WM-3/TPS-L2/WM-3 motor disassembly and maintenance

[To anyone seriously considering performing this repair or maintenance, read through all of the comments on this thread as there is some extra advise and comments which will give more insight into risks and how the repair can most effectively be completed, thank you.]

After a month or so of building up the courage, I finally got around to maintaining the motor to my WM-3. The motor works just fine, however it was a bit noisy so i wanted to do something about that. This repair was done thanks to a video out there on YT by ‘Living Analog’ who im sure some of you have seen. He posted a video where he briefly displayed disassembly of the motor. For the sake of keeping things simple, i will not go step by step on exactly what he did as you can watch the video itself here:



Instead i will comment on my own experience to aid some of you who may want to do this in the future. This repair is applicable to all MNF-1600B motors as found in the TPS-L2, WM-3, WM-2, etc

1. In the video, he desolders the wires to the motor, in practice this is unnecessary and i found it easy to handle with the wires still soldered

2. When bending back the clips on the metal sleeve of the motor, use a pocket knife blade. is thick enough to push it back, and small enough to remain rigid and be easy to work with

3. You may have a hard time pushing the motor backwards and out of the metal sleeve, but this is normal. Use a flat head screwdriver to push it out, try not to pull it too much.

4. He makes it look remarkably easy to remove the motor shaft, when in fact it is not possible to do by hand. Instead i recommend prying it out of its socket with the tip of your tweezers (WARNING 1: there is always the risk that you could damage the plastic socket which holds the motor shaft in place) (WARNING 2: there are 3 small nearly invisible washers inbetween the spindle and the motor hub entry, make sure to retrieve them or ser them aside unless they are still on the spindle shaft.)

5. As shown in the video, an exacto knife is a must for getting the glue off the plastic housing to separate it from the metal contacts in the clear plastic ring.

6. In the video he shows the lubrication of the motor shaft and hub, but not of the contacts in the clear plastic bearing or the pole on the copper tube. Use alcohol on a qtip to clean the pole on the copper tube, and use a tooth pick to pick out any junk in the bearing in the clear plastic piece. Afterwards, apply drop of oil onto the pole on the copper piece.

7. For returning the motor back in it’s sleeve, it is easiest when the metal retainers are put back in postion so you can return the rubber ring on the spindle end back into it’s normal position

just as a side note, when the spindle shaft is taken out, there may be a small white ring around the end of the spindle shaft. It was decayed on mine and upon completion of the repair, wow and flutter was normal and i could see no evidence that this small piece was significant to the performance of the motor

 

That's a very interesting topic indeed ! Congratulations for doing it !

From what I can see in the video, the back bearing is impossible to oil without taking the rotor out of the casing, contrary to a previous suggestion on the forum to oil it with a very small needle though the inspection hole on the back (oil will never get there).
This kind of service will become necessary at some point, especially on older units with many hours of use. However, what I am afraid of is many people which have never opened a motor before will attempt this repair with potentially bad consequences. It's certainly not the easiest one to repair, unlike bigger deck-type motors.

I have discovered this channel (Living Analog) a couple of months ago and I'm really not impressed with the quality of his repairs, although he does get the job done. However, what is shown in the videos is far from what I would consider a professional repair;
He left the motor pulley with a lot of scuffs, not sanding it (this will both increase the wow&flutter and wear the new belt much faster), he always puts a lot of grease in points where oil should be used instead. Many times I hear a ton of wow&flutter in the test done at the end of the videos.
I'm writing this because I assume many of these units end up on ebay listed at 800$-1000$, which is far too much for a unit that has not been serviced professionally.

@Xjmczar Great description of the procedure, I really like it !

I have a couple of questions about the process:
1. In your opinion, what is the risk of damaging the brushes when disassemblying/reassemblying the motor ? In my opinion, that's always a potential problem, as those can easily get bent even on big deck-type motors.
2. What kind of result did you get after cleaning/relubricating the motor ? How much quieter did it get ?
3. What wow&flutter figure did you get ?

 

Thank you, im glad you found it informative! In regards to your questions:

1. The risk is high, however i actually found it easier to disassemble and reassemble than some carbon brushed motors in some decks ive worked on. The brushed are a little sturdier and more rigid making them easier to work with and manipulate your way around when reassembling. It was a nightmare nonetheless trying to reassemble, but nothing a lot of time and wiggling couldn’t fix.
2. So my motor previously was quite noisy, to the point where at high volumes, the oscillation was audible in playback. After the procedure the motor was much quieter and definitely seemed a lot less strained sounding.
3. There was not a significant decrease in the wow and flutter levels, its around 0.221 RMS, I believe still a tad higher than spec, but likely down to the fact the belt groove on my capstan is slightly uneven due to a hardened belt which melted itself onto the groove. I will say however than the speed playback is a lot more consistent. Before, I had problems with the speed being a bit slow or a bit fast, but so far it is no longer an issue.

Just some closing thoughts, i agree that not a lot of people should really be doing this, and in all reality i wouldnt even recommend more qualified people doing so just because of how risky it is. My biggest concern is over the spindle shaft as i discovered after prying it out, there were tiny groves which guided the bottom of it in and out of its socket located at the bottom of the copper tube. Prying it out is the only way to remove it, however there is no way to know whether it is in the groove, or being stripped in the process simply due to the amount of force you have to put in to it.

 

I think the wow&flutter figure you got is the typical one, all the TPS-L2/WM-3s I repaired are in that ballpark after a complete service: lubrication of most rotating parts, cleaning of belt grooves, new rubber (sometimes including a new pinch roller, although that usually does not make a difference unless it has indentation marks).
In regard to the motor what I typically do is lubricate the front bearing, clean/sand the pulley groove if there are stuck depositis and spray contact cleaner on the back hole.

The TPS-L2 article on walkman-archive states a 0.219% WRMS wow&flutter figure, so consistent with both your measurements and mine: http://www.walkman-archive.com/gadgets/walkman_sony_01_tps-l2_eng_v3.htm

Given the service manual does not actually give a specification for wow&flutter, I tend to believe that is the normal value (given in part by the servo circuit itself, which I think is not the best SONY designed).

 

The wow and flutter performance certainly shows its dictation DNA, but it’s tolerable. Suffice to say I don't particularly enjoy listening to classical music on it.

 

Exactly, your unit is now performing as it should have been when it left the factory over 40 years ago, remarkable! One can enjoy it for the next 10-20 years . As stated before in another thread, Sony only provided w&f specifications for a select range of portables: as far as I know, the ones mentioned in this thread and the D(D) line. So others might be ‘even worse’ but have been used my millions of people to enjoy music on the go.

 

@Xjmczar I've looked at the TCM-600 schematic and noticed the servo circuit is nearly identical to the one used in TPS-L2/WM-3 (uses the same CX183 IC and even the implementation is extremely similar). No surprise here, it's something that I expected to see.
What I have noticed about this circuit is it's pretty slow. After initial turn on, the tape speed is initially a tad slow and only after 10-15 seconds is gets to the speed set by the speed adjustment trimpot.
This suggests the bandwitdth of the servo circuit is not high, so it's not able to compensate for fast transients in tape speed variation, but only to constant variations in load (like a cassette where the hubs are a little harder to turn). That explains, in part, why the wow&flutter figure is on the high side.

The only downside of this mechanism is the auto-stop function. The spring that acts on the cam puts a significant additional load on the motor, which does generate a periodic wow, once every couple of seconds (when the cam pushes on the lever).

@Emiel The ~0.20% WRMS figure is in my experience something pretty typical of a belt driven walkman (at least for the early ones). Depending on the music, it's a figure that can go mostly not noticed or it may ruin the experience on some types of music.
Belt driven walkmans that I have measured were between 0.12%WRMS and 0.25% WRMS (most being between 0.15%WRMS and 0.19% WRMS). DD units without quartz lock go below 0.1% WRMS, while those with quartz lock get below 0.05% WRMS.

 

Thats very interesting @Valentin as that would certainly explain the speed variations. Though for however much load the auto stop mech puts on the motor, it seems relatively negligible since you can't hear it that well during playback.

Anyways I wanted to share some more information since I did run into issues since this maintenance was performed:
1. When the motor spun up, the motor would make a dreadful squeaking noise, eventually dissipating within 5-10 seconds.
2. Motor noise got a bit louder
3. Wow and flutter increased

Inspecting the motor before taking it apart again, I noticed the squeaking sound was caused by the position of the motor spindle. As in if it was further out or in, in this case the spindle being closer to the chassis seamed to be the position which caused it. So if the motor spindle was nudged out a little, the squeaking would stop.

Going back into the motor I had 2 objectives:
1. To reduce the noise of the motor (wow and flutter as well)
2. To find the cause of the squeaking

I tackled the squeaking noise first, and as it turned out, the squeaking was caused by how far in place the motor spindle was in the socket of the copper tube. Now at first I was confused as the spindle should just be able to sit in with minimal play. But then I discovered that the missing white plastic washer on the bottom of the motor spindle was to blame. When I first took it apart, my washer was cracked and crumbled apart. Reassembling it without the washer meant that the motor spindle sat too far in, or at least that the housing the spindle was grinding up against the socket in the copper tube. The solution then was to make sure not to force the spindle deeper than it should be, accounting for the space the plastic ring would of made.

However I then encountered another issue which was that the copper tube seemed to be brushing against the inside of the motor casing. The solution to this was that the plastic tape around the copper was bubbled up and was falling apart from me handling it, so I promptly removed it. It does not need to be replaced since the TPS-L2 lacks this tape so the WM-3 should function as normal.

The final issue then was the motor noise. At first I believed I didn't put enough oil on the back bearing, but this wasn't the case. The extra oil I put on did not have an effect on the noise.
So then I turned my attention to the rod at the bottom of the copper tube which sat in the back bearing. As it turned out, the copper tube did not sit straight, meaning the tube wobbled in motion. I found that using a flat head screwdriver I was able to manipulate the tube to sit straight.

Putting the whole thing back together and testing it for an hour, everything is working the way it should. Motor noise is down, Wow and Flutter is normal again, and the squeaking noise is gone (At least I hope)

I'd also like to make a revision to the warning I put out about possibly stripping the socket in which the spindle shaft sat. Upon further investigation this is actually a non issue, it is very easy to tell whether or not the spindle is catching in the socket when reassembling. All you have to do is turn the spindle while you lower it down and then feel for when it catches, that's when you can safely push the spindle back into place.

But I'd also like to give another warning. I didn't see this last time, but at the bottom of the spindle shaft housing is another clear washer which sits between the spindle socket and itself. Most cases this washer will stick to the spindle shaft housing bc of the oil, but remove if necessary.

In conclusion that white ring around the bottom of the spindle shaft was important, but reassembly and function is not completely compromised if it is broken. It will require you however to put more work into making the spindle sit right.

 

Very useful information about the motor, @Xjmczar !

This is a very good reference for repairing the MNF-1600B motor. Some pictures would also be helpful, so if you happen to repair a second one in the future, I suggest taking some pictures as well in the process.

In regard to the initial speed variation that I was talking about, the problem is caused by the servo itslef, as no motor squeaking can be heard during this time. The difference is not high (something like 10-15Hz lower), but noticeable.
I can hear the pitch of the music is a tad slow at the beginning and gradually coming up to the correct speed.
It's an aspect that I have seen both on units on which the motor was completely silent (you could only hear it if you stick your ear to the unit) and on units with somewhat noisier motor (which can be heard without sticking your ear to the unit).
NOTE: have not seen this issue on other walkmans that use the MNF-1600B motor, but use a different servo IC.

I mention this because, from what you say, it seems there are 2 independent problems that both cause similar symptoms: the one that I mention is a limitation of the circuit design itself and the one you are mentioning is a motor in need of repair.

By the way, how silent is your motor after the repair ? Can you hear it when the unit sitting on a desk and you on a chair in front of the desk ?

 

It is very quiet. Without a tape inside you can barely hear it, with a tape inside, the sound of the tape completely drowns out the minimal noise the motor makes. And yes if I should take a similar motor apart in the future I shall take some pictures. I only refrained from doing so since I didn't think I could offer too much more visual insight than what the video already showed with my phone. I would've if I had my proper macro camera with me, however I've lent it to a friend for a bit so I don't have it with me.

Also I forgot to mention in my last comment, but I also have to revise what I said about the risk of damaging the brushes. There isn't actually that big of a risk of damage than I initially thought. When I had all the pieces apart, I took the copper tube and mounted it in the brushes to test the strength and difficulty of the operation. To my surprise it is actually very easy to mount without damage. All you have to do is push it in dead straight. I did it a couple times just to see if any damage was done but no, the brushes were straight and still in the same place as before.

In my opinion, the riskiest part of the whole operation is the removal and reassembly of the spindle shaft. There are so many variables like broken plastic rings which could drastically alter the direction and success of the repair.

But thankfully, lubrication of the brushes can be done without removing the spindle shaft, copper tube, and the plastic spindle shaft housing from each other. Living Analog's video would suggest that the spindle should be removed before removing the back piece from the upper body, at least these are the directions I took with my repair. In reality I would suggest not touching the spindle housing at all since the necessary lubrication can be done from the front. And as I found out, many of the noise and performance issues have to do with those brushes.

 

@Valentin, here are some photos with some comments which will be useful for the repair and just general knowledge:

Also see R1/R2, these photos show how one can make sure the copper tube is perpendicular with the spindle.

R1) insert the spindle into the socket.

R2) Look vertically down on spindle and copper tube, making sure the spindle is not leaning. If it is leaning, hold the base of the copper tube and using multiple fingers, bend the copper tube in the direction of the lean. This process is trying to avoid the copper tube from rubbing on the inside of the spindle housing which causes wow/flutter and noise.

 

Very informative @Xjmczar ! This thread should become sticky, as so many SONY walkmans use this motor or a similar one without the FG coil (which I bet is very similar on the inside) on models where the FG coil is on the flywheel side.

I am surprised to see that there's no significant amount of wear both on the brushes/commutator and on the shaft (where it comes into contact with the bushings). This means this motor will last a very very long time and when it becomes noisy, simply service it and it's like new.

 

Very nice tutorial & photos! These will certainly help someone. When I tried disassembling my first one, it was rather scary since I don't know what to expect

I think another risk that should be considered is the possibility of the FG coil separating from plastic brush housing. The fit is loose, but tight enough for things to get stuck if there is any corrosion in the way. Improper handling may also separate them both when disassembled.

In any case, that's certainly something you don't want to happen. The wires of the coil are really thin.

Maybe it depends on the climate, but corrosion on the FG coil is something I found more than once here, in the relatively humid tropics.


And here's the insides of a similar motor (without the FG coil) from the National/Panasonic RQ-SJ1/KJ1/WJ1



This has another problem. The magnet inside has started to chip by itself, resulting in chunks interfering with the rotation of the coil. Note that the coil in these doesn't look a "nice" as the Sony ones.

 

Hello - I am new here, having just found this site and thread in trying to bring my TPS-l2 back to life. I am the original owner, and it has been stored and not powered on for 15 years or so. I purchased and installed new belts, it powers on as I can see the red indicator lamp, but there is no movement of the motor at all. My question is, are there any steps I can take to determine if the motor is the problem? I have time and lots of patience and would feel comfortable disassembling the motor, I just want to make sure that the motor is the issue before I proceed.

 

First of all, try to rotate the motor by hand and see if it turns freely.
If it does, you can test it using a bench power supply or even a AA battery: red and white wires are the motor itslef, violet ones are the FG coil (which you leave disconnected).

 

The motor spindle does turn easily, I have not disconnected the violet wires, should I disconnect them before testing or can they stay as is?

 

Okay, so I went ahead and disconnected both violet wires then connected to a battery, pushed the play button, nothing at all from the motor. I don't have a great deal of experience, but this seems to indicate to me that it's time to refurb the motor. Please let me know if I'm missing something. Thanks!

 

You connected what to a battery ? It's completely unclear what you did exactly.
What I stated is the white and red wires are the power to the motor, while the 2 purple ones are FG coil. Hope you didn't power the FG coil with a battery...
So what should be done is disconnect all 4 wires from the circuit and power the motor (using the red and white wires, NOT the purple ones) using a AA battery (just one - 1.5V).
If the motor spins when powered via the 2 red and white wires, means it's good.

The purple wires are from a coil that reads back the speed of the motor for the servo feedback loop. When testing, those can be left floating.

 

I'm sorry I was not clear, I just was not sure if the violet wires had to be disconnected from the circuit board. I unsoldered the violet from the board, left them floating, then connected red & white from the motor to the battery. So now I am ready to follow the procedures here on this page and pull the motor apart.

 

What confused me was this, as you stated you pushed the PLAY button, given you were supposed to test the motor out of the circuit, so the PLAY button doesn't have any role.

Then, given the motor doesn't spin when powered externally, but it does turn freely means it's not a lubrication problem.
You either have damage to the brushes/commutator or there's an open circuit on one or more the rotor coils.
Let us know what you find and if the motor can be fixed/how you fixed it.

For reference, the violet wires can remain connected or not it doesn't really matter.