In normal play, the radial motor current is controlled by a function of the radial error signal. The mean current should therefore remain almost zero throughout the playing of a disc, reflecting the very slow movement towards the outside.
Coil current will be modulated about the mean by any eccentricity of the track on the disc, and by its own process of correcting error. It's a feedback system with several loops and sources of delay, so it is never perfectly settled.
In your case, I take it that the arm moves to its innermost extremity and focuses, but then can't move from that position. Normally it should focus, find the TOC and read it. Pressing play in stop mode presumably initiates a search for the TOC if it has not already been read, or for the first track if it has. I guess that's why your machine shows a change in voltage across the coil. I also guess that the voltage should reverse polarity, from driving the arm inwards to driving it outwards, until the TOC or the first track is found.
Have you checked the voltages on each leg of the driver transistors? A loss of one of the power supplies, or a transistor failure, could cause the problem..
I could check voltages on my own CD104 for comparison, but I'll need to dig it out and make space on the bench.
Coil current will be modulated about the mean by any eccentricity of the track on the disc, and by its own process of correcting error. It's a feedback system with several loops and sources of delay, so it is never perfectly settled.
In your case, I take it that the arm moves to its innermost extremity and focuses, but then can't move from that position. Normally it should focus, find the TOC and read it. Pressing play in stop mode presumably initiates a search for the TOC if it has not already been read, or for the first track if it has. I guess that's why your machine shows a change in voltage across the coil. I also guess that the voltage should reverse polarity, from driving the arm inwards to driving it outwards, until the TOC or the first track is found.
Have you checked the voltages on each leg of the driver transistors? A loss of one of the power supplies, or a transistor failure, could cause the problem..
I could check voltages on my own CD104 for comparison, but I'll need to dig it out and make space on the bench.
Yes the -/+ voltages are present, and I do get -/+ 11.5 volts at the bases of both transistors, as per service manual.
Also, I inspected the pre-amplifier circuit just to make sure and I could not find anything wrong there. The μA 741 is good.
My guess is that the radial motor amplifier works. I could not find a single place where the quoted voltage is wrong.
How do I check on the radial error signal then? Quoted values in PLAY mode are spot-on.
Bad CDM? I doubt it. I've seen wrecked machines with perfectly good ones, and this example is mint!
Also, I inspected the pre-amplifier circuit just to make sure and I could not find anything wrong there. The μA 741 is good.
My guess is that the radial motor amplifier works. I could not find a single place where the quoted voltage is wrong.
How do I check on the radial error signal then? Quoted values in PLAY mode are spot-on.
Bad CDM? I doubt it. I've seen wrecked machines with perfectly good ones, and this example is mint!
Soooo... it suggests that there is a problem with the pickup finding focus on "normal red book standard" discs. Odd problem to have. Even more unlikely is that the pickup is not in quite the correct position at rest and the "poor and wide tolerance" CDR just manages to overlap into that range.
Its worth looking at the laser current during focus search operation. There will be a low value resistor in the laser drive circuit to measure the volt drop across and hence calculate the current... although that in itself isn't perhaps much help. You could note the value and try altering it slightly (I wouldn't normally advise that but in this case worth trying). You can then reset it back having noted the value.
Sometimes pickups suffer from internal contamination of the optics.
Its worth looking at the laser current during focus search operation. There will be a low value resistor in the laser drive circuit to measure the volt drop across and hence calculate the current... although that in itself isn't perhaps much help. You could note the value and try altering it slightly (I wouldn't normally advise that but in this case worth trying). You can then reset it back having noted the value.
Sometimes pickups suffer from internal contamination of the optics.
How does that affect focus?
Laser current measures just fine, perhaps a little high if anything (1,38 mA).
Hmm. I took it from the initial post that the focus was working OK.
A high laser current may be a bad sign. It adjusts itself to maintain a constant output from the monitor diode AFAICS, and the trimmer adjusts the voltage generated by a given monitor diode output current, so basically it's a brightness control. A high laser current may indicate that it has adjusted itself upwards to the limit in a fruitless search for brightness.
However, it reads CDRs, which is mysterious. Perhaps laser currrent was adjusted using a CDR, and it is too bright for CDs? Seems a bit far-fetched but you might try turning it down. If the trimmer hasn't been moved since the Middle Ages, you'll need to clean it to get a reliable contact.
Are we sure that it's the focus that is failing on CDs?
What's the eye pattern like when reading CDR?
A high laser current may be a bad sign. It adjusts itself to maintain a constant output from the monitor diode AFAICS, and the trimmer adjusts the voltage generated by a given monitor diode output current, so basically it's a brightness control. A high laser current may indicate that it has adjusted itself upwards to the limit in a fruitless search for brightness.
However, it reads CDRs, which is mysterious. Perhaps laser currrent was adjusted using a CDR, and it is too bright for CDs? Seems a bit far-fetched but you might try turning it down. If the trimmer hasn't been moved since the Middle Ages, you'll need to clean it to get a reliable contact.
Are we sure that it's the focus that is failing on CDs?
What's the eye pattern like when reading CDR?
The player was never opened before, so nothing had been re-adjusted.
I haven't seen the eye-pattern - did not think of it, since no CD could be read!
What should I expect then?
Something else: it won't read the CDR to its end. It seems to lose focus at some point (here track 13/20).
I haven't seen the eye-pattern - did not think of it, since no CD could be read!
What should I expect then?
Something else: it won't read the CDR to its end. It seems to lose focus at some point (here track 13/20).
Last edited:
CDR eye pattern may be as much as 20% lower amplitude, but otherwise the same as in the manual. You're looking for a stable pattern with clear diamonds.
Failure to read later tracks can have a few causes. Now the machine is playing, you can do all the checks and adjustments in the manual. One that may change due to wear is the height adjustment, set by a screw under the bottom spindle bearing. Are CDRs the same thickness as CDs?
Failure to read later tracks can have a few causes. Now the machine is playing, you can do all the checks and adjustments in the manual. One that may change due to wear is the height adjustment, set by a screw under the bottom spindle bearing. Are CDRs the same thickness as CDs?
Do mean 138ma ? That's way over what would be expected. Normally current is around the 50 to 60ma (ish) region.
Philips were the odd one out in how they wanted laser current adjusting and measuring and used to specify a volt drop across a series resistor to the photo diode array rather than actual laser diode current. The actual current sensing resistor is usually a low value one in series with the emitter of the laser diode drive transistor.
The definitive test is always the quality and amplitude of the RF signal and it should look like this (scroll down page) You can see the effect a lens clean has too. Contaminated optics certainly reduce the amplitude of the signal. I'm no expert on optics but I imagine that "uneven" contamination affects the relative levels of "light" falling on what is a large area photo array. Does it affect focus directly ? I'm not sure tbh. Looking through internally misted binoculars gives poor detail definition. Wouldn't the same apply to light travelling through a pickup lens array ?
http://www.diyaudio.com/forums/digital-source/226288-sony-cdp790-kss240-restoration-project.html
The RF should be as clear as shown (but not with CDR which has indistinct transitions). The amplitude should be around 1.2 volts (as a ball park figure, I haven't got the CD104 manual to refer to)
Dried out electroylitics (specifically some 33uf caps on the servo PCB) were a common issue on later models) although I can't quote you any specifics for the 104.
Exactly. That's how they measure laser current. You get 575 mV across this resistance (470 Ω).
I tried to see if table height was off but it was actually perfectly aligned.
The eye-pattern is slightly fuzzy and shows exactly 1V PP, dead stable.
Last edited:
Correct scope grounding (using a ground that the RF amp uses) is essential to get the clearest eye-pattern.
Don't know what to suggest really over your player. I would perhaps "experiment" (just to see its effect) of altering the laser current slightly. The problem may all be pickup related anyway in the end.
Don't know what to suggest really over your player. I would perhaps "experiment" (just to see its effect) of altering the laser current slightly. The problem may all be pickup related anyway in the end.
Manual shows 1Vpp. Seems strange for a CDR but perhaps this machine compensates for the difference better than others?
I'm disappointed about the spindle height adjustment, as it might have explained everything, including the failure to read later tracks, where the effects of warping on disc height are greater. Looking at a few of my CDs and CDRs, there is a big difference in thickness, but it's the wrong way round...the CDRs are thicker so wear on the spindle bearing would show up on CDR first it would seem.
"Fault finding in CD players from eye-pattern observation" would be an attractive title for a paper. I found this page useful for fuzziness...not the same pattern but the principles are similar I guess...
http://cnx.org/content/m18662/latest/
I don't know if CDRs are inclined to show more or less noise and jitter. My CDM1 machines are a bit fuzzy too, but they still work well.
We're struggling here. What other differences are there between CDRs and CDs?
I'm disappointed about the spindle height adjustment, as it might have explained everything, including the failure to read later tracks, where the effects of warping on disc height are greater. Looking at a few of my CDs and CDRs, there is a big difference in thickness, but it's the wrong way round...the CDRs are thicker so wear on the spindle bearing would show up on CDR first it would seem.
"Fault finding in CD players from eye-pattern observation" would be an attractive title for a paper. I found this page useful for fuzziness...not the same pattern but the principles are similar I guess...
http://cnx.org/content/m18662/latest/
I don't know if CDRs are inclined to show more or less noise and jitter. My CDM1 machines are a bit fuzzy too, but they still work well.
We're struggling here. What other differences are there between CDRs and CDs?
Thanks for the information about the eye-pattern noise.
The fact is that any CDR I throw in the drawer will be read until about 60% when it stalls, with a lot of noise from the mech (not mechanical noise). It doesn't stop right away, but after a few long seconds of agony!
Back to track one it will gladly play again.
I tried an old - but as new - 1982 CD, extremely carefully made as they were at the beginning, but nothing commercial plays. The two miserable attempts, and stop. No TOC.
The CDRs play well, no skipping, perfect sound - well, you know what I mean.
As you put it, we are struggling indeed. Surely someone has already experienced this before?
The only discrepancy to the diagram is that I get -7V at pin 1 of IC 6212. Philips indicates -4.3V. All the other values are spot-on.
The fact is that any CDR I throw in the drawer will be read until about 60% when it stalls, with a lot of noise from the mech (not mechanical noise). It doesn't stop right away, but after a few long seconds of agony!
Back to track one it will gladly play again.
I tried an old - but as new - 1982 CD, extremely carefully made as they were at the beginning, but nothing commercial plays. The two miserable attempts, and stop. No TOC.
The CDRs play well, no skipping, perfect sound - well, you know what I mean.
As you put it, we are struggling indeed. Surely someone has already experienced this before?
The only discrepancy to the diagram is that I get -7V at pin 1 of IC 6212. Philips indicates -4.3V. All the other values are spot-on.
Is pin1 ic6212 correct [10.3V] when not playing? What about test point 51? This is back to tracking error processing. What's k-factor? There's a bit in the manual (annex IV) about checking it.
I don't understand much of the block diagram, or how it relates to the circuits.
There are four associated test points: 46, 49, 50, 51, that should all be equal in level but carry different waveforms when not playing. At a wild guess, k-factor (and d-factor) are ratios or functions of the outputs from the photodiode array, that relate to the shape and alignment of the spot of light on the disc. The CDM1 manual might be illuminating. If your processing is adding an offset, perhaps it thinks its centred on the track when it's not, quite.
There's a dynamic (playing) k-factor check too. There's no explanation of what it means or what to do if it's wrong.
I wonder if the width of a CDR track is greater than a CD, leaving less space between? That would make CDR more tolerant of radial error, perhaps.
I don't understand much of the block diagram, or how it relates to the circuits.
There are four associated test points: 46, 49, 50, 51, that should all be equal in level but carry different waveforms when not playing. At a wild guess, k-factor (and d-factor) are ratios or functions of the outputs from the photodiode array, that relate to the shape and alignment of the spot of light on the disc. The CDM1 manual might be illuminating. If your processing is adding an offset, perhaps it thinks its centred on the track when it's not, quite.
There's a dynamic (playing) k-factor check too. There's no explanation of what it means or what to do if it's wrong.
I wonder if the width of a CDR track is greater than a CD, leaving less space between? That would make CDR more tolerant of radial error, perhaps.
That could be normal.
The fact it plays CDR has to be a starting point. If it gets that far its 98% working.
One check I don't think has been mentioned in all this. Check the rails (all of them) with a scope for noise and ripple as well as correct voltage.
Hi again. PSU checked, ripple in good order, caps checked fine.
Now, I've got this CDR which has some visible eccentricity. I can see the servo doing its hard work: the swing arm doing just that—swing back and forth!
But the CDR plays fine nonetheless. Of course, still no luck with red book CDs.
Now, I've got this CDR which has some visible eccentricity. I can see the servo doing its hard work: the swing arm doing just that—swing back and forth!
But the CDR plays fine nonetheless. Of course, still no luck with red book CDs.
I really can't think of anything else tbh for this problem. I haven't been able to turn much up other than some mention of two versions of the system control processor (there's a CD104 and 104B identified by the number on the back. CD104/05 or 104/65 I think) This corrected a problem with the radial drive whereby the drive transistor/s ? overheat if a disc is left in the tray. As you mention not being happy with the radial arm behaviour it might be worth looking at.
I've definitely got the one where the transistor gets very hot! So the RAFOC being permanently close to the spindle seems to make sense then.
Would tired caps inside the CDM affect anything by the way?
The CDRs play so fine (listening to the latest PJ Harvey at the moment), even the bad ones.
Would tired caps inside the CDM affect anything by the way?
The CDRs play so fine (listening to the latest PJ Harvey at the moment), even the bad ones.
Last edited: