Once again - this level of information is exactly what I was looking for, and your explanations are as clear as it gets. Thanks Valentin. Regarding the dummy load - like you said, moving forward I'll be using the oscilloscope for most of the same tests that I was previously attempting w/ my USB interface, so it sounds like in that case the 47k dummy load makes sense to use, as the high input impedance of the oscilloscope and/or multimeter is such that it will have a negligible effect on load. As for setting up the line in / mic in input impedance - Based on your link above about impedance bridging in audio, we're looking to maximize the load impedance and minimize the source impedance. So for line input calibration, say we are using the 600 ohm resistor in parallel to create a constant output impedance for our signal generator / attenuator. 1) With the line input impedance of the D3 at 47k input impedance, is it the case that the D3 represents the load, and the signal generator / attenuator the source? Now for mic in calibration - it looks like the manual states that the mic in wants a 'low impedance mic' on the input, but then it also looks like the mic input itself is low impedance (300ohms). So I'm a little confused - 2) Is it the input impedance of the mic input, or the source impedance of the hypothetical mic being plugged in that's the low impedance part of the chain? 3) Also wondering what effect the 10kohm resistor in series is meant to have in conjunction with the 600ohm parallel resistor? I was thinking maybe the 10kohm resistor in series cancels out the 600ohm resistor in parallel, creating a a 10kohm load on the mic input? And how does the 300ohm input impedance factor into this (if at all)? I wish I could return the favor for all of this help, Valentin - thank you again! John
I'm glad you asked the questions, John, because other people will find the answers useful too. I made this tutorial (more are on the way), because these devices need a complete repair to be in factory specifications, which most people don't do. I have seen many ebay auctions where replacing a belt, without even cleaning the pulley's grooves is called a refurbish, which in reality it's far from it. In the DD units, including the WM-D3, most people just replace the center gear, put some new lubrication and call it a day. Some don't even use a good quality gear, as there are more versions available on the market. Some units require a complete relubrication of every rotating part (including the motor) to bring the wow&flutter into spec for example. Not to mention the electronic adjustments, which are critical on Dolby equipped units. Very few people actually bother to even check these, although I did not find many decks. recorders or walkmans to be perfectly adjusted, especially given most were never re-adjusted. Some were not adjusted correctly from the factory, some have been dropped and the head wear alaso plays a significant role. 1. Yes, the D3 represents the load and the signal generator the source. Regarding the second question, I made a mistake, the input impedance of the MIC IN is not 300 ohms, it's higher than that. 300 ohms is the impedance of the source they expect you to put in. 2. No, it relates to the mistake I made. I will edit my previous post to correct it. It is exactly as in LINE IN, the MIC IN is the high impedance part of the chain (has a few kiloohms input impedance, not stated in the manual) and the mic itself is the low impedance part (300 ohms). 3. Let me explain more clearly: - the 600 ohm resistors in parallel acts as a constant load for the attenuator (it's related to the source part of the chain, hence low impedance) - the 10K ohm resistor put in series act so the load "sees" a high impedance connected to it (although the source impedance itself is not high, 600 ohm). So the 10K and 600 ohm do not cancel each other out, one acts for the source, the other for the load. I hope I explained it in a clearer way now. Regarding your question if the 10K creates a load on the mic input: note that the term load only applies to a SOURCE part of a chain. Another note regarding the first question is the input impadance of the LINE IN is roughly 10K (it's given mostly by the resistance of the REC LEVEL pot, RV301 on the schematic), not 47K as I stated (it is not given in the service manual). I edited the previous post to correct this. But, as I said in that post, I point out that the 10K in series with the source is still not needed.
Had a few mechanical snags during the D3 calibration process, which I’m happy to report are finally resolved. I’ll report back shortly on those and how I approached fixing them in case it’s helpful. Otherwise the actual calibration has been successful, with just the final bias calibration left to do. Even better has been the learning process throughout, and I have @Valentin to thank for that once again. I’ve seen the same few people post (usually in other forums) about how the only way to work on these units is after hundreds of hours of experience with some hundred odd decks, or else send out to a professional. My thinking is that not all of us, for better or worse, have gone into audio repair as a full time career, but we’re keen to learn nonetheless. For those of us who aren’t professionals, I’d credit threads like this one as just about the best resource publicly available. Hats off to all here taking the time and effort to share what they know!
Hi again, Back to close the loop on my WM-D3 calibration. All calibration procedures were completed, including tape speed, azimuth, playback level, record level, and bias adjustments. A couple of challenges along the are called out below - Azimuth Stability While adjusting the azimuth I noticed that my readings were highly unstable - I'd remove the ceramic trimmer tool from the azimuth adjustment screw, stop the test tape, and after playing again to check my work the reading would shift back out of phase. It was also the case that, while making the adjustments, if I put pressure in the wrong spot the azimuth reading would jump significantly. After comparing with another D3 I have on hand, I saw that the one I'd been working on was missing a spring that connects the head plate to the cassette door. Perhaps there was a revision to this section, or maybe it'd been removed by someone over the years, but regardless I ended up swapping in the spring from my other unit, which fixed the issue. Azimuth now remains stable. Low Record Level Another issue that I ran into - and that I think is alluded to in the initial calibration tutorial - was that I wasn't able to get my record level set high enough that the monitored output while recording matched exactly the level when playing the recording back. The adjustment pot that controls record level would bottom out - or seemingly close off completely - at a point that sat below the level required. I've read here and elsewhere that Sony may have had more sensitive tapes in-house than the tapes others were using with their machines, which is one potential explanation. I'm assuming that @Valentin 's recommendation to use a Dolby test tape for higher levels is because of this as well, but I might be misunderstanding that piece. Anyway, I simply adjusted record level as high as it would go, which ended up being only 1 or so dB away from matching the level monitored during recording. Misaligned Center + FFW Gear Finally, despite attempts at adjusting the center gear arm, I kept running into an issue that caused the FFW gear to slip out of contact with the center gear, producing a screeching sound and stopping the take up reel from moving. Lots of time spent inspecting the area, the motor housing (which I ended up swapping entirely with one from another D3) and all sorts of other potential solutions that ultimately failed to correct the issue. In the end I could see that my new center gear was 'wobbling' - for lack of a better term - and that there was a point in its rotation where it wasn't making a sufficient connection with the FFW gear. In the end, I glued a tiny piece of rubber to one of the center gear arms sitting below the FFW button, which added just enough height to force the center gear lower when pressed down, since it now made contact sooner with the FFW button. This allowed the center gear to maintain contact with the FFW gear during FFW. I can't imagine that's a clear explanation for most - though I've added an image here showing the small black rubber piece glued to the arm. As mentioned, I actually have a couple more D3's, each of which have their own idiosyncrasies that I'll be working through. That being said, happy have one here that's fully calibrated and sounding great.
Good work, @JoFugd ! Regarding the Dolby level tape, I recommend that one because I got consistent results using it and all decks, recorders and walkmans that I aligned using that tape played Dolby encoded tapes perfectly, even the pre-recorded ones. People on other forums stated that the test tapes recorded to the DIN standard, like the P-4-L300 were not recorded correctly. Given that all SONY units I have aligned had their Dolby levels off (and Dolby encoded tapes did sound muffled before the adjustment), I tend to believe that is true. A third reason is regarding the age of those old tapes: - I would not trust a tape that is so old to have the exact levels it had when it was first recorded - those old tapes are very hard to find nowadays and extremely expensive even if you do find them In respect with the low record level, the fact that you need to turn the trimpots all the way to get the level where it needs to be clearly indicates a problem. It should not be that way for sure. On all units I have repaired, I only rotated the trimpots a little. I would check the following: - Are you sure the playback levels are set correctly ? If they're too low, that could explain the situation. What test tape did you use to adjust those ? - Make sure 100% that the azimuth is correctly set. When using a 6.3Khz test tape, the L and R should be 90 degress out of phase and at roughly the same levels. - Are you 100% sure the bias is set correctly ? The bias level does influence the record level. What tape did you use to adjust the bias ? I recommend using a TYPE II (cobalt doped ferric) tape to adjust the bias, because it has a good enough treble response. On a TYPE I you will need to make the the treble response much higher to compensate for the response of the tape. And if the 10Khz response is too high, the 330Hz will be too low. Also keep in mind that TYPE Is are the most sensitive to bias levels, compared to TYPE II and TYPE IV. In reality you need to make a compromise when making these adjustments, as you won't get perfect response on all types of tapes. I personally consider that using a TYPE II is the best compromise. As a sidenote, I have seen some decks which have separate bias and rec level adjustments for each tape type. Only those will get perfect response on all tape types, but that is rare and certainly not to be found on a walkman.
I have some updates regarding the WM-D3: 2 problems I discovered in a recently repaired unit and their respective solutions. RECORDING PROBLEM @JoFugd I suggest checking out this one, as it seems to be similar to what you have described When doing the adjustements (Dolby level, rec level and bias) I noticed the record level on left channel doesn't get nowehere near what it should be, even with the trimpot at the maximum setting. I immediately suspected the bias, so I checked the bias level immediately after the capacitive divider. After some measurement and soldering/desoldering pads I found out the bias level is much lower on the left channel and I could not get it anywhere near what was on the right channel, even with all the capacitors soldered. The values were 5.3Vrms on the right channel, 3.1Vrms on the right channel. At this point I suspected that something was loading the bias oscillator, so I decided to desolder the wire going to the REC/PB switch. After desoldering, the bias level immediately came back to 5.3Vrms, identical to the right channel. The problem was obviously after that wire. I started by measuring the resistance between the hot point of that wire and the ground point (postitive side of C306). The resistance measurement revealed nothing, as both L and R resistances were the same. Decided to start measuring the indutance with the LCR meter: the test revealed a significantly lower inductance on the left channel (28mH on the left compared to 109mH on the right - test done at 10Khz). At this point, I was starting to suspect the head itslef, so decided to measure it out of the circuit. After desoldering the wires going to the head, the head measured just fine, so it was clear that's not the problem. It was only now when I started to suspect a parasitic capacitance loading down the bias oscillator (that would also explain the lower inductance measured with the LCR meter). With the wires to the head disconnected, I measured the capacitance of the circuit: 6pF on the right channel and 70pF on the left. Now it was very clear this was the problem. But I still did not know where that parasitic capacitance was coming from. At first, I suspected some oxidation inside the REC/PB switch; I was ready to desolder, open and clean it, but decided to do a clean around it first. Some residue came out from the REC/PB switch area (that wasn't very visible at first glance). Now the parasistic capacitance was gone. VU-METER PROBLEM The VU-meter indicates lower peaks than the real signal peaks. I observed this by playing the same tape on a WM-D6C. The D6C was peaking at +3dB, as the WM-D3 was not going higher than -5dB, sometimes barely ligjhting up the 0dB led (Dolby levels set correctly). The funny part was it seemed more like a slew rate problem, as the VU-meter would correctly indicate levels of constant amplitude signals, but would not respond to transient peaks fast enough to display them. I could not find anything wrong: did clean and resolder some of the components around the CX10043 IC and the IC itself with no result. I was hoping there was a similar problem with a parasitic capacitance in the feedback of the amplifiers that would slow them down. In the end, I increased the feedback resistance R336. Initially tried 220k, which increased the gain too much; then I tried with 180k, which seems about right but it's still too high. EDIT: Correct value for R336 is 165K. Package is 0805. Note that 165K is not a standard E24 value (5% tolerance), but it is an E96 value (1% tolerance). So you will only find this value with tolerances of 1% or less.
Hello, thank you for all this information, i also looked here for a Vu-related problem, I had static discharges/rumbling sounds, like thunder at irregular intervals. But the most apparent error was the VU-meter, at headphone volume "0" correct level with a Dolby level tape, but the led level raised to +5 (max) if i increased the headphone volume. The internal record level pots had to be turned fully up to have a correct recording. Reason was a interrupted ground connection, most probably due to a shock. Resistance between the ground point / screw hole and negative battery connection was 40 ohms, voltage drop 0.7V Repaired it with the yellow wire.
This tutorial was very helpful. Thanks! The question that remains is: Without a dolby test tape is it possible to check the playback level? I have the original ABEX tapes for azimuth 8khz and w&f 3150hz, both are -10db. I've already tested the R and L channels and they have the same value. But when I record with my D3 I get a lower volume, not much but noticeable. Not knowing if the playback levels are right, I don't dare to change the recording levels
Glad you found the tutorial useful ! From my experience the playback levels set from the factory are off. This is what I have noticed by both listening tests with pre-recorded Dolby-encoded tapes and measurement with Dolby test tape (200nWb/m ANSI). An imbalance between right and left channels is something I also noticed. Most SONY devices have the levels off, it's not just the WM-D3. I have seen very few SONYs with the correct levels (some D6Cs) and I suspect those that have correct levels have been adjusted before, although there's no way to know for sure. I am confident the Dolby test tape I use has the correct levels because if I measure a unit which doesn't have adjustments (like the WM-DD33), the levels are absolutely perfect ! Those units without adjustments happen to be the ones which sound good when playing Dolby-encoded tapes without doing anything on them. This is the reason I recommend using 200nWb/m ANSI Dolby test tape, not the 160nWb/m DIN test tape which was originally used. I would also not trust old calibration cassettes from the 80s, only fresh ones recorded recently. For testing PB levels you need a Dolby level tape, the azimuth and w&f ones aren't useful for this purpose. The one I recommend is the one mentioned in first post. In regard to recording, the levels as well as the bias adjustment will depend on the specific tape the unit is calibrated on. Do keep in mind magnetic properties of tapes can vary a lot, even between different generations of the same tape, so there's no perfect adjustment for all tapes. The fact that recorded levels are lower than the input will affect Dolby encoding rendering muffled sound, which will be even worse if PB levels are lower than should be. Do keep in mind that you will need to adjust both bias and rec levels. They influence each other so you will need to go back and forth a couple of times for a perfect result. As a conclusion, it's up to everyone what they choose to do. Playback levels are more of a problem and those should be done in my opinion, while record levels and bias can offer decent results as they are from factory. If one wants perfect recording for a specific tape, then doing an adjustment with that one is the thing to do. Then, if one wants to record with Dolby NR, again doing the rec calibration will improve results significantly.
I just read all these pages, Wow. Valentin, Do you work on walkmans for other people, yes/no? I am not ready to do so yet but I would like to keep people that I have read about in my thoughts when I do get around thinking about getting my walkmans repaired so I can play them. This month of February 2024 is a one year anniversary of me still not finishing repairing a boombox just to give a idea as to how far backed up I am, because other people wanted something else of mine right then...
@Cassette2go Yes I do work on walkmans for other people, as @Emiel has already mentioned. Hope you will be able to finish the boombox repair project and then get to some walkman repairs also.
I wish I could get finished but it may take me more months or even a year or two, before I can be only working on my walkmans. I may get side tracked doing a walkman here or there but my primary goal is to be rid of the bulk of the huge boombox amount of box's I bought to hear to understand how they sound but not to have a collection of them, this is where people confuse me with being a hoarder, I am not I just bought too much too fast and now I tonight I just got a offer on a model I am considering to get to repair...
