Sony Walkman WM-DC2 restoration.

Only explanation I can come up with is that you have way more friction in PLAY than in FF/REW. Frozen clutch maybe ?
Given the W&F goes back to normal even when slightly pushing on the motor suggests the problem is mechanical.

 

That is a good point! Haven't thought of that. I will open and clean it.
But I should probably get a torque cassette to actually verify it or is there any other way to test if it's in a normal range?

Good thing I got this unit in a fairly good price cause the fixing is getting expensive with all the new tools

 

I cleaned the clutch, added the tiniest amount of PDP65 to the plastic ring. Did not help. Shucks!

 

Could this be somehow related to an electrical issue? Yesterday I thought I got the problem fixed after adjusting the motor position: the W&F was sitting arouind 0.09. I was listening to a tape and the batteries died halfway through. After I put in a fresh pair of Eneloops, the W&F jumped up like crazy and the vibration came back.

 

It's not impossible to be electrical, although that would mean the servo loop only locks when motor current is above a certain threshold.

As for the torque, check by hand: you should be able to stop the takeup reel with one finger. A frozen clutch would make the torque so high that it would require significant force to keep it stopped with 2 fingers.
Eventually compare the torque with another walkman.

 

Yes, the clutch and torque seem to be OK. Compared to my other DD units and D6C it seems to be about the same.

 

I took everything apart again and tried to run the motor and capstan without the center gear assembly. I still get a weird vibration from the motor that disappears when I gently add motor tension or touch the flywheel.

I also checked IC601 pin 15 with the oscilloscope.
With AC coupling, 50 mV/div and 1 ms/div, I get a very regular repeating waveform at roughly 280 Hz. The DC voltage at pin 15 is around 2.1 V.
When I lightly press on the motor and the mechanical vibration disappears, this modulation also becomes much smaller/disappears.

Does this waveform look normal for pin 15 of the CX20084, or does the amount of modulation seem excessive?





I am fresh out of ideas of what to check next. Could this be some kind of servo-loop instability, where the circuit never settles properly at the target speed and instead keeps correcting back and forth? Or does this still sound more like a mechanical resonance that the servo is simply reacting to?

 

I've seen cases where motor shaking (very low frequency vibration, a couple of Hz only) occured because the new capstan ring had some eccentricity on the outer diameter.
This however didn't affect W&F in a significant way as it does in your case.

Waveform doesn't look normal at pin 15, you should see just very small corrections, not something periodic.
Will attach some screenshots from other pins so you can make some comparisons (don't have one with pin 15).

 

Thank you, Valentin!

I'll check pins 9–14 so I can see where my waveforms first start to differ.

 

Here are the measurements from pins 15 → 9: https://imgur.com/a/pin-measurements-8hRlxle

I took two measurements of each pin: one with the motor vibrating as it is, and one while lightly adding motor tension, which makes the vibration disappear. The other pins don't seem to change nearly as much when motor tension is added, but the difference at pins 14 and 15 is very obvious.

 

Looks like the servo may be oscillating. Waveform is correct when motor is pressed.
My suggestion is to check all resistors and capacitors in the servo loop, since many were changed.
Also the MLCC caps make sure are C0G dielectric, not X5R/X7R, as on the latter the capacitance varies significantly with applied voltage which can introduce instability.

 

I started from Q601 transistor and went down the line. I had checked the C601 capacitor previously and it was OKish but decided to check again and it was faulty! Maybe I fucked it up when I resoldered it back, I dunno... I took a replacement cap from my donor DD3 and POW, the vibration was gone, everything runs smoothly and without problems. Wow & Flutter is hanging around 0.06%~. Feels good, man.

Thank you, Valentin, for all the help!

 

Out of curiosity, what was exactly wrong with C601 ? High leakege ? Short circuit ?

 

The C601 is a 10 nF capacitor, first time I went through the components it measured 8,7 nF. I don't know if that is enough difference to make the servo act as weird as it did but now that I took it out for measurement, I couldn't get a reading at all. Maybe I overheated it first time or somehow was able to destroy the contacts now that I took it out... What can you do when you have sausages for fingers (plus my soldering iron is a piece of shit.)

 

Value of that capacitor isn't critical, but having a small capacitance there is.
Purpose of that capacitor is exactly to increase stability by mitigating the Miller effect.

8.7nF reading seems ok, but no reading at all means the cap is open circuit, case in which circuit becomes unstable.
This explains why increasing the motor current (hence Q601 collector current) stabilized it.
C601 is a MLCC capacitor, these are pretty sensitive and can be damaged relatively easily. They usually fail short-circuit, but in this case it seems it failed open.