Sony WM-DD33 speed issues (and more..)

Hi everyone!

Up until now, my journey of restoring my old Sony DD Walkmans have been more or less straightforward.

However, I recently bought a WM-DD33 with many issues, feeling confident that I would be able to solve them. I was wrong.

It has at least two issues that I need some advices in order to fix them.

1) it has a spot of rust on the chassis, and I would like to know what is the best way of treating it and the right paint to use.

2) it has speed stability issues.
When using a 3kHz calibration tape, almost all the times I press play, it locks to +-2800Hz. On the other times, it locks to 300.5Hz. When this is the case, it has two behaviors: on some occasions the W&F is good (0.05%) and other times not so much (0.2%).
I thought the issue was on the quartz lock, so I unbridged the PLL to remove it. But now, even adjusting the RV601 I can only get +-1kHz!! I also checked the capacitors and they seem ok.

Any advices or suggestions are more than welcome.

thank you!

Best Regards,
Jose

 

Hi Everyone,

I have taken some measurements with batteries - IKEA LADDA 2450 - without cassette and PLL unbridged, and I read the following voltages on IC601 pins:

1 - 1,163
2 - 1,171
3 - 1,180
4 - 1,460
5 - 1,522
6 - 1,492
7 - 2,606
8 - 2,603
9 - 1,06 to 1,13
10 - 0,19 to 0,21
11 - 0,500
12 - 0,155
13 - 0,170
14 - 0,166
15 - 1,960
16 - 0.00 (GND)

There are many values that are way off.
I will get my oscilloscope out of storage and chip the waveforms at pin 4.

Best Regards,
Jose.

 

Some of the values are indeed off. But the fact PLL is disconnected and battery voltage is not 3V will influence this circuit.

I suggest changing the strategy because we're really not interested in all the voltages but rather if some key areas of the circuit are working:
1. The voltage reference. This is present at 1.52V and it's close enough to the specified 1.6V to believe it's fine.
2. The sawtooth waveform at pin 4. You should have this waveform there even with the PLL disconnected.

When these ICs don't work (usually due to walkman being powered with reversed polarity), the reference voltage is not present.

However, before suspecting anything electronic I would want to make sure there's no mechanical problem. Would check the following:
1. Is the motor rotating freely ? Remove the motor and rotate by hand in both directions.
2. What current does the motor draw with no load ? Power the motor with a bench power supply or battery (1.5V) and measure the current consumption. Should be 10-20mA.
3. Is everything clean mechanically ? Clean the rubber ring with IPA, motor spindle, capstan and pinch roller just to be sure.

 

Hi everyone,

Sorry for being away for such a long time, but unfortunately I had to deal with some prolonged health issues. The good news is that I am ok now, so I can return to my hobbies

I have done a couple of things:

• changed the capacitors to new ones;
• looked for broken traces on the PCB, but could not find any;
• check with the oscilloscope the wave forms on pin 1 of IC602 and on pin 4 of IC601 - When it is locked correctly the waveforms are very good. When it is locked at a lower speed I can see the waveform on pin 4 IC601 a little bit erratic and with the top tip cropped.
• cleaned everything, specially the rubber ring;
• I can turn the motor by hand but it seems the shaft is bent? Is this even possible?
• Also, could a faulty capstan FG cause this?

I will now remove the motor and check the current load when spinning with a stable 1.5V power supply.

 

What do you mean by motor shaft being bent ? Does the spindle have wobble ?
Bend it can be, but only if someone opened the motor before and bent it by accident or if the motor was dropped.
But if that was to be the case, it would be impossible to obtain W&F of 0.05% WRMS.
However fact shaft isn't bent doesn't rule out a potential problem with the motor.

The FG coil usually causes the motor to run fast. Also, on these DDs FG signal is also used for auto-off. So if there was no FG signal, motor would stop after a couple of seconds.

Suggest desoldering the motor and running it (installed in the walkman) on a 1.5V power supply.
See if you can obtain a consistent speed and consistent W&F (even if high without servo control).
This way you will be able to have some direction if the issue is mechanical or in the drive circuit.

 

Hi Valentin,

thank you very much for your help!

I have used a bench power supply and dialled to 1.5V - The actual value measured by an external voltmeter was 1.51V.

I then used a Amp-meter in series to measure the current draw and I got two values:

• if the motor is in horizontal position - laying flat on the table - I got 7.7mA
• if the motor is in vertical position with the shaft pointing up, I got 8.2mA

Are this values normal? Should I expect differences depending on the orientation?

Regarding the shaft being bent, what I see is that when it is spinning and I looking to the shaft dead on, the center of the shaft is not in same spot - it describes a quind of an elipse - ensemble I saying it wobbles.
Someone before me tried to fix, that I am sure, because this Walkmans center gear had already been replaced and I bought it with this issue already described.

 

Current draw with no load (it applies to pretty much all coreless motors) should be 10-20mA.
In your case, below 10mA is perfect no concerns there at all. Fact it varies a bit with position is normal, because friction from the washers is higher when vertical.
Your variation of 0.5mA is negligible. If you had a 5mA variation, for example, that would be a concern.

About the wobble, check if entire spindle wobbles.

At this point, adjust the external voltage so tape speed is closer to 3kHz and check if you are able to obtain a consistent speed.

 

Hi,

I did not have time to properly do the test suggested but I briefly tried to lock to 3khz using the external power supply connected to the motor and I couldn’t. W&F was all over the place around 0.8%. I will try to take my time later today to calmly redo the test.

What motors are compatible with the DD33? Only the DD30?
I can try to use another to test the mechanism.

 

Any motors from the center gear DD walkmans (DD, DDII, DDIII, D3, DD100, DD30, DD33, DC2) are interchangeable.
The ones that aren't are the DD9 and low-cost series (DD1, DD10, DD11, DD22).

Note there are 2 versions of the DD motor, a later one with Namiki sticker and earlier ones with SONY sticker. Both are made by Namiki and interchangeable.

 

Thank you Valentin for your help.

I today received a package of Walkman DD parts that I bought from eBay, and I now have two motors (and other parts) to also test. They aren’t from DD33, but from DDII but since they can be used I will try and see if I get any luck

 

Hi!

So, armed with replacement parts I discovered that:
- new motor, it runs smoother but I can see it shaking inside the cage;
- exchanged the capstan with another one and the shaking disappeared, so I assume whatever bent the motor also made something on the capstan. I used the same new capstan rubber ring on both so it was not that causing the shaking.
- cleaned, oiled and greased following the guide. Noticed some noise of teeth rubbing and discovered that the height of the small cog on the FF was shorter by a millimeter. Someone cut it? I changed it and all is good now.
- I connected the motor to the Bench Power Supply and I can make it work around 3010Hz but varies too much and WF is around 0.1-0.2%. The PS reported 1.5-1.6V and 0.08-0.09A.
- I then connected the motor back to the PCB and run the test and I get the same initial issue: sometimes it locks at 3KHz and WF is 0.08% and other times it is at 2.6KHz with high WF.

Could other mechanical parts do this, or I can assume it is electrical at this point? Should I try to add ruggedness to the capstan part that is in contact with the tape?
Could a bad adjustment to the tape path cause this?

 

W&F of 0.2% is ok given there's no servo control. How much does the speed vary in this setup ?

Roughening the capstan is something would recommend doing as there's no downside in doing it. Seen a couple of situations where a polished capstan had the symptoms of low speed and high W&F.
As far as mechanical problem goes, a 400Hz drop in tape speed would mean a lot of resistance which you would be able to feel by hand and would also create some audible noise.

Before drawing the conclusion, measure the voltage at the motor when speed drops. If the voltage drops also, it's obvious problem is in the servo.

 

Ok, so I just tested with batteries, measuring 2.68V under load and these were my findings:
- when synced at 3KHz, I measured at the motor 1.69V +- 0.01V.
- when not synced, I measured 1.60V +-0.01V.

I also noted that with fresh batteries (rechargeable, measured 2.78) the issues does not occur so often.

 

Would be interesting to test it with alkalines to see if there really is a difference. Although you would need thorough testing to be sure this is not a glitch.

The 90mV difference you measured explains the difference in tape speed, so this confirms it to be a servo problem.
In regard to the +/-1V, you probably meant +/-10mV (0.01V) instead.
When the servo feedback loop is stable, the third decimal should vary quite a bit, while first 2 should be stable.

Given you say the sawtooth waveform gets chopped off at the top when servo becomes unstable, I would check/replace C616.
It's possible this capacitor is starting to fail after a certain voltage, which would explain the behavior.
It would also be interesting to see what happens with reference voltage (pin 5) when servo becomes unstable.

 

Yes, I meant 0.01V (10mV) - edited and fixed on my previous post.

I have just replaced C616 with another 10nF 100V capacitor and it remains the same.

when it is not synced, pin 5 does not display any waveform, just a constant value of 1.5V. On pin 4 the wave form seems to be constantly rising until it starts to be cut off and resets itself. It does seem to be some capacitance issue but the caps are brand new and I can seem to find any broken trace (yet).

this last values and waveforms were taken with an external power supply set to 3.0V

 

Hi.
I have DD33 with exactly the same behavior. But it got worse over time.
It started exactly as you described. Sometimes locks on lower frequency (and I see erratic behaviour on pin 4 of servo IC), but it could work normally sometimes. After some time, it got even worse - depending in the moon cycle, it would not start at all. But giving a spin to the capstan allowed it to start at 3000/2600 speed with 50/50 chance.
Now when pll jumper is disconnected, I see no waveform on pin 4, and speed of the capstan is maybe 10rpm. Speed can be adjusted from standstill to 10 rpm. But according to the manual, it should be adjusted normally, but without quartz lock, wow/flutter should be higher. In my case there is almost no movement of the motor in that state.

What I've tried: Replacing all electrolytic caps: No improvement. Checked the replaced caps with esr meter and capacitance meter - they are good.
Checked timing capacitor (C616): It is good.
Checked all other signals: they are logical: 20mv p-p on pin 1 during normal rotation, vref is 1.54v, motor drive amplifier behavior is stable.
Checked speed potentiometer and R621 - they are within specs.

Then I started experimenting. Accidentally I discovered, that with PLL jumper connected, motor can be started not only by rotating it, but by touching one of the C616 pins (the one which goes to servo IC CHA1423N).
I dug deeper, and discovered that during touching the pins (I've set my scope to single waveform) and during touching the capacitor, voltage on pin 4 was increased by 0.08V, and this started waveform generation.
I've added pull-up resistor from pin 4 to pin 7 of servo IC (I think it was 2.2MOhm), and after that had stable start.
Disconnecting the PLL jumper, I've discovered, that now speed can be adjusted, but it is till lower than 3000Hz on any position of the potentiometer. The saw waveform period on pin 4 was 1.6mSec, but according to dd33 schematics it should be 1.2mSec, and also voltage was 1.5V p-p, but according to schematics it should be 1.2V p-p.
To increase speed, I reduced value of R621 from 62kOhm to 33kOhm, and now I've had a possibility to set speed to 3090 as per datasheet. Then I've connected PLL jumper, and had locked it to 3000Hz 100% of the time. But only on 3V power supply. Reducing voltage to 2.6V brings back the erratic behavior.
Also, with disconnected PLL jumper I see almost 6.5V peak signal on IC602 (quartz lock chip) pin 5, which is connected through 1MOhm resistor to pin 4 of servo IC601.

What concerns me:
6.5V feedig to the IC601 could have damaged internal logic of the chip. As per service manual, inside the IC601, after sawtooth generator, the signal goes directly into the comparator, which I think compares it to Vref minus one diode voltage drop (which at 1uA current should give us roughly 1.2V).
Sawtooth waveform peak voltage is 1.5V in my case, but 1.2V in datasheet.
I've ordered another dead walkman for way too much money to try to repair mine. Hope servo IC is good in it, or at least I can take some measurements from it. Only the latest DD walkmans use the same servo IC, DD33, DD22 and FX100, which is sad.

I'll give an update when I'll have more information.

 

Fascinating!

It is exactly the same issue I have. Could it be that someone plugged a higher voltage power supply? I know that normally the issue is reverse polarity but perhaps this issue is caused by too high voltage.

Anyway, I got a bunch of parts but unfortunately the DD33 PCB on this lot did not have IC601 Someone had already removed it.

I tried with a 3.3Mohm and it worked as you described. Although with this value I could only lower to 2.8V until it stopped locking. Will try other values later and will leave feedback.

 

I like reading about these detailed diagnostic "chases" of strange problems / mysteries !

 

Before doing more tests with different resistor values on the pull-up, I started checking the signals on all the pins and comparing them when it is locked vs non-locked.

When it is locked, pins: 15, 14, 13, 12, 11, 9 all are stables and constant.
When they are not locked, the value fluctuates and have some positive and negative spikes.
It is very weird...

The symptom I have right now is when I play start it is not locked. If I press the capstan wheel to slow a little bit and let go, it syncs.

 

So, I have double checked the waveforms and on Pin9, I found it odd that I was seeing impulses even when in sync. When not locked, those impulses turned almost like a square waveform.
This is connected to the auto off, so I removed resistor R616 and now it always locks in sync. Also, waveform on Pin4 is smooth and stable, so I think there is something wrong with the auto off mechanism that is triggering the symptoms and perhaps IC601 is not faulty at all.