DAC) did fill some space here
Still my technical knowledge is limited.
I do now attach the CDP-X5000 servo circuit. Direct connections between servo control CXD-2515 and servo driver IC BA6297 are marked red for convenience. Also CXD-2515 & BA6297 circuit diagrams are added in this gif as well as BA6297 application example from the manufacturer.

Maybe someone could take a look...:)

All the best,
Salar
 

Attachments

  • CDP_X5000_Servo_Circuits_600dpi.jpg
    CDP_X5000_Servo_Circuits_600dpi.jpg
    463.9 KB · Views: 99
Same as I described, except some of the processes have been incorporated into the pick-up block amp, which has provision for focus bias and tracking balance adjustment, but not focus balance or tracking offset. The other functions are incorporated into IC101, which is why it has such a long description.

If you want the change the relative gain in the drives to tracking coils and sled, then try changing the gain of one of their respective drivers, which is set by external resistors AFAICS, but I can't read the text. Or attenuate one of the driver inputs.

It's a complicated system though, so tackling one problem may cause others.
 
The output of each driver's first opamp is available at a pin labelled "out", for example pin 18. In the datasheet example, the error signal goes to the inverting input via 10k, and 10k is also used for feedback, giving unity gain. All the examples in the sheet are unity gain which is not very auspicious maybe. Anyway, the gain is the ratio between the two resistors, so you could increase gain to the sled, for example, by increasing the feedback resistor a little. You may be the first person in the world to try. It will be interesting.
 
Thanks a lot to you both! I am only persistent because it is still cold, so still time to test, and I paid about 1200$ for pseudo-rugged high-end gear, that does not work in room temperatures that were normal to our grandparents - or somebody living in a classical japanese house in the winter...
Do I see it correctly that I should try to alter R135 in the X5000 datasheet? (Still have 2 hrs left to buy precision trimmers)
 
As far as I understand, R131 and R133 sum up to 66k, whereas R135 is 33k.
I am a complete layman. But this is no unity gain as being proposed in BA6297 application example by the manufacturer.
Guess this does not matter anyway, wether there is unity gain (this means 1:1 I guess) or a 2:1 ratio.
Still this gives me headroom to play a bit with feedback resistor R135/33k replacing it with a 50k cermet trimmer...
I would like to monitor the result with my scope. Guess I use R135 for that?
I also guess R136 should be not touched...?
 
Last edited: