This is a complete step-by-step tutorial for how to adjust the Dolby levels, record levels and adjust the bias on your WM-D6 or WM-D6C. Applies to all generations and all PCB revisions. This tutorial starts on the assumption that the head is not worn or pitted and the azimuth is correctly adjusted. It also excludes electronic problems, such as bad EQ modules. Those could also cause tapes to sound muffled when played, but can also cause the unit to record tapes incorrectly. There are 2 EQ modules, one for playback, one for record. This is a common problem which should be fixed on any D6C. New EQ modules can be found here: https://www.ebay.com/itm/263537248339 Regarding the rec level and bias adjustment, some people like to do the adjustment to one specific tape and only use the unit to record on that type of tape. While this is the ideal scenario, I imagine that in reality most people will want to use the walkman with different types of tapes. I personally like to do that, because I also collect blank tapes. The good news is: if you align the machine using a good cobalt-doped ferric (TYPE II), it will work fine on most tapes (even on TYPE I and TYPE IV) without significant differences. The only big difference are true chrome tapes (like BASF Chromdioxid series). These have very different magnetic properties and are best recorded on a high-end will external calibration pots anyway. In the pictures below is the test rig I have built myself to load the output with 47K. It is very simple, using components that can easily be found.
DOLBY LEVEL ADJUSTMENT - You will need either a digital oscillscope (an analog one with digital measurements will also work) or TrueRMS multimeter. - You will need a Dolby level tape (200nWb/m ANSI). I don't recommend using a P-4-L300 tape (160nWb/m DIN). - You will need to load the LINE OUT with 47K resistor. If you don't, the levels will be higher than they need to be. The level you should be getting is 340mV RMS with a Dolby tape 200nWb/m ANSI. The VU-meter will indicate -5dB when the level is adjusted correctly. You need to adjust 2 trimpots: RV-101 LEFT CHANNEL RV-201 RIGHT CHANNEL Stop the unit and start it again a few times and check the level remains the same.
RECORD LEVEL ADJUSTMENT - You will need a function generator, digital audio player or computer. If using a DAP or computer sound card, it is necessary to use an uncompressed wav file, as mp3 will not have a constant level. - You will also need the oscilloscope or multimeter used previously. - You will need a blank tape of your own choosing. But it needs to be one that has not been recorded before. - You will also need to load the LINE OUT with 2 47K resistors. Feed a 330Hz 0dB (775mV RMS) signal to LINE IN. Put the unit in record mode and measure the value on LINE OUT. Write it down on a piece of paper. Now rewind the tape and play it back. Check that the value in playback is the same as the value you wrote down. If not, adjust accordingly. You will need to repeat this procedure a couple of times to get it right. It doesn't need to be perfect, a few mV RMS difference is ok. The tripots that need to be adjusted are: RV-102 LEFT CHANNEL RV-202 RIGHT CHANNEL Note that the REC level trimpots should be rotated counterclockwise to increase the level. The idea is to match the level that you get when monitoring the recording the same as the one you get on playback.
BIAS ADJUSTMENT - You will need a function generator, digital audio player or computer. If using a DAP or computer sound card, it is necessary to use an uncompressed wav file, as mp3 will not have a constant level. - You will also need the oscilloscope or multimeter used previously. - You will need a blank tape of your own choosing. Needs to be the same as the one you used for setting the record level. Feed a 10Khz -10dB (250mV RMS) signal to LINE IN. Put the unit in record mode and measure the value on LINE OUT. Write it down o a piece of paper. Rewind the tape and play it back. Check that the value in playback is the same as the one you wrote down. If the value in playback is higher, you need to increase the capacitance. You can bridge more than one pad if necessary. If the value in playback is lower, you need to decrease the capacitance. You can unsolder more than one pad if necessary. Of course, repeat the process a couple of times until you get it close. Given there are only discrete values of capacitors, you won't be able to get a perfect result. In order to have a good dynamic range, it's better that the output at 10Khz to be a litttle lower than the input, not a little higher. That's because the tape's SOL is the biggest limitation.
Yes, very good tapes which I recommend. To be noted HPR sells both 200nWb/m DIN and 218nWb/m DIN (positions 5 and 6 on the list). The Dolby level standard (which is what I refer to in this tutorial) is 200nWb/m ANSI = 218nWb/m DIN.
In the case of playback frequency response, using HPR test tape, what is recommended if the PB response is not flat? I know there are no PB response trimmers; perhaps altering the value of one of the PB amp caps (C11, C116) would flatten it?
On a D6C with head in perfect condition, playback frequency response will be flat (see pic attached). If there are dips or an early rolloff in the FR, head is the first thing that should be checked (wear and azimuth adjustment). Tweaking the preamp shouldn't be necessary if the head is original. When the response is not as good as it should, I proceed to do a head lapping.
I guess "flat" response is a relative term. Your graph shows ~-5db @ 18 KHz. Is that normal for a WM-D6C? Is your graph showing overall record/playback response, or just playback?
The service manual specification is 15kHz, which the measurement corresponds with having -3dB@15kHz. FR is playback. For recording, FR will only be flat on the tape unit was calibrated on. By "flat" I was reffering to FR being flat in the area where is should be flat (with no dips or early rolloff), not to a completely flat 20Hz-20kHz. For that to happen, we would need to have the -3dB point above 20kHz (which only a handful of decks have). On analog tape machines, some rolloff should be expected on both ends due to limitations of the head and tape itself.
Yes, you are correct...it IS within spec; perhaps even a bit better, being essentially +0, -3 db from 40 to 15 KHz. BUT a 1000zxl will do record/playback on Type IV 20-20kHz +/-.75dB. I don't expect a WM-D6C to match this; I'm just wondering (out loud) if it could be made BETTER by a cap change. I am also considering subbing in two OPA 1642s for the stock NJM4558s.
Yes there are decks out there which can get close to a flat 20Hz-20kHz on a TYPE IV, but it's not something I would expect in general, even less so of a walkman. The D6C can likely be made better in that regard (but it's to be noted noise floor will increase), will analyze the circuit an create a separate thread about that. About the OPA1642, I have tried an OPA (likely this one) in the past on the D6C that didn't work at all (possibly due to the JFET input stage). So can only suggest to try it and come with feedback later on. What does work on the D6C as an upgrade is the NE5532, but that one has higher quiescent current than the original. It's still within the capability of DC-DC converter, but it will deplete batteries a bit faster.
Well, that's very interesting. WHAT didn't work at all? How did it sound worse? Did it oscillate? I think that the OPA1642, being a faster chip, could possibly oscillate. It needs a 100nF cap across the rails and 22 pF caps across the feedback paths to limit its bandwidth to eliminate oscillation. Bipolar or film caps on its input/outputs would avoid any anomalies caused by switching to JFET from bipolar. The 5532 has always been a favorite of mine in my other gear; maybe I'll just go that way if you say that it definitely works. In other applications, I would just put a high quality MillMax socket in and swap away, but there isn't room for that here. I really don't care too much about having a deck that wins spec wars. I just want it as FLAT as possible (+/- 0.75 db) from 40Hz to 15KHz; the rest is just academic to my ears.
Didn't work - there was no sound at all with the OPAs. Can't remember for sure if it oscillated, too much time has passed since then. It is a possibility, but don't remember this to be the problem. It can likely be made to work with some modifications, but at the time I was looking for a plug and play replacement that didn't require any mods. NE5532 proved to work without any mods, only disadvantage being the higher quiescent current mentioned previously (8mA compared to 3.5mA of the NJM4558).
Well, a total of 9 mA more current probably won't make a lot o' diff in most cases. The one thing that was very attractive to me about the OPA1642 was its rail-to-rail performance; but in checking further, in this low-voltage application the NE5532 provides more than enough output voltage swing---not as much, but enough. How much sonic difference did you feel the NE5532 made over the 4558?
