Sony WM-DC2 Rebuild - Stops During Rewind

Doing more of the same and expecting different results is not a good strategy. It's more than likely not a lubrication problem.

Regarding the flywheel top screw, I recommend adjusting that mechanically not by current consumption.
Adjustment goes the following: you remove the motor and adjust so that up down play is within 0.05-0.1mm.
You can use a feeler gauge or piece of paper (which is 0.1mm in thickness) or simply do it by feel. It should have a tiny bit of play you can feel by hand, but not much at all.
It's also to be noted this adjustment should not depend on the capstan ring, assuming the dimensions are close to original.

Rather than cleaning/re-lubing the mechanism, I would try ordering a different capstan ring. Something like this for example: https://www.aliexpress.com/item/1005003562728980.html
It's possible the capstan ring you have it way too thick, especially considering the FixYourAudio ones don't have the angle which originals ones had.

Try to figure out where that increased consumption is coming from, as 150mA is 50% extra.
Run it without a tape, see if stopping the takeup reel is what increases the consumption, etc.

 

Hello - and thanks yet again for the advice.

Apologies again for the delay in responding - I feel I'm somewhat losing my mind with this project! Here's what I've recently found:

- As already noted: the current consumption whilst in Play starts of around 150mA.
- With the tape removed from the mechanism, this drops to around 115mA in Play. To increase the current up to 150mA requires quite a firm grip on the takeup reel - more resistance than I would expect, given how freely the tape appears to wind by hand.
- With the motor removed from the unit and engaging Play, it only spins for a second before it automatically shuts off - but as far as I can tell from my multimeter during this brief period, the current reaches about 105mA during this time. Obviously this is a big difference from the ~10mA current draw at 1.6V supply voltage for the motor disconnected from the PCB.

I'm certain I had previously tried fresh batteries with no difference in behaviour - but I've now tried this again... with different results. I guess I was having a bad day the first time I tried it...

Comparing the input voltage to the PCB between battery sets:
- Voltage in Stop is 3.0V (new) vs 2.4V (old);
- Voltage in Play is 2.8V (new) vs 2.1V (old). The current draw was still high at 145mA in Play, with the new battery set.

And the problem with prematurely stopping in Wind is still mostly present: in Rewind, the tape gets almost to the end before it halts. In FF, it does reach the end entirely without issue. Flipping the tape (to eliminate the chance of a tight spot at one end only) shows no difference in behaviour.

I'm using one of the original units' capstan rings still - I've not replaced this with a new one.

 

The difference between cassette in and without cassette current draw is too high. You should see ~18% increase with cassette in. What you have is a 27% difference, which is almost double.
Also want to correct a mistake: the spec in the manual is without a cassette in. So you should have ~100mA without cassette and ~118mA with.

Takeup torque on these DDs should be around 30 g x cm, which is not a lot. You should be able to stop the takeup reel easily with just one finger. If you need two fingers to stop it, takeup torque is excessive indicating clutch not slipping enough.

Since the motor draws normal current taken out of the walkman (no load), next step is to identify why the consumption is increased when installed in the mechanism.

Regarding fresh vs used batteries, unit should be able to wind the tapes fully even with somewhat used batteries. If batteries were the issues, you would see the LED dimming a lot to the point of being completely off.
Same goes for the cassette itself, as long as it can be turned (not completely stuck), unit should be able to fully wind even 120min tapes without triggering auto-off.

 

Hello:

Thanks - though I think the original batteries are now beyond serviceable usage, as I'm seeing the LED dim. The following is all with the fresh set of batteries installed.


I've been digging in to see if I can find out more about the current consumption...

With the mechanism assembled, I see the following current draws with no tape inserted:
- Play: 113mA
- Rewind: 108mA
- FF: 126mA

With the two gears on the take-up spindle removed, I see:
- Play: 108mA
- Rewind: 108mA
- FF: 99mA

So, whilst the 108mA draw is obviously higher than the specification of 100mA, the big jump here is the draw in FF - where the only difference between the two tests is whether the take-up hub is being driven or not. (Given the zero difference between Play and Rewind modes when the gears are removed, I'd suggest there's not an inherently large load presented by the spool hubs.)

So I compared the feel of the two hubs - there was a bit of play in the supply hub, but no play at all in the take-up hub. Consequently I've pulled it apart and put it back together, such that there is a little play - I wonder if I was a bit over-eager in my tightening of the screw that holds the top take-up gear in place, and that over-tightened the components on either side of the spindle.

Since doing this, the behaviour has been a little better (though it does still fail to consistently rewind entirely to the end of the tape). The current draws I'm now seeing with no tape inserted (and the two gears on the take-up spindle reinstalled) are:
- Play: 105mA
- Rewind: 107mA
- FF: 119mA

Perhaps I need to slacken the take-up hub even further - but I thought I'd take a couple of videos, just in case something leaps out.

Here's the mechanism in FF mode, near the end of the tape - all looks good: (attached IMG_9601.mp4.zip)

Here's the mechanism in Rewind mode, near the end of the tape. You'll hear it briefly snatching as it goes, with little kicks in the motor position - and then the motor completely kicks off to one side when it finally stops (before the end of the tape). To me, this seems like a mechanical behaviour rather than an electronic one: (attached IMG_9600.mp4.zip)


What do you think? Thanks for your advice as always!

 

105-108mA I would consider within spec, which further tends to confirm something I suspected from the beginning: you likely have a faulty CX20084 and the problem is not triggered by increased current consumption.
And think the compromise solution would be to remove that R616 and stop the unit manually every time at the end of FF/REW instead on relying on auto-off (mechanical auto-stop on PLAY will still work).
Another option is try to experiment with a higher value, like instead of 47K, put a 470K and see if that fixes the problem.

Regarding the take-up hub, the problem is not tightening the screw too much, but the clearances between the parts themselves.
Solution to this is remove the hub and file a tiny bit of material from the plastic bushing top part. This, in turn, will give you the necessary clearance.
In case of takeup reel, the play is fixed and determined by length of the shaft and of plastic bushing, unlike the supply one where you can press the gear at different position to obtain the desired clearance.

 

I'm glad to report that the machine is now working correctly!

I tried this - didn't seem to help, so I gradually increased the resistance. By the time I got up to 570K, the unit would briefly pause (where it had previously completely stopped), before then continuing. An improvement, sure - but I still wasn't convinced by it.

Seeing the struggles the mechanism seemed to be having with the eccentricity of the centre gear, I took it apart to try and re-centre this. Although there doesn't seem to be much play in how the toothed rim fixes onto the central hub, I did manage to improve it somewhat (rewinding no longer kicks as it did in the previous video).
I also noted that there seemed to be some stiffness in a couple of the elements of the underlying mechanism (see the attached picture), so I took these out and cleaned out their undersides, which allowed them to move much more freely.

Since putting it back together, the problem appears to be resolved - even with the original 47K resistor back in place! So I'm not sure which part of this made the most difference - but feeding back here in case it helps anyone in the future.

@Valentin - so many thanks for your continued input on this. I'm so very grateful for your generous time and expertise!

 

Those 2 parts in the pictures are related to the mechanical auto-stop.
Fact those were harder to turn makes even more a point in fact mechanism chassis should really be removed and fully disassembled, cleaned and re-lubricated.
This way things like takeup reel support (which is a problem in ~70% of the cases) can also be adressed.

Hard to say what exactly fixed it, problems with eccentricity seen before (they're common with the injection molded gear, CNC one is much better) but never seen that to cause auto-off to trigger.
Think question is what is the current consumption difference before vs after ? Is there a significant difference ?

 

Before the last steps, I was seeing this:
- With tape inserted, current in Play was 145mA;
- With no tape, current in Play was 113mA;
- With no tape, current in Rewind was 108mA;
- With no tape, current in FF was 126mA.

Since the latest work, I'm now seeing:
- With tape inserted, current in Play is 130mA;
- With no tape, current in Play is 107mA;
- With no tape, current in Rewind is 106mA;
- With no tape, current in FF is 108mA.

So, curiously, despite Rewind mode being where I was seeing the erroneous auto-stop problem happening, that's actually the mode where the difference in current before/after is the least.