Dear forum members, I had purchased a Sony WM-EX655 which I was told was in much better condition than it actually was. As part of the repairs to get the walkman to work I have completed the following: - Replaced the rubber belt with "FixYourAudio" brand replacement for this walkman - Reattached one of the pichroller springs (the black one near the centre of unit) - I had to bend back the playback head brackets as it was causing the unit to not close properly (I do not know how the previous owner bent it) - Cleaned the playback head and pinch rollers - Cleans pots/switches with contact cleaner - Calibrated the playback speed using Digital Oscilloscope. It seems to fluxuate -0.025KHZ to +0.025KHZ out of range (or possibly even more). There is is also alot of amplitude variation and when I listen to the tone on the calibration tape (3.15kHZ) I can clearly hear this wobble as I can when listing to any tape My question would be, what should I try next in order to try eliminate this wobbling noise? The DC motor seems to spin freely so I havent taken off the magnet to clean as it seems ok. The slipping take up reel clutch shouldnt really be able to cause a wobble as far as I know unless it was not slipping consistantly at the correct torque? What do you think? I have not 'demagnetized the head' by the way as I have seen some debate on how useful this actually is to the playback quality. Im not sure if it could be something wrong with the pinch roller or capstan. I'll include some pictures. Any list of suggestions for me to try eliminate the wobble noise? I have also noticed when playing the B-side via auto reverse it plays the other side of tape but with no high end. I only hear low frequency end of spectrum. Im not sure if this is related to the wobble issue or not. Its become clear to me now that this high end cut off would most likely because my playbackhead isnt perfectly aligned. It would not be picking up all of track 3 and track for (left and right of b-side) completely. At least I'd assume this. What do you think? In the meanwhile I will try process the signal of each side in XY mode on oscilloscope to look for perfectly diagonal line on each side After playing around with it some more and bending back playback head I've noticed a repetitive cutting in and out sound about 500ms apart. On closer inspection I think theres a flat spot on the b-side pinch roller. That would explain it I think. Edit: actually the pinch roller seems okay so I dont know why its doing that. I'll clean capstan and pinch roller while in play mode to clean all sides with door sensor forced down tomorrow at work. I just cleaned thoroughly, still clunking. Anyone know how to remove the pinch rollers on this model? I was going to swap them around to see if it moves the fault to the other side
Demagnetizing the head won't solve the lack of HF, as it's a problem that exists only on one side so it's a tape path problem. Head azimuth is non-adjustable on these units, lack of HF can be caused by: head wear (unlikely), bent head bridge (you mentioned you straightened it) or tape curling at the roller (most likely). Assuming you don't have a mirror tape, try to observe the tape at the capstan/roller (remove the frame if necessary), see if it moves up or down after the roller engages. Swapping the pinch roller with other side is a good test, but keep in mind alignment problems (roller not parallel to capstan on vertical axis or at an angle on horizontal axis) can also cause the same problem. The high W&F may or may not be related to the tape path problem. Given the high W&F/speed variation exists on both sides, it would seem not related. Do you have a WFGUI screenshot of the measured W&F ? 25Hz speed variation is high. In regard to potential causes of W&F, on these series are: - cracked clutch gear: https://stereo2go.com/forums/threads/wm-ex-series-tac-tac-sound-repair.7108/ - wear on the motor bushing; fact motor is spinning freely doesn't tell much, you need to check lateral play; - wear on capstan bearings;
Hi Valentine. Thank you for the advice. For the pulsing noise problem on only the b-side auto reverse playback, I have taken the plastic frame off for a good inspection and I can clearly see the pinch roller is bopping back and forth during playback the with b-side capstan. I feel no play on this capstan pully so I think it may be the pinch roller. I have another EX655 so if I can lift up pinch roller bracket, I'll swap it with this other one I have to rule out faulty pinch roller. I'm not 100 certain the other ones pinch roller is healthy but it doesnt bop back and forth like that This is a video showing the bopping and if you turn on audio can hear the pulsing along with the tone on the test tape Update on pulsing noise issue: Switching over pinch roller (with bracket) from my other unit solved this so its definenly the roller. For W&F analysis (on common a-side and b-side playback), I've been using my Siglent Oscilloscope on normal waveform mode and also XY mode. Its not producing a linear line on XY mode so somethings not right. The gear does not seem cracked but I haved ordered a new one and when it arrives I'll be able to rule out the clutch gear (this is assuming the clutch internal mechenism is still ok as I only ordered the outer plastic gear teeth. On the motor it spins freely and I see no lateral play at all. Simmilarly I see no lateral play on the capstain flywheels. This confuses me to what the problem could be. Perhaps its the clutch Im going to spend some time learning WFGUI and then I'll upload some screenshots Update WF screenshots from WFGUI - A-Side - B-Side (auto reverse) Capstan 2 Deviation seems to be around around 20hz max on A-Side and 13hz max on B-Side. (Assuming I calculated it right. 3150hz * (deviation (0.006% max for A-Side, 0.004% max for B-Side))
Not sure I understand what you measured with the scope, are you reffering to azimuth ? XY 45 degree line means the channels are in phase. For azimuth you need full track 10kHz test tape. The 3.15kHz speed/W&F tape is no good for checking azmith. But as long as it sounds good, don't bother with azimuth measurements, as you will rarely get a perfect measurement on such autoreverse mechanism with no azimuth adjustment. As far as the W&F goes, figures are pretty typical of this mechanism. FWD side has higher W&F because it also drives the geartrain. The clutch gear, as long as it's not cracked, don't replace it. Things you can try to improve the W&F are: lubricating the capstan bearings, motor (have a tutorial for that, but keep in mind it's tricky) and eventually try a DeckTech belt instead of the FixYourAudio one. But at the end of day don't expect getting a 0.1% WRMS figure on this walkman, it just won't happen. Question is: do you hear W&F with music ? Fact you hear it with test tone doesn't mean much, as with a constant tone it will be audible even at very low values like 0.04%.
Sorry for the misunderstanding. I wasn't testing for azimuth anymore as when I bent the playback head back I no longer am hearing a noticle cut off on the high frequency end. I also do not have have a 10khz test tape unfortunently (but only the w&f 3.15khz calibration tape) Regarding the XY test I was conducting on the oscilloscope: I was not only really looking for any phase shift between the left and right channels also any ampliitude imbalance between the channels. - if there is a phase difference between the 2 channels it will show a tilted ellipse on the scope which points to a azimuth alignment issue. - if the channels are in phase but the is amplitude imbalance it will show a horisontal or vertical ellipse apparently I think I will leave the azimuth alone for now like you said as the auto-reverse side is playing nice now with the new roller. Early on when I was testing, I examined the triggered waveform (in normal voltage time mode) of the left and right channel of the headphone output signal when playing the 3.15khz test tape and i noticed alot of amplitude fluctuations. This would be the equivalent of tremolo in music lingo. Any idea what would have caused that? I need to retest after work to confirm if its still doing it I am a bit hesitant to pry out the thin PCB off the motor so for now I'll the other things. I would like to lubricate the capstan bearings. In order to do this, i first lifted off that plastic cap on the tape side of capstan and then i tried to gently pry off the capstan pulley from the belt side. It seems very tight. Any tips? What sort of tools did you use? The auto reverse side sounds quite good to me last night and I didn't notice W&F (when playing ordinary music). The Forward side I did notice some sort of fluctuation. It wasnt really bad, but noticeable. I sort of wonder if that pinch roller needs replacing also or perhaps its tremolo I'm hearing and not vibrato Side question: which portable cassette players do you think are the best in terms of quality, serviceability and longevity?
Regarding Lissajous curves, there are many patterns depending on phase and frequency difference between 2 signals (in case of test tape there is only phase difference): https://en.wikipedia.org/wiki/Lissajous_curve For azimuth, the amplitudes need to be maximum and also in phase. But a full-track (recorded with a mono head) tape is required for that, otherwise the recorded L/R signals may not be in phase to begin with. Check if head is parallel to the mechanism chassis visually (with door removed). If that's the case, it's ok. Some amplitude variation is normal at higher frequencies, it's analog tape after all. Not sure how much of a variation there is to tell if it's normal or not. The black caps on the capstans are just cosmetic and for the tape not catching in sharp edges, there is a split washer underneath that holds the actual capstan. Make sure to lube both bushings and clean the capstan thoroughly after otherwise your tapes will get contaminated with oil. You can try replacing the pinch roller on FWD side, although given that's what drives the geartrain, the higher W&F is more likely related to that as previously stated. Yes, depending on music (notes with long decay reveal W&F the most) 0.3% and even 0.2% will be noticeable. If W&F <0.2% is a must, walkmans from EX series to consider are the ones with large motor: EX7, EX9, EX20, EX2000. However do keep in mind none of these have the volume pot on the device, but on the remote. As far as the best walkman in terms of quality, serviceability and longevity, the above mentioned would meet these criteria if motor + capstan bearings are lubricated. AIWA Alpha-2 mechanism walkmans (https://stereo2go.com/forums/threads/aiwa-alpha-2-mechanism-grooved-head.9629/#post-76999) also tend to obtain better than average W&F figures, but these are not easy to service. And of course there are the SONY DDs, which will render excellent W&F, but again not the easier to service. But if you want only one walkman as a daily driver, I would prefer a brushless motor as brushed ones will eventually wear out with heavy use.
He playback head surface was not completely parallel with mechenism chassis and capstans. I've straightened it a little more. The first tape I put in got tangled up. Im not sure why that is lucky it was a tape I didnt really like. the subsequent tapes have been fine. I've made some videos on oscillscope playing the 3.15khz tone: Voltage / Time Forward Voltage / Time Reverse XY Forward XY Reverse My observation on the xy forward is that I see a near perfect circle and for the xy reverse I see a tilted ellipse. Based on what google gemini has regergitated to me this means the following: (I know gemini is not as accurate as a pro such as yourself but interesting to see what it says) Based on it's response, it sounds like my bench mark should be a straight diagonal line. So my forward playback is currently extreamly bad and reverse playback is a little better. I think I must have it bent back too far. I had another look and noticed it still wasnt perfectly parallel so I bent it some more. This has definently helped as I am now getting a tilted ellipse in a much less circular form. Or better way to put it, im getting closer to a perfectly linear line. XY Forward 2nd Adjustment XY Reverse 2nd Adjustment I'd say if I kept bending it it would eventually be close to a linear line (perfect azimuth) Thanks for explaining there is a slit washer under the capstain cosmetic cap, I didnt see it before. By the way that cap is not only cosmetic as I found it it was the reason my tape got twisted before I think. The black cap forces the tape on the correct side of the capstan. I noticed without it on when i put in the calibration tape the tape went to the wrong side of capstan. I couldnt work out why it wasnt playing and it was because tape film was on wrong side. I assume thats why the other tape earlier got twisted but not certain. Before I pull out capstan I'm going to order watch lubricant oil. Do you recommend Moebius Synt-A-Lube 9010 for walkman bushings? Thanks for all the suggestions of the good walkmans to look out for that give low w&f and good lengevity. I'll keep an eye out for them. I was thinking with the brush motors couldnt we just replace the brushes or are they near impossible to find as a replacement part? I replaced the brushes on my cars alternator about a decade ago and the process was pretty straight forward. But I guess alternator brushes are much eaiser to find and this specific one 'Toyota' was designed to be eaisly replaceable. That of course helps alot
Yes, as stated in previous post in phase --> 45 degree line, but: 1. You need a full track test tape (meaning recorded on mono head); 2. On such a unit you won't get a perfect measurement anyway; For this reason I would bend the bridge to look visually perfect. Any watch oil is fine, you don't need something special for this application. The brushes yes we could replace, but there are 2 problems: 1. The commutator also wears, which will also need to be sanded/replaced; 2. Neither the brushes nor the commutator are available to buy separately;
I've asked the equiptment company whom I bought my calibration from if it has been recorded as a full track tape on a mono head. I will let you know. Yes, I agree visual alignment seems adequate. I've placed an order for some pinch rollers and watch oil. I'll have to sit tight until it arrives Yeah it sounds like the brush motors are going to wear out faster than the brushless. I might have a go at servicing the motor bushing. My fear is that I'll accidently cut a trace on one of the coils. Worst case I guess I could repair the trace
Now in practice brushed motors will last a very long time. But if a walkman is used like 5h a day, every day, the motor will eventually fail after a number of years. Really depends if you have just one walkman which will be a workhorse or the wear will be split among 3 or 5 walkmans. As for the EX motor, there are a couple of tricky parts: 1. Removing the top magnet; you need to be careful not to pry the whole assembly (with bottom magnet) as that will break the PCB; it's probably the easiest mistake to make; 2. Removing the PCB from the motor case without damage; here heating the case with a soldering iron will soften the glue and help a lot; 3. Re-soldering the motor holder (plastic part with pins); this absolutely needs to be done while motor is firmly attached to the mechanism chassis, otherwise PCB will pull up once screws are installed; For 1, you need to be very careful to hold a finger on the pulley, otherwise there is a tendency for the whole rotor assembly to slide up.
Thanks Valentine. I'll update this thread when I have some more results after lubricating bearings. I might try service the motor on one of my less valuable walkmans first as it seems risky doing it first time
Hi Valentine. Today I run the walkman test cassette with 3.15khz tone on the walkman and run the wfgui_8_60.exe with input interface set to the walkman output. I adjusted the speed potentiometer on the forward playback so the 'frequency' field read about 3150hz +-20hz. When I go to auto-reverse playback the frequency goes up to 3230hz +-10hz Then after i stop and leave for a few seconds and try forward playback again it seems give me a random centre freuqency each time for the forward playback. Like the second stop play: 3170hz, then the third stop play: 3130hz. The autoreverse playback however seems to be consistant on 3230hz as the centre. Does this point to any particular fault in the device? Do you expect this behavour from this walkman or does it seem like theres a problem? I know forward playback needs to drive the larger gear train like you said, but even so this doesnt seem quite right? (I've not lubricated the capstans yet though still waiting on watch oil)
First of all, keep in mind you need to allow some time for the unit to stabilize (5-10 seconds) before reading the speed/W&F figure. In the first few seconds of running, there are some fluctuations in any unit that should be ignored. Then, fact the speed is consistent on REV, but not FWD can be an indication problem is mechanical in nature. Usually I would suspect a dirty speed trimpot in such situations, but fact REV doesn't have fluctuations would suggest otherwise. I would check the pinch roller pressure is there and consistent with every engagement, capstan surface is even and capstan is rotating freely even when roller is pressed against it. No, it doesn't seem right ~100Hz difference between FWD and REV is too high to be considered normal. Difference should be within 20-30Hz.
I replaced the roller and bracket with my other ex655. It seems to have made no difference. The forward and reverse playback is still about 100hz apart I've manually pushed pinch roller onto forward capstan and rotated the capstay pulley with my finger both ways and the contact with pinch roller looks good and i dont really feel any resistance. Im a bit confused as to what the problem is. There is a bit of mechnical noise on forward and reverse playback and both sound different but nothing New pinch rollers and watch oil still on the way. Should come in a couple of days I agree that it seems mechanical. Can't figure it out. Perhaps I dont have head perfectly aligned and edge of tape is rubbing against something
The pinch rollers I ordered arrived and they have slightly smaller outter diameter. It causes the roller to bob up and down, annoying. I ordered a generic Sony WM EX roller though so it was my fault. Do you recommend I order this one or do you have a recommended alternative I could try? I'm located in Australia
Replacing only the rubber tire I do NOT recommend, because almost always such roller assemblies become eccentric. Generally speaking, don't recommend replacing the original rollers unless it has a problem like indentation mark, it's not round, rubber is hard as a stone, etc. You can try the FixYourAudio ones, like this: https://fixyouraudio.com/product/sony-wm-ex560-pinch-roller/. They're not listed for each individual model, but are the same. Or if you want to search more options, the dimensions are 6 x 5 x 1.5mm (outer diameter x height x pin diameter). Note height is the total height, including the hub.
Thanks. I'll order from Fix your audio! Today I recieved watch oil so i removed the forward capstan, cleaned bushings with ipa and lubed them. Interestingly in the software W&F metre it seems to believe things have made things alot better. I recalibrated the speed and now the forward and reverse is about 30hz apart. There is also much less variation in pitch and it seems to start up fairly consistantly each time so a huge improvement! It seems to have been the problem Thank you very much mate. Some of my tapes like my very old personal recordings have a bit of W&F but it must be the tape stretched. I have one question: how do you usually remove the slit washer that holds the capstan on. I used a flat head screwdriver and bent it out but it doesnt seem like the best way as i bent it a little bit. It didnt come off nearly as eaisly as the clutch slit washer for me. Forward Reverse
30Hz difference can be considered in spec. Even the manual gives +/-1.5% REV speed differnece, which means 45Hz. 30Hz is just 1%. Yes, W&F can be caused by the cassette itself: if hubs are warped (and tape doesn't spool evently) or slip sheets are worn, you will have ecessive friction. It also depends A LOT on the music you're listining to. That 0.13% figure shouldn't be audible with any music, while the 0.22% will be noticeable with some genres (which have long note decay), but not so much with others. As for removing the washers, I usually press on one side with a fingernail, while pullin on the opposite side with a pair of tweezers. However it's hard to remove them without any bending, if it bends just bend it back before re-installation.
I just read the manual and found the part your refering to. Yep thats true so it seems like the variation between forward and reverse is all good now. I would say the W&F 0.22% is not desirable but not all that bad. I'll think I'll leave it for now. Thanks mate! Oh it sounds like I did the capstan slit washer removal as good as possible then. I was a bit concerned they bending may damage it. I eventually got it to click back in
