I remember some Sony manuals showing a drop of hypoid oil to be used on the chromed slide rods. That was on early players like the CDP101 and the D50 portables. They even went as far as to say if the mech was tilted at some quoted angle that the pickup should slide under its own weight freely end to end.
As with most things the engineering on first generation stuff was excellent, then as the electronics got better (servos, error correction etc) the integrity of the mech becomes less important as they relied on the better servos to keep it all together.
As with most things the engineering on first generation stuff was excellent, then as the electronics got better (servos, error correction etc) the integrity of the mech becomes less important as they relied on the better servos to keep it all together.
Hi Karl,
Exactly. The head was supposed to be free enough to move under gravity only.
I think the determining factor in allowing the mechanisms to decline in quality was the improvement in DSP error correction, and more importantly, concealment. When the KSS-150A mechanism first appeared, we all laughed at it. It was a cheap, nasty thing comparatively speaking. When the KSS-150A heads began to fail, we all said "what do you expect?". The KSS-210A fixed those issues and Sony learned a lot from that. Today, I am happy to see a KSS-210A or KSS-240A type head and mechanism.
Other little horrors sold to the public were the HOP-M3 and SF-90 heads. No sleeve bearings resulted in fast wear. Plus, they would begin to dog track on the rails, then bind.. Went through buckets of those cheap heads. One thing though, the SF-90 family had a very good eye pattern. The HOP-M3 was highly variable (what quality control?).
Then Philips out did everyone for crap. The VAM series of transports. Pure junk with zero quality control. They didn't even release a schematic of the RF board. This transport was the final straw for most manufacturers and they abandoned Philips in droves (finally). Philips was very nasty in that they did not make spares available. The CDM-9 transport was popular, but the failure rate was higher and spares prices went through the roof. Many players hit the trash due to this. Too bad.
-Chris
Exactly. The head was supposed to be free enough to move under gravity only.
I think the determining factor in allowing the mechanisms to decline in quality was the improvement in DSP error correction, and more importantly, concealment. When the KSS-150A mechanism first appeared, we all laughed at it. It was a cheap, nasty thing comparatively speaking. When the KSS-150A heads began to fail, we all said "what do you expect?". The KSS-210A fixed those issues and Sony learned a lot from that. Today, I am happy to see a KSS-210A or KSS-240A type head and mechanism.
Other little horrors sold to the public were the HOP-M3 and SF-90 heads. No sleeve bearings resulted in fast wear. Plus, they would begin to dog track on the rails, then bind.. Went through buckets of those cheap heads. One thing though, the SF-90 family had a very good eye pattern. The HOP-M3 was highly variable (what quality control?).
Then Philips out did everyone for crap. The VAM series of transports. Pure junk with zero quality control. They didn't even release a schematic of the RF board. This transport was the final straw for most manufacturers and they abandoned Philips in droves (finally). Philips was very nasty in that they did not make spares available. The CDM-9 transport was popular, but the failure rate was higher and spares prices went through the roof. Many players hit the trash due to this. Too bad.
-Chris
I just scored another success with an $11 USD generic KSS-240A.
I have this Denon DCM-320 5-disk changer:
It had a rather odd fault. Focus search and initial spin-up of disk worked fine every time. I could even see a decent RF signal of about 1.2V P-P for a second or two. Then, at the point where a functional CD player moves the sled to the first track, this machine would stall out. The RF signal dropped in level and got totally random. I don't think the EF tracking signal was getting out of the pickup. Occasionally it would actually start working, maybe one time out of 10. But not for long.
So I tossed in one of my $11 USD generic KSS-240A pickups.
Bingo!
It now works perfectly and steadily.
I was able to improvise some cribbing to hold the transport while bench testing it:
CD changers are always a pain to bench test because the transport is buried underneath the changer mechanism. Fortunately this one has long cables for the transport.
-EB
I have this Denon DCM-320 5-disk changer:
It had a rather odd fault. Focus search and initial spin-up of disk worked fine every time. I could even see a decent RF signal of about 1.2V P-P for a second or two. Then, at the point where a functional CD player moves the sled to the first track, this machine would stall out. The RF signal dropped in level and got totally random. I don't think the EF tracking signal was getting out of the pickup. Occasionally it would actually start working, maybe one time out of 10. But not for long.
So I tossed in one of my $11 USD generic KSS-240A pickups.
Bingo!
It now works perfectly and steadily.
I was able to improvise some cribbing to hold the transport while bench testing it:
CD changers are always a pain to bench test because the transport is buried underneath the changer mechanism. Fortunately this one has long cables for the transport.
-EB
Oops! Not so fast! The replacement KSS-240A performed flawlessly for a whole day.
Then the exact same original fault returned:
- Focus search is fine.
- Disc spins up at normal speed.
- RF signal amplitude & eye pattern is perfect for 1 second.
- But in this machine the RF signal amplitude drops 50% & gets very noisy as viewed with the scope.
- The sled never moves.
- There is an audible hissing sound from the transport.
- The spindle motor RPM begins to vary & eventually speeds way up or stops completely.
- Perhaps there is no tracking error signal coming out of the KSS-240A?
- Or the later sections of the servo system are failing to process it?
- Or perhaps the sled drive motor has a “dead spot?”
BTW this Denon DCM-320 5-disc changer still has an old garage sale label on it:
“$10 - works sometimes.”
I set it aside years ago because this fault was so intermittent. I inspected & re-inspected the main PC board closely for bad solder joints but found none. Checked all electrolytics (they were all fine). Measured power supply voltages & ripple (also OK). Tried a different flex cable between KSS240A and main PC board (no change - still had same fault).
-EB
Thanks!
Unfortunately after extensive testing I didn't find any issues with the sled or its drive motor. When the malfunction occurs the sled servo gets disabled (0V output to the sled motor). Also when I can get this machine to initialize a disc then it always plays it all the way through to the end. The drive voltage going into the sled motor never exceeds +/- 1V except while the sled is returning to the home position or jumping to a different track on the disc.
In fact, when this machine malfunctions, both the sled drive signal and the tracking coil drive signal get disabled. Oddly the tracking coil drive signal settles to a DC level of about 1V rather than 0V.
The focus drive and spindle motor drive continue to operate in a "somewhat controlled" manner, but with a considerable amount of noise and instability.
I'm getting suspicious that the serial pathway for digital "command" inputs is flaky. This isn't a standard SPI or I2C bus.
It does have 3 lines: Clock, Data, and "XLT."
All 3 are outputs from the MCU to the servo system.
The MCU briefly pulls the "XLT" signal low to latch in the data after 8 bits of serial data have been sent.
Another odd thing is that two input pins of the CXA1372 servo processing IC are left floating.
These are:
Pin 24, "S.STOP" or "sled stop." In some CD players this pin is connected to the "home position" sensor switch for the sled. An external pullup resistor on the PC board is required according to the CXA1372 datasheet. But for the DCM-320 this task is handled by the MCU. I'm always suspicious of floating input pins, even in the case of a bipolar IC like the CXA1372. I intend to connect this pin to +5 with a 10K pullup resistor to see if it makes any difference.
Pin 28, "LOCK." This is an input which shuts down the sled servo to prevent over-run. According to the CXA1372 datasheet there is an internal pullup resistor of 47K. This pin isn't connected in the DCM-320.
Sigh! This project is turning out to be more difficult and time-consuming than what I expected. Still, I do expect to learn something from it.
-EB
RF drops... I think I would be looking at a focus issue here.
Sled doesn't move... may be in response to low RF
Hissing... is what the focus coils can do when focus gain is to high.
Spindle motor... again that may all be related to low RF.
Think what happens when you do a track jump, you lose the normal tracking but focus is still preserved and amplitude of the RF is preserved. yes the RF looks awful but only because the tracking (normal tracking anyway)is disabled during that operation.
I'd be investigating the focus side of things.
Excellent recommendations Karl & Chris! Especially using the "test mode."
I find it interesting that a Japanese CD player has a service mode similar to what Philips used back in the day.
Service mode for this Denon DCM-320:
- Mode 1: (STOP button) shuts everything down.
- Mode 2: (PLAY button) Activates focus servo & spindle motor only. Tracking servo & sled remain inactive. (This is the perfect mode for adjusting E-F balance.)
- Mode 3: (PAUSE button) Additionally switches on tracking servo & sled movement. In this mode the machine should play a disc.
This diagnostic header is actually just a group of jumpers on the main PC board.
It doesn't have a socket or stand-up header pins, but it works fine with small EZ-hook clipleads.
I'm still amazed that this machine always succeeds at the initial focus search & TOC read. This works 100% perfect every time.
The fault always starts just after reading the TOC, at the point where the sled would normally move a bit in order to begin playing the first track on the disc.
Another unusual behavior:
- The entire front panel FL display goes "dark" when this fault starts up.
- The disc continues rotating & the focus servo continues to operate but the focus/tracking actuator becomes audibly noisy.
- There is no tracking signal applied to either the tracking actuator coil or the sled motor.
- Sometimes pressing the STOP button multiple times or holding it down will recover things to where the disc rotation stops & the display lights up again.
- Other times the machine remains in the "locked up" state until power is cycled.
If that coincides with the fault then this might be a way in... it has to be a clue worth following.
Its worth looking at basics like supplies etc to the FL display. Try and find why that goes out. Switch your thinking over from looking for a playability fault to looking for a no display fault. It may well lead somewhere.
Could the focus amp have some problem that is effectively pulling a rail down? Maybe both devices in an NPN/PNP pair turning on... one maybe intermittent.
Ah-Ha!
I think that has a vertical DSP PCB on the main PCB.
Desolder the DSP PCB carefully. With a little liquid solder flux (made for electronic soldering!), and using an electronically controlled temperature soldering station with a grounded tip. Apply flux around the DSP chip, using a chisel tip, resolder the pins. The flux will keep the solder from shorting pins and make a nice factory looking job. Then do the same to the memory chip. Clean up with lacquer thinner and a cotton swab - cardboard or wooden shaft only!! Inspect the connections carefully for shorts or idsconnected pins.
Install the vertical PCB in the main PCB again, use the flux and clean up afterwards. Plug in and enjoy your success!
-Chris
I think that has a vertical DSP PCB on the main PCB.
Desolder the DSP PCB carefully. With a little liquid solder flux (made for electronic soldering!), and using an electronically controlled temperature soldering station with a grounded tip. Apply flux around the DSP chip, using a chisel tip, resolder the pins. The flux will keep the solder from shorting pins and make a nice factory looking job. Then do the same to the memory chip. Clean up with lacquer thinner and a cotton swab - cardboard or wooden shaft only!! Inspect the connections carefully for shorts or idsconnected pins.
Install the vertical PCB in the main PCB again, use the flux and clean up afterwards. Plug in and enjoy your success!
-Chris
I recall that small vertical PC board from some of the single-disc Denon CD players. However in this DCM-320 5-disc changer everything is located directly on the main PC board. No daughter boards. Also the flat flex cable from the KSS-240A goes into a ZIF socket on the main PC board. There's no small PC board underneath the transport.
A new test result of interest:
Quoting from the thread titled "Nakamichi CD4 problem:
When this Denon DCM-320 is acting up and simply refusing to actually play any disc, then if I spray a little bit of freeze spray onto its CXA1372S servo IC followed by power-cycling the machine, then the result is always that it begins to work perfectly (for a period of time).
There is absolutely no bad soldering anywhere on this PC board. It looked pretty good to start with and I touched up every location that might have been even slightly dodgy.
Now I'm trying do decide whether I should drag another CXA1372S out from one of my parts units and try installing it in this DCM-320. Interestingly this CXA1372S doesn't even warm up when the player is running.
-EB
Understood. I will wick off most of the existing solder and redo with 63Sn/37Pb. This CXA1372S servo IC is in a “through-hole” 48-pin SDIP package. This same chip also comes in a QFP package (CXA1372Q) for applications like the little PC board directly underneath the disc transport mechanism in many Sony-brand CD players which use the KSS-240A.
The DSP side is a Sony CXD2500AQ in an 80-pin QFP surface-mount package underneath the PC board.
Results of today’s tests:
Flex the PC board and tap on all the components: No improvement.
Unplug/replug all cables that have receptacles on the main PC board: No improvement.
Apply freeze spray to every IC except the CXA1372S: No improvement.
Apply freeze spray to CXA1372S: This always gets the machine to function for a period of time.
-EB
This Denon DCM-320 5-disc changer has a relatively unsophisticated power supply. All of the servo and motor drives operate on unregulated +/-8V DC rails. The digital section is powered by +5V regulated down from the +8V rail with a linear regulator. There is also a regulated -5V rail which powers the dual PCM61 DACs along with the same regulated +5V rail which feeds the MCU, servo, & decoder IC. The VFD has a -24V supply which is used only for the display itself.
The +/-5V rails are always totally stable.
The +/-8V rails vary by about +/-300mV depending on which motors and servos are operating.
I haven’t observed any larger variations in these DC rail voltages.
Curiously the analog line-level audio outputs come straight out of the PCM61 DACs. There are no additional opamps in the analog audio output section except for the dual channel headphone amplifier IC. The PCM61 operates on dual +/-5V supplies. I believe the headphone amplifier IC also operates on +/-5V.
It seems that many Japanese CD players use +/-5V rails for everything in the analog audio output section. That can be an issue for people who wish to mod their CD players with different opamps: There aren’t many opamps which work well with only +/-5V supply voltages. Some clip frequently & severely on +/-5V supply rails.
-EB
Update on my Denon DCM-320 5-disc changer restoration:
I confirmed its CXA1372S servo IC is intermittently failing:
It's likely that one pin goes open-circuit intermittently.
Giving the top of the IC package a very brief dose of "freeze spray" instantly fixes the intermittent every single time.
I've repeated this experiment more than 10 times.
Nothing else corrects the fault.
I haven't tried to identify precisely which pin is intermittent.
I suspect one of the data input pins (DATA, CLK, XLT).
Working together these 3 pins receive 8-bit serial commands from the MCU.
However answering this question this doesn't really matter.
I just need to replace this IC. Hopefully I have a parts machine to get one from.
So... I will set this unit aside for a bit.
I'll continue with 2 other multi-disc changers which are on my shelf:
To keep this thread on-topic I'll only post here about CD players which use the KSS-240A.
I'm still surprised this Denon DCM-320 actually has a bad IC chip.
IC chips don't fail at anywhere near the rate of stretched belts, dirty lenses, dry solder joints, bad electrolytic capacitors, bad motors, or faulty optical pickups.
This particular IC chip doesn't even get warm during operation.
There's no real reason why it should fail now, except that it may have been like this ever since it was manufactured.
One potential weak point with IC chips of this vintage is their incredibly thin "bond wires" which connect metalized pads on the silicon die to the external pins. The probability of one faulty internal connection rises with the number of pins. The CXA1372S has 48 pins.
-EB
I confirmed its CXA1372S servo IC is intermittently failing:
It's likely that one pin goes open-circuit intermittently.
Giving the top of the IC package a very brief dose of "freeze spray" instantly fixes the intermittent every single time.
I've repeated this experiment more than 10 times.
Nothing else corrects the fault.
I haven't tried to identify precisely which pin is intermittent.
I suspect one of the data input pins (DATA, CLK, XLT).
Working together these 3 pins receive 8-bit serial commands from the MCU.
However answering this question this doesn't really matter.
I just need to replace this IC. Hopefully I have a parts machine to get one from.
So... I will set this unit aside for a bit.
I'll continue with 2 other multi-disc changers which are on my shelf:
- Sony CDP-C201
- Onkyo DX-C120
To keep this thread on-topic I'll only post here about CD players which use the KSS-240A.
I'm still surprised this Denon DCM-320 actually has a bad IC chip.
IC chips don't fail at anywhere near the rate of stretched belts, dirty lenses, dry solder joints, bad electrolytic capacitors, bad motors, or faulty optical pickups.
This particular IC chip doesn't even get warm during operation.
There's no real reason why it should fail now, except that it may have been like this ever since it was manufactured.
One potential weak point with IC chips of this vintage is their incredibly thin "bond wires" which connect metalized pads on the silicon die to the external pins. The probability of one faulty internal connection rises with the number of pins. The CXA1372S has 48 pins.
-EB