[Eryan]

Hello forum!

I just got my hands on a couple of Sony D-90 (D-9 for the folks in the US). Both were in very good cosmetic conditions, but both were not working. They turned on, the CD turned, but the laser did not move (the laser sled that is); pointing to the dreaded middle gear. So, I got my repair manual, soldering iron, and started repairing.

The problems inside both units were the usual suspects: middle gear of death was green and broken in several pieces, effectively blocking both the sled motor and the sled from moving. A few caps had leaked (both SMD and surface mounted) but only one of the units had damage to the traces. Replacing some of the caps required to find material to weld to, since the pads for a couple of caps were corroded away.

So, after replacing the gears, changing the caps and cleaning everything I reassembled and gave them a try. And they both did not worked. Or at least, they showed the exact same issue as the beginning: the laser sled was not moving.

I disassembled both units again and started troubleshooting. The sled motors work. I disordered the wires and drove them with a power supply. They started running at ~4,2V. I then welded the wires again and turned on the Discman while measuring the voltage on the SD+ and SD- terminals. To my surprise, there was no voltage applied. The sled motors were receiving no power from the motor driver. The issue happened with both an external power supply and batteries.

I went back to the schematics and realized that the D-90 uses a Motorola MCP1715FU IC to drive all four motors (CD, sled, focus and tracking). Inside it, there are four H-bridges that drive the above components. I am guessing that the motor driver got damaged trying to move the blocked sled motor. I cannot see any apparent damage on the chip. I also checked the transistors feeding the IC, (Q437 and Q453) but they both look fine.

Therefore, I would like to know if someone has experienced something similar. The middle gear problem is widely spread, and someone might have had similar issues with the electronics.

I'd appreciate any suggestions or ideas.

 

[SonyNutter]

How did you go? Does it do a focus search?

I am trying to fix a Sony D22 with the same chip. Suspected the MCP1715, but have since validated its fine. The service manual for the D22, which has a watermark of freeservicemanuals.info, has a block diagram of the chip, and each output is identical for both motors and both lens coils. So if any of those functions work then you can be sure that at least the supporting circuits for the chip are fine.

I'm really struggling with mine. Does focus search, locks, good eye pattern, then fails to load the track table and starts again. Starting to think its some kind of logic failure.

 

[SonyNutter]

Oh, note that you can put the player into test mode and advance and retract the sled manually using the track skip keys. You close a solder bridge near the system control IC. You can also command a focus search, so again you can use the focus circuit as a reference.

With those features, you could measure the command signals on pins 42, 41, 37 and 36 (in the D22 pin 41 is the sled) and compare them to the outputs, pins (2, 44), (7, 9), (32, 34) & (25, 27) respectively

 

[Eryan]

How did you go? Does it do a focus search?

I am trying to fix a Sony D22 with the same chip. Suspected the MCP1715, but have since validated its fine. The service manual for the D22, which has a watermark of freeservicemanuals.info, has a block diagram of the chip, and each output is identical for both motors and both lens coils. So if any of those functions work then you can be sure that at least the supporting circuits for the chip are fine.

I'm really struggling with mine. Does focus search, locks, good eye pattern, then fails to load the track table and starts again. Starting to think its some kind of logic failure.

My D-90s are still a work in project. The sled issue was purely mechanical. I spent a few hours polishing the brass pins for the gears with a cotton loop and some polishing compound. After that, the gears moved freely, but that did not solve the focus issue. I ended replacing every single electrolytic capacitor, and swapped the laser from a donor unit. The Discman now plays CDs, but only on batteries, and skips quite a lot. Due to the laser swap, I need to recalibrate it. I'll get an analog oscilloscope this week to try to calibrate it (the dreaded Sony potentiometers).

In your case, the issue might be power related. From what I have read, these units tends to be picky regarding power delivery. Does the issue happens both on batteries and AC, or only in one of those? Have you checked the electrolytic capacitors in the power stage/regulation?

 

[Eryan]

Oh, note that you can put the player into test mode and advance and retract the sled manually using the track skip keys. You close a solder bridge near the system control IC. You can also command a focus search, so again you can use the focus circuit as a reference.

With those features, you could measure the command signals on pins 42, 41, 37 and 36 (in the D22 pin 41 is the sled) and compare them to the outputs, pins (2, 44), (7, 9), (32, 34) & (25, 27) respectively

I am aware of the test mode. In the D-90, you need a wire to bridge the pin. I was not aware however of the pinout for the command signals. I might try that on a D-11 I am trying to revive (swapped the caps on the power stage/regulation and now it fails to play any CD).

 

[SonyNutter]

Stupid website. 'Sorry we had a problem'. Just lost my whole reply

So its not the power supply. I can see the whole circuit for power regulation and the power brick and the battery go into the same point. I have adjusted the circuit to the required 5.3V and 3.5V and behaviour has never changed. They are not fussy about power supplies.

Have noticed that the voltage to the sled motor out of the Motorola chip is asymmetric depending on whether it is going in or out (by operating the sled manually in service mode, using the track forward and back keys). Measuring the input to ch 2 of the Motorola chip ERR2 (pin 41), which comes directly from the Sony servo controller chip, the voltage swings are symmetric. The PWM out of FLIN2 (pin 5) changes its duty cycle depending on direction, causing a different voltage at HIN4 (pin 4). So something in system around the Motorola chip reacts differently depending on the direction of the command. Can see the stable sync going in (or out?) on pin 25 and a good sawtooth on COSC (pin 22). So looks like the part that generates the PWM is OK. It looks like a reference voltage could be off. But from what I can see all reference voltages are pulled from the same power rail and are set with voltage dividers, which shouldn't drift. So I don't quite know what is going on. Wish I could find the datasheet of the Motorola chip to understand more.

I suspect the playback problem is caused by the gain asymmetry, as channels 2 and 4 on the Motorola chip, sled and spindle, seem to be linked. Focus and tracking are OK, can see a good eye and the FOK signal triggers. But the spindle speed seems to hunt about once per revolution, seen by a stretch and contraction of the eye pattern and spiking of the spindle motor voltage. This could be caused by the gain asymmetry, basically by the servo over accelerating the disk and then having a weak ability to 'catch' it.

Sounds like you already know this, but be careful adjusting lasers, have read they are very easy to blow and have short lives if overdriven. All optical blocks I believe have a feedback photodiode in the optical block and the Automatic Power Control (APC) uses this. Sony blocks I believe are preadjusted at the factory, and have the individual diode current printed on the sticker. You measure and check the current by measuring the voltage across a resistor, and two test points are usually provided on the circuit board for this. The D22 has them, but are unmarked and need the service manual to find them. I have played around with adjusting a couple of diodes where I had a weak eye pattern amplitude, and where I had many spare laser blocks, but it didn't seem to make playback any better. Measuring Eight-Fourteen Modulation (EFM), which is the down-stream cleaned-up (squared-up) version of the RF eye pattern, on a 2ch digital scope with no persistence, I could see that the signal was correctly interpreted despite the low amplitude RF eye pattern. So I have decided that low laser power is rarely an issue if the focus system is working.

This book has a great section devoted adjusting lasers https://archive.org/details/understandingandservicingcdplayers

It also has much general advice for making adjustments. I have never heard of any adjustment be referred to as a 'dreaded pot', and i've revived a few players just through a tune up.