Nakamichi 600
This is the original 'Mark I' which I bought off ebay. Purchasing a 'spares or repair' machine is always a risk, which is why I never pay 'top money' for anything.
Many ebay sellers just want your money, and care not if you're unable to fix the deck - so be aware readers. With that in mind, I especially enjoy and make a point to inform them that I've repaired the deck.
Some of the Repairs
Repairs are many, and what follows was not necessarily the order in which several repairs were made to this deck.
Initial problems -
- Seized DC motor
- Pinch roller had slightly hardened surface
- Heavily worn record/playback head
- Heavily worn erase head
Strangely, for a deck of this age, the drive belt was fine, although it will be changed later.
The initial motor seizure was relatively easy to deal with as I just freed the frictional neck by applying switch cleaner on to the pulley-side bearing, and 'worked' the pulley until the motor was free. Then after cleaning the neck, ISOFLEX PDP65 oil was liberally applied to the said bearing, and the motor was up and running.
Many hours later in the repair cycle, the DC (FG) motor was opened up and serviced.
The Rotor: Be aware of the spacer or washer left-side which is hidden from view. It is very easy for this to get trapped inside the permanent magnet chamber in the second picture.
The Rotor: Be aware of the spacer or washer left-side which is hidden from view. It is very easy for this to get trapped inside the permanent magnet chamber in the second picture.
%20Motor.jpg)
The bearing on the brushes side of the assembly was very dry, and so again ISOFLEX PDP65 oil was injected into the bearing cavity. At this point a spacer was lodged (out of shot) into the stator chamber where sits the permanent magnet.
For a 50 year cassette deck that has seen many hours of use, the brushes appeared not to be heavily worn.
For a 50 year cassette deck that has seen many hours of use, the brushes appeared not to be heavily worn.
%20Motor2.jpg)
Carefully returning the motor to its casing.
%20Motor3.jpg)
It is important to carefully re-assemble any such motor - we want the least possible internal friction for the servo controller to work the motor. So my advice is to: re-assemble, and before tightening the three screws that secure the left-side cap, spin gently the motor shaft and 'feel' for the least torque load on your fingers and thumb. Then incrementally tighten each one in turn until the whole unit is done.
- Free-running, zero load on motor, no belt etc: Im ~ 35mA/36mA (from a cold star, but will rise a little as it gets warmer) Good!
- PLAY action + No Cassette: Im ~ 65mA /66mA (cold start), Expected range!
- PLAY action + a slightly frictional cassette: Im ~ 71mA ... 80mA or higher {From cold start, and depends on the cassette and spool tape mass} Expected range!
Worn Record/PB Head
It was obvious that the head was worn from the playback level, but moreover - the lack of treble in the audio. As a quick check I demagnetised the heads and later adjusted the azimuth just to see that this wasn't purely an azimuth issue, and so it proved to be so.
Disassembled head with wear grooves.
Disassembled head with wear grooves.
There were two options - lap this original head or find some replacement.
However, a 'straight forward' replacement isn't so easy to find, there are probably few or zero amount of NOS Nakamichi heads out there on the internet so I was stuck.
In my stocks of various electronic artefacts, I had a Sony head that would fit, it had a short penetration depth unlike most other possible options. From memory its inductance was comparable, and so this would disturb too much the bias oscillator resonant frequency.
The Sony head proved very successful, although it lacked playback sensitivity. I could easily achieved 30Hz ... 19Khz within +- 3dB (Ref: -20dB 160 nWb/m DIN standard) after a record/playback session. Truthfully, I was astonished by this amazing response, especially from the Sony head, which was originally from a Sony TC-K44 and lapped,
However, just now and then the pre-amp would become unstable with very low frequency oscillations of the order of about 5Hz, yes five cycles per second. This instability could begin after 50 seconds or 50 minutes!
At first I thought this issue may have some instability within the pre-amplifier? After all, there had been some re-capping done within the pre-amp before I bought it. Since I had not played the deck for long with the original head fitted, I could not determine for certain if this was a pre-amplifier issue or not?
I was at a loss to find what the exact cause was, and after lapping the original head, I returned the Nak head to its original owner - this Nakamichi 600. Finally, no instability!
You may notice the two small 'pits' close to the gap on the right channel. They could have been there before lapping?
Provisional record/playback frequency response measurements also returned surprisingly great results: 30Hz to 19Khz +-3dB at the theoretical 160nWb/m DIN level.
Pinch Roller (13mm x 8mm x 2mm ): The reader will have noticed too that the pinch roller had already been replaced. Incidentally, there is apparently only one compression setting available between pinch roller and capstan.
Record/PB Multi-switch
As this is an old deck, I thought it prudent to remove and clean the record-playback switch. These almost always require some intervention, at the very least the application of a switch cleaning spray solution.
This is a very time consuming and delicate job.
First, removal by de-soldering. (Note - the circuit board 'patches')
The sliding contacts were cleaned with switch cleaner solution and a soft glass fibre brush.
The sliding contacts were cleaned with switch cleaner solution and a soft glass fibre brush.
Care must be taken at all stages.
Erase Head Problems
At the time of writing this has not yet been resolved.
Although full erase for Type I tapes is excellent, Type II is more problematic. I first noticed this since the left channel would not fully erase - there was, and still is a periodic resurfacing of sound in the left channel. Studying the erase head reveals that the erasure track fully covers both L/R channel record tracks, so probably no erase head height issue there.
Closer inspection revealed another story however - severe wear on the erase head!
Both ferrite and the plastic surfaces are badly worn, and since Type II tapes require a substantially higher erasure B-field strength, then perhaps it is not that surprising that full persistent erasure is not happening? The effective cross sectional area is being reduced and with it the relative permeability of the ferrite.
Lapping an erase head isn't that easy since we are dealing with two different materials - both of which will have different wear coefficients.
Initial inductance of the worn erase had was estimated at 78uH, and post lapping approximately 73uH.
Post lapping improved erasure, but it has not eliminated it.
Alternative Erase Head
A direct replacement would be desirable, but it is highly unlikely. So another is sought.
Neglecting fitting issues for the moment, most cassette deck erase heads I have measured possess an inductance between approximately 800uH and 1200uH.
As these figures stand, they are useless as a replacement for the Nakamichi 600 for at least one good reason - oscillator resonance.
At the heart of the oscillator is the L302 ferrite cored transformer which was removed and examined 'offline'.
Unloaded, the primary or transistor side had an inductance of approximately 272uH, and the secondary or 'tank circuit' side, 585uH. Both sets are variable by turning in, or turning out the ferrite 'slug'.
This oscillator circuit has a single erase head bias current setting for both Type I and Type II tapes - which means the field strength is the same regardless of tape setting.
The Original Service Manual Schematic
The Original Service Manual Schematic
Solder Side
'Unofficial Patches' refers to direct soldered links found on the solder side of the board.
Resonance: A good estimate of the resonant frequency of the tank circuit can be got from the equation ...
Where C315 = 39nF, and L is the effective inductance of the secondary windings and the erase head. The record head inductance has been ignored (approx 70uH) since that 'tapping' and circuit is also highly resistive.
Without proof, the effective (non-mutually coupled) inductance of two coils is given by
Here we have 1/Lt = 1/585uH + 1/73uH or Lt = 64.9uH.
So then, we can expect the secondary side to oscillate at approximately ...
f(osc) = 1/ {2*pi*Sqrt(64.9uHx39nF)
or
f(osc) = 100Khz.
If a 'new' erase head is going to be fitted, it will need an effective inductance in the region of 60uH ... 85uH. The final oscillator frequency can be 'fine tuned' using the ferrite 'slug' in the transformer.
From measurement, the 'old' erase had an inductance of 73uH which at a measured frequency of 106,000Hz offers a calculated impedance of around 2*pi*106,000*73/10^6 or 48.6 ohms.
The high frequency bias voltage was measured under the original erase head 'load' and was found to be 18.6v rms.
Translated, we obtain the working bias current Ibias.
Ibias=18.6/48.6 amps or 383mA.
This is the rms erasure current working to erase the tape, it is quite a high value, much higher than found in other cassette decks.
There is some heat generated by this 'large' current, part of it will generated from I^2R losses, and swirling eddy currents in the ferrite.
Based an an estimate of head resistance to be about 2 ohms, I^2R losses alone will generate (0.382)^2*2 or 0.3 watts. And that doesn't include eddy current and skin effect generated thermal activity.
The 300mW doesn't look much on paper, but consider the small size of the erase head. Surely over time, this will have had an impact on the softness of the plastic and ferrite area sections? Hence the excessive wear?
Potential Erase Head Replacement
The original erase head is still in circuit, operating normally and successfully for 'Normal' or Type I tapes.
For a replacement to work, it will need an inductance of about 60uH to 85uH as stated above, and is able to erase the tape with its generated magnetic B-field.
Alternatively, another head could be employed with higher inductance, providing of course that the tank circuit still works on the original C315 = 39nF and an approximate and effective 60uH to 80uH inductive load.
I have ordered an erase head with L ~ 390uH and DC resistance of about 2 ohms.
Alone this won't work, but if a dummy load in parallel can be found to make 60Uh to 80uH, then there is a chance of success.
Again, we use 1/Lt = 1/L1 + 1/L2
Here, our Lt is chosen at 75uH (say), and the dummy load (in parallel) L1 = 390uH.
So then ... 1/L2 = 1/Lt - 1/L1, gives 1/L2 = 1/75 - 1/390 or L2 ~ 93uH.
So, if indeed this 390uH head can be sourced, and fitted correctly (inc modifications), then the bias erasure current that the new head will 'sink' will be given by the current divider rule ...
Ibias = 0.383 x 93/(390+93) or 74mA.
Sufficient Bias Field: If and when this head is fitted, will it generate enough field strength to erase Type I and Type II tapes with 74mA AC rms?
Many high impedance erase heads erase their tapes with much lower currents than the 383mA calculated above.
For example, the erase head on the Sony TC-K44 has an inductance of 796uH, and for Type II erasure, the quote rms voltage is 32v.
Ibias = 32/{ 2*pi*105,000*796 x 10^(-6) } = 61mA.
So there is hope that this idea may work, we shall see.
Alternative Tank Circuit Capacitance
Another option is to fit a new erase head in circuit, but lower the value of C315? However, the potential problem still remains - can the transformer's output voltage of around 18.6v rms drive enough Ibias current, hence enough
B-field density and strength?
Additionally, I suspect under an effective higher impedance load, the high frequency bias voltage will be a little higher than 18.6v rms, perhaps closer to 20v rms, since the overall load will be less?
* Here was a break in the work.
Attempt at Paralleled Dummy Inductor
So the experiment continued, and the new erase head has in fact an estimated inductance of LH ~ 333uH at 100Khz testing frequency. Bear in mind, inductance is frequency independent, but the tester uses a test sine wave to estimate the value, so it is best to estimate at approximately the bias frequency.
You can also see that I've had to cut some of the platform away in order to obtain a 'fit' on to the original Nak 600 assembly mounting points.
At this point the erase head was known to be too tall, a careful and slow grinding away of excess plastic had to be done get the head to sit at the correct height. This was to ensure that both Left and right record tracks were going to be exposed to the erasure field density B.
So what of the results of this test?
Well, the short and simple answer is - it worked, and worked well for Type I tapes only. Again, Type II was not being fully erased, but yes - in this configuration both tracks were being partially erased equally, unlike before.
I could have left this 'as it was', but I knew this was never going to be a permanent solution - the fact that Type II tapes were not being fully erased had nothing to do with my decision to cancel this approach.
So why cancel this configuration?
The answer was simple - the dummy inductor was by-passing a lot of AC current during a record session. This I knew before hand, but I was intrigued by its possibility, and I confirmed to myself that it worked. The little 100uH inductor is rated at 150mA, but during a record session a calculated by-pass current of about 205mA was 'warming' the coil up!
At this point the frequency of oscillation was as calculated to within small margins of error, somewhere around 100Khz to 110Khz.
Return to Default Erase Head Load of 333uH
The next and final configuration was to reconnect the new erase head, but this time alter C315 to suit our objective of approximate 105Khz resonance.
Doing the Math, I forecast that C315 needed to be approximately 10.7nF.
Since I have no polystyrene or polypropylene '10nF' caps at this moment I found some 10nF mylar 100v rated capacitors that I could try.
With a measured and new C315 in place of 10.2nF, the record oscillator fired up nicely, but the resonant frequency was considerably higher than expected, somewhere around 125Khz!? So somewhere I made an error in calculation - not sure where, maybe I'll find it later?
Generally, the effective oscillation frequency can be got from ...
Knowing that to decrease the resonant frequency fosc, I found a sweet spot where I added 6.8nF to the overall tank circuit capacitance. Measured, both caps in parallel came to 16.7nF. This did the trick, and fosc ~ 105.4Khz with no instability issues or frequency drift.
Incidentally, the oscillator is still 'fine adjustable' using the ferrite cored transformer L302. It means that I can later opt to install an 18nF 'tank circuit' capacitor, then 'tune up' the oscillator back up to 105Khz.
 |
| A temporary set up using Mylar 100v rated capacitors. Polyester or polypropylene caps will be added later. |
With the oscilloscope set to a high 10M ohm input impedance on the probes, the measure record-mode erase voltage was 18.5v RMS. Actually, just looking at the previous value, it is about the same.
The new Ibias AC current for the 'new' head can be estimated for both tape types, since this Nakamichi 600 has just one high frequency erasure voltage setting.
From ...
From ...
Ibias = 18.5/{ 2*pi*105400*333x10^(-6))
or
Ibias ~ 84mA.
For readers who are interested in the circulating currents 'through' the C315 capacitor, we have ...
IC315 = 18.5 x { ω x C315 } = 18.5 x 2*pi*105400*16.7x10^(-9)
or
IC315 ~ 205mA {AC RMS}
Depth Erasure Figures
Type I tapes received full erasure, but Type II (TDK SA) typically attenuates a Dolby Level 400Hz test tone by approximately -25dB, and fairly equal in both channels, maybe slightly less in the right.
Future Plans?
I have no plans to either - go back to the old worn Nakamichi 600 erase head, and no plans to alter the modified configuration.
Since I don't use Type II tapes very often it is pointless, and I sense - almost impossible to resolve the issue. Purchasing a 'Nak 600 parts deck' would be a waste of money and time. It would also be unjustifiable to disassemble a good Nakamichi 550 to access a potential donor head.
The penetration of the new erase head is deeper than the original Nakamichi 600 head. Of note - there is enough clearance so that the erase head's guide do not clash with the cassette shell when the door is both opened and closed.
Initially, I has to push the guide in further to get adequate clearance.
A Falstad Simulation
A Falstad Simulation
I like to analyse without simulators 'by hand', yes it is sometimes difficult, but rewarding. Falstad is a good simulator, and I was able to confirm my calculations with Paul Falstad's circuit simulator.
***************************************
{This and every article in this blogger is subject to blogspot HTML bug fixes, general alterations and corrections without notice}
06/11/2026
