Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02271551 1999-05-26
A DISK RECORDING SYSTEM AND A METHOD OF CONTROLLING
THE ROTATION OF A TURNTABLE IN SUCH A DISK
RECORDING SYSTEM
Back round of the invention
Field of the invention
The present invention relates to a disk recording
system, for recording a disk such as a compact disk.
The present invention also relates to a method of
controlling the rotation of a turntable relative to a
IO recording head in a disk recording system.
Summary of the prior art
Many systems are known in which information is
recorded on a disk-shaped medium and may subsequently
be played back. Generally the information is arranged
either in substantially circular rings or in a
continuous spiral track on the disk. An example of
the former arrangement is the magnetic floppy disk or
hard disk, where the information is divided into
sectors lying in concentric tracks. Examples of the
latter arrangement include the conventional gramophone
record carrying sound information in analog form in
the spiral groove in its surface, the optically read
videodisk carrying video information in analog form in
a series of pits arranged spirally on the surface (or
on an interfacial boundary) of the disk, and the
compact optical disk carrying audio or other
information in digital form in a series of spirally-
- CA 02271551 1999-05-26
2
arranged pits. The gramophone record, the videodisk
and the compact disk are a11 examples of media
available to the consumer which cannot normally be
recorded on by the consumer; recording takes place on
a master disk which is subsequently replicated by
various processes such that the disks bought by the
consumer are close copies of the geometry and the
information content of the master disk.
The process of recording information on any of
these media usually shares in common the fact that the
disk or master disk is rotated at a speed typically
anywhere between 16~ r.p.m. (for some gramophone
records) and 1,800 r.p.m. (or even higher for some
videodisks ) while the point of recording ( which may be
a magnetic head, a mechanical stylus, or a focused
Light beam) is traversed between the inside and outer
edge of the disk at a slower rate. Normally it is a
requirement of the recording system that the
rotational motion of the disk may vary only slowly, if
at a11; generally this is easily ensured by the
inertia of the disk itself, together with that of the
mechanism which rotates it. The radial motion of the
point of recording on the disk is, however, not so
easily controlled. In the case of magnetic disk
recording, it is usual that the recording head must
move in discrete steps between the separate concentric
tracks; by contrast, in the cases of gramophone
records, videodisks, or compact disks, the recording
CA 02271551 1999-OS-26
3
head must move continuously relative to the disk in a
generally radial direction in order to lay out the
information in a spiral track, and it is
characteristic of these cases that the smoothness of
the radial motion is more important than the absolute
accuracy of radial positioning. With a gramophone
record, for example, any radial motion having
significant energy in the audio frequency band, even
if it represents only a small fraction of the average
groove spacing, will appear as a corresponding lateral
movement of the pickup when the final record copy is
played, and this will be audible as a noise
superimposed on the recorded audio signal. With
videodisks and compact disks there is not only the
possibility that any sudden radial motions of the
recording head will cause the player to fail to follow
the track on the final disk, but also the more serious
likelihood that such motions will be dangerous simply
because they will result in significant changes in
spacing between successive turns of the spiral track.
Since this spacing is typically only 1.6-l.7um, and
any reduction in spacing has the effect of increasing
the crosstalk between tracks (resulting in
interference in the picture from a videodisk, or an
increased likelihood of bit errors with a compact
disk ) it is desirable to maintain a tolerance of at
most +O.lpm in the track spacing, and' preferably a
much closer tolerance than this.
CA 02271551 1999-OS-26
4
To obtain the necessary radial tracking motion,
it is usual to move the recording head along a
straight line which passes through the axis of the ,
disk, in other words radially. When recording
gramophone record masters this is commonly achieved by
mounting the recording head on a linear slide or
rolling mount and moving it by means of a rotating
headscrew and nut. Satisfactory performance is
achieved by careful engineering; the stiffness of the
leadscrew drive is great enough to overcome residual
friction in the mounting. In videodisk and compact
disk mastering (recording) a similar technique may be
used, in which the optics which produce the focused
beam are moved over the rotating master disk. To
avoid the disadvantage that part of the optics are
thus movable while the remainder (owing to the size of
the light source, normally a laser) have to be fixed,
it is alternatively possible to move the entire
turntable (which carries the master disk) together
with its rotary bearing along a straight line, using
a leadscrew, while the recording head remains fixed.
In long-playing optical videodisks, or optical
compact disks used for audio or other data in digital
form, a constant linear velocity mode or recording is
normally used because it allows the maximum recording
time consistent with operation at the optimum linear
velocity (which determines the bandwidth of the signal
which can be recorded) throughout the recording.
CA 02271551 1999-OS-26
Constant linear velocity recording, however, adds
complications to the system, because neither the speed
of rotation of the disk, nor the speed of radial
motion of the recording or playback head relative to
5 the disk is constant. In a playback-only system this
may be no great problem because both rotational and
radial motions are normally controlled by servos
governed by the information already laid down on the
disk. Record/playback systems (for example read/write
data recording systems) using pre-grooved disks are
also simply implemented by servos of this type.
However, in the case of a master recording system for
videodisk or compact disk, where the master disk is
initially devoid of groove structure, there is a
problem in generating the required motions as it were
from first principles.
If information is to be recorded in a spiral
track of pitch P at a linear velocity v then, if the
instantaneous radius is R and the rotational speed is
w (radians/sec.) at time t,
cu = v_
R Equation 1
and dR - _Pcp Equation 2
dt 2n
from which dR - Pv Equation 3
dt 2rcR
EP-A-011495 discloses, in the context of defining
or following a plurality of substantially circular and
concentric information tracks, an arrangement in which
a
CA 02271551 1999-OS-26
6
the relationship of Equation 1 is obtained by
generating a signal dependent on the radius R, and
generating therefrom an alternating current signal of
frequency inversely proportional to the radius R. The
angular rotational speed c>J of the turntable is then
synchronised to that alternating current signal. The
same alternating current signal is used to control the
radial velocity dR, in accordance with Equation 2
dt
using a lead screw drive. A similar method is
described in EP-A-011493, in which an alternating
current signal with a frequency inversely proportional
to the radius R is generated, in this case by a
digital division process.
It may be noted that the methods disclosed in EP-
A-011495 and EP-A-011493 both require that an
alternating current signal is generated first, to
which the turntable rotation has to be synchronised..
As discussed above, it is possible for either the
recording head or the turntable bearing to be made the
moving element. However, whichever is the moving
element, a lead screw drive system is not completely
satisfactory in videodisk or compact disk mastering
because of the great smoothness of motion required.
A very precisely ground lead screw would be needed,
and sticking or slipping of the lead screw mechanism
could cause significant problems.
If a drive mechanism other than a lead screw is
CA 02271551 1999-OS-26
7
employed, generally it will be less stiff than a lead
screw. The smoothness of the radial tracking motion
is then strongly dependent on the attainment of very
low friction in the bearing on which the recording
head or (as the case may be) the turntable bearing
unit moves.
The recording head and turntable may be
mounted so that the axis of the turntable is movable
relative to the recording head, with the direction of
that movement being defined by an air bearing.
The use of such an air bearing has the advantage
of providing very low or essentially negligible
friction in the desired direction of motion, combined
with great stiffness to motion in other directions.
In order to control the relative motion of the
recording head and the turntable bearing assembly, it
is clear that in order to retain the full advantage of
using a friction-free air bearing to support this
motion (whether by moving the recording head or by
moving the turntable bearing assembly) the driving
mechanism should not of itself add friction.
Friction-free means of causing relative motion are
known, for example electric motors using a current-
carrying coil of wire mounted in a magnetic field, or
movable permanent magnets acted on by a variable
magnetic field, or induction or hysteresis motors
using movable electrically conducting or ferromagnetic
CA 02271551 1999-OS-26
elements acted on by a travelling alternating magnetic
field. Generally such driving means have in common
the property that in response to some control signal
they will exert a controllable force on the movable
element; however, when applied to a movable assembly
mounted on an air bearing, they do not
characteristically act in such a way as to define the
position of the movable element, but only in such a
way as to define its acceleration, since the
mechanical behaviour of the moving element on its
friction-free bearing is dominated by its inertia.
This behaviour is in contrast to that of a leadscrew
actuator, which is mechanically stiff and directly
determines the position of the movable element.
Additional suitable means are therefore necessary to
control the position of the movable element.
It is well known in the art to control the
relative motion of two elements by providing position
sensing means to detect the motion, together with an
amplifier responding to both the output of the
position sensing means and an externally supplied
control signal, the output from the amplifier
providing the control input to the motor or other
driving means in such a way that the resulting motion
is constrained in accordance with the externally
supplied control signal. In other words, this is a
negative feedback servo-loop. Such a system
characteristically has an upper frequency limit (or
CA 02271551 1999-OS-26
9
bandwidth) such that for motions having components
predominantly below this frequency limit the motion is
well controlled by the feedback loop whereas for
i
motion having components predominantly at higher
frequencies the feedback loop exerts little control.
In the arrangement described above, relating to
the control of a massive assembly mounted on a
friction-free bearing, the choice of bandwidth (which
may be readily adjusted by changing, inter alia, the
gain of a suitable amplifier) may be a difficult
compromise. In the absence of feedback control a
system consisting of a massive assembly mounted on a
friction-free bearing is very_vulnerable to external
vibration, since in the presence of such vibration
causing the normally "fixed" elements of the system to
move, the tendency of what would normally be termed
the "moving" elements is to remain stationary owing to
their inertia, resulting in a large relative motion
between the two. If the bearing is a rotary one, the
sensitivity to linear vibrations may be reduced by
balancing the moving assembly. There remains,
however, a sensitivity to vibrations which have a
rotational component about the axis of the bearing.
To suppress such relative motion the feedback loop
must have a large bandwidth, covering a11 the
frequencies at which external vibrations may be
present.
If this is done the external vibrations are
CA 02271551 1999-OS-26
1~
attenuated as regards relative motion between the
"movable" and "fixed" elements, but at the same time
any noiselike or other fluctuations inherent in the a
signal generated by the position sensing means
increase in importance; for the action of the negative
feedback loop is such as to attempt to hold the
reading obtained from the position sensing means to a
set value, determined by the external control signal,
so that the inherent fluctuations in this reading
therefore appear, with opposite sign, in the actual
position of the movable element. Specifically, those
inherent fluctuations which fall within the frequency
bandwidth of the negative feedback loop become imposed
on the actual position of the movable element.
Increasing the bandwidth therefore attenuates external
disturbances but increases the effect of fluctuations
in the position sensor reading. In the context of
videodisk or compact disk mastering, it may be
difficult to find a bandwidth great enough to remove
the external vibrations which does not unacceptably
increase the uncertainty caused by these inherent
fluctuations.
It should be noted that the above discussion
holds equally, whether the output of the position
sensing means is directly representative of the
relative position of the movable and fixed parts (so
that the feedback loop acts in such a way that the
external control signal controls the relative
CA 02271551 1999-OS-26
11
position), or whether the said output is
representative of their relative velocity (in which
case the external control signal controls their ,
relative velocity). Similar considerations govern the
choice of bandwidth in both cases.
There is therefore a need to reduce the effect of
external vibrations without relying on a negative
feedback loop of the high bandwidth.
Passive damping of the motion of the movable elements
may be achieved, for example by a fluid-filled dashpot. With such
a dashpot, an outer element is secured to e.g. a fixed element (or
frame) of the system and an inner element is secured to the
movable assembly mounted on the friction-free bearing described
above. Such a dashpot offers viscous resistance to the motion of
the movable elements. By contrast to the effect of inertia alone,
which is such as to stabilise the motion of these elements with
reference to an external (inertial) frame, the effect of such viscous
damping is to tend to stabilise the motion of the movable elements
with reference to the frame of the machine itself, and thus to
attenuate rather than to accentuate any vibrations in the frame as
far as relative motion between the "movable" and "fixed"
elements is concerned. Moreover, this attenuation occurs by
purely passive means, and does not add noise or fluctuations
as the negative feedback
CA 02271551 1999-OS-26
12
servo loop described above would have done.
It is necessary, by way of explanation to make a
clear distinction between the effects of friction and
the effects of viscous drag. Friction occurs when two
solid parts are in contact, and is characterised in
that in order to cause a sliding motion between the
parts it is necessary to apply a force which exceeds
a certain threshold, no matter how slow the desired
motion may be. Viscous drag, as produced by a
dashpot, by contrast sets up a force in resistance to
relative motion which decreases as the speed of
relative motion decreases, so that the rate of motion
may be readily controlled by varying the applied
force, down to the smallest speeds. It is an object
of the present invention to eliminate friction forces
by the use of an air bearing, and to replace them by
viscous forces by the use of a dashpot.
To control the motion of such a system a negative
feedback servo loop, as described above, may be
employed. However, since the dashpot provides the
means of attenuating the effect of external
vibrations, and also since the required speed of
motion changes only gradually with the radius of the
recorded information on the disk in the case of
constant-linear-velocity (CLV) recording, and may not
change at a11 in the case of constant-angular-velocity
(CAV) recording, this servo loop may have a small
bandwidth (i.e. a long response time) such that any
CA 02271551 1999-OS-26
13
fluctuations introduced into the motion as a result of
inherent noise in the position sensing means are not
important. A bandwidth corresponding to a response
time of 5-10 seconds has been found by the applicants
to be suitable.
A rotary bearing may be used to support the
relative movement of the recording head and the axis of the
master disk in an arc rather than a straight line. In one case, the
recording head remains stationary while the turntable
bearing unit moves, and the arc of relative motion is
such that there is a position where the recording head
lies directly on the axis of the master disk. This
enables information or visible markings to be recorded
as close to the centre of the final record as may be
desired. The axes of both the turntable bearing and the second
rotary bearing which supports it may be vertical, so that there is no
gravitational force tending to move the turntable bearing unit one
way or the other.
It may be thought that it is functionally
inferior to allow the recording head to move in an arc
relative to the axis of the disk rather than in a
straight line. In fact, in the case of gramophone
record:mastering, since it is almost universal to play
gramophone records by a stylus mounted on a swinging
arm, it may well be that, if the master disk is
CA 02271551 1999-OS-26
14
recorded by a stylus moving relatively in a curve of
similar geometry to that of the playback stylus, then
the tracing distortion (arising during playback from
the variation in orientation between the stylus and
the recorded groove) may actually be less than with a
linear motion. In the case of optical recording of
videodisks of the "Laservision" type and compact disks
using a focused light spot, the recording process is
not strongly sensitive to the orientation of the
recording head to the recorded grooves. Furthermore,
at least in the case of videodisks and compact disks
recorded at a nominally constant linear velocity
(CLV), the effect of the arcwise motion on the linear
recording velocity is negligible notwithstanding that
the arcwise motion has an angular component relative
to the disk, because the arcwise motion is very slow
compared with the rotational motion of the master
disk.
ZO In summary, the viscous dashpot may damp and
control the radial element of relative motion of recording head
and master disk in a disk recording system where the
relative motion is required to be in the form of a
smooth spiral.
Preferably, the master disk rotates relatively
rapidly on a first bearing and the relatively slow
radial element of the said relative motion is
supported by a second rotary bearing so that the
CA 02271551 1999-OS-26
radial element of the relative motion is a circular
arc.
Preferably, the force to produce the relative
motion is provided by a direct-drive electric motor
5 using the moving-coil, moving magnet, induction or
hysteresis principle.
Alternatively, the force to produce the said
relative motion is provided by a spring, whose other
end may be moved controllably by a geared motor
10 assembly.
A bearing arrangement may be provided for supporting the
relative motion of the recording head and the master
disk in a disk recording system where the relative
15 motion is required to be in the form of a smooth
spiral, in which the master disk rotates relatively
rapidly on a first bearing and the relatively slow
radial element of the said relative motion of the
recording head and the first bearing is supported by
a second rotary bearing so that the relative motion is
a circular arc.
Preferably the second rotary bearing is then an
air bearing, the entire assembly of the master disk on
its turntable and bearing is supported on a bracket
mounted on the rotor of the second rotary bearing, and
the recording head is fixed.
Means may also be provided for sensing the relative motion of
CA 02271551 1999-OS-26
16
two parts whereby a movable set of conducting elements
may move in a transverse direction between two sets of
fixed conducting elements carrying alternating
voltages of opposite phase and the capacitively
induced voltage on the said movable elements forms the
input to a phase-sensitive detector, the reference
input of which is the said alternating voltage applied
to one set of fixed conducting elements, so that the
d.c. output of the said phase-sensitive detector is a
IO voltage representative of the relative position of the
fixed and moving elements.
Preferably, the relative motion to be sensed is
a rotational one, and the fixed and moving elements
have the shape of circular sectors.
Preferably, an additional voltage representative
of the rate of relative motion of the fixed and moving
elements is obtained electronically from the voltage
representative of the relative position of the same
elements.
Preferably, the force to produce the relative
motion is controlled by a linear servo amplifier whose
inputs are the voltage representative of the rate of
relative motion obtained as above and a reference
voltage representative of the desired rate of relative
motion.
CA 02271551 1999-OS-26
16a
Summary of the Invention
In accordance with the invention there is provided a
method of controlling the rotation of a turntable relative to a
recording head in a disk recording system, comprising:
generating a first signal corresponding to the radial
displacement of the recording head relative to the turntable axis,
generating a second signal corresponding to the rate of
rotation of the turntable,
generating a signal representing the product of the radial
displacement and the rate of rotation of a turntable on the basis
of the first and second signals, and controlling the rotation of the
turntable in dependence on the difference between that product
and a reference value.
The invention also provides a disk recording system
comprising:
a recording head,
a turntable for supporting a disk adjacent the recording
head the turntable being rotatable relative to the recording head;
means for moving the recording head relative to the axis
of the turntable,
means for generating a first signal corresponding to the
radial displacement of the recording head from the turntable axis
means for generating a second signal corresponding to the
rate of rotation of the turntable,
means for generating a signal representing the product of
the radial displacement and the rate of rotation of the turntable
on the basis of the first and second signals, and
means for controlling the rotation of the turntable in
dependence on the difference between that product and a
reference value.
From Equation 3, if R = 0 at time to, then-------------------------
CA 02271551 1999-OS-26
17
Equation 4 below holds:
R = ~ Pv ( t-to ) ~~ Equation 4
n
From this, it follows that:
cu = ~ P ( t-to ) ~-~ Equation 5
nv
Thus, both R and ~~ are non-linear functions of
time.
It would be possible to generate these functions
of time in a digital computer, and feed the functions
to digital to analog converters to obtain voltages
representative of the required values of R and c~. It
would then be possible to use those voltages to
control servo systems governing the radial and
rotational motions. However, such an arrangement has
the disadvantage that the values of R and w thus
obtained necessarily change in a stepwise manner, the
magnitude of those steps depending on the resolution
of the digital to analogue converters.
It would also be possible to use a digital
computer to generate a value of the rate of change of
radius R, given by equation 6 below.
dR - ~ Pv ~ Equation 6
dt 4n( t to )
This value, together with the value of a~ derived
by Equation 5, could be used to derive values which
could then be output via digital to analog converters
and fed to servos controlling the radial and
rotational velocities. A single digital to analogue
converter could be used for both values, since they
CA 02271551 1999-OS-26
18
have the same time dependence. Such an arrangement
would have the advantage that there are not the
stepwise variations in R, but only in its rate of
change dR and the effect on the recorded disk is
dt
much less severe. However, the provision of an
appropriate computing system means that complexity is
high.
Applicants have therefore proposed to obtain the
necessary relationships between R and ~ by analog means.
At its most general, the proposal is that signals are
generated corresponding to R and cu; the product of
these two signals is then used to control the rotation
of the turntable in dependence on the difference
between that product and a reference value.
A signal VR corresponding to the radius R is
generated, as is a signal representative of and
proportional to the rotation velocity c~. The former
is preferably a voltage signal and the latter is
preferably a frequency signal. Such a frequency
signal may be produced by a slotted strobe disk
attached to a shaft of the disk recording system,
which may be sensed by suitable optical means. The
signal VR and the frequency signal may then be combined
in a multiplying discriminator circuit which generates
a suitable signal proportional to the product Rc>>.
This product may be a voltage signal, which may then
be compared with a reference voltage Vo representative
CA 02271551 1999-OS-26
19
of the desired linear velocity. The difference in
voltages may then be used to drive a servo amplifier,
which in turn drives a motor which rotates a turntable
of the disk recording system.
Thus, it is possible to form a servo loop which
tends to hold the product Ra.~ constant and equal to the
desired linear velocity. A separate discriminator
circuit may also be used to derive from the frequency
signal a voltage V~ representative of the quantity
2n
where P is the desired track pitch. This voltage V
may then be compared with a voltage VS derived from the
voltage VR, which represents the quantity dR.
dt
The difference between the voltages V~ and Vg may
then drive a servo amplifier whose output causes
radial motion. Thus, the action of this second servo
loop is such as to tend to hold VS = V~ and thus to
hold the radial velocity dR equal to the desired
dt
value
2n
Preferably the disk recording system is used for
optical recording of videodisks or audio or data
compact disks.
Brief description of the drawings
An embodiment of the present invention will now
CA 02271551 1999-OS-26
be described in detail, by way of example, with
reference to the accompanying drawings, in which:
Fig. 1 is a schematic view of a disk recording
system according to an embodiment of the present
5 invention;
Fig. 2 illustrates a linear dashpot which may be
used in the present invention;
Figs. 3a and 3b illustrate a first rotary dashpot
which may be used in the present invention, Fig. 3a
10 being a sectional view and Fig. 3b being a detail of
the dashpot mechanism;
Figs. 4a and 4b illustrate a second rotary
dashpot which may be used in the present invention,
Fig. 4a representing a sectional view and Fig. 4b
15 illustrating a detail of the dashpot mechanism;
Fig. 5 illustrates an arrangement using a spring
for applying a the torque to the shaft of the main air
bearing in an embodiment of the present invention,
together with a means for monitoring the rotation of
20 the end of the spring remote from the shaft;
Figs. 6a and 6b illustrate another arrangement
using a spring for applying a torque to the shaft of
the main air bearing in the embodiment of the present
invention, together with a means for monitoring the
strain on the spring, Fig. 6a being a perspective view
of the arrangement and Fig. 6b being a detailed view
of part the arrangement of Fig. 6a;
Fig. 7 shows a local servo loop for controlling
CA 02271551 1999-OS-26
21
the motor arrangement of Fig. 5 or Figs 6a/6b of the
first embodiment of the present invention;
Figs. 8a and 8b illustrate a capacitive
electrical sensing device for use in the embodiment of
the present invention, Fig. 8a being a sectional view
and Fig. 8b being an exploded view of the disks of
Fig. 8a;
Fig. 9 is a block diagram of an electrical
circuit for processing the output of the sensing
device of Figs. 8a and 8b;
Fig. 10 shows the variation of the output Ve from
the circuit of Fig. 9;
Fig. 11 shows a differentiating circuit for
generating a further output voltage from the output
circuit of Fig. 9;
Fig. 12 shows a servo amplifier for use in the
present invention;
Fig. 13 shows the servo system for the disk
recording system of Fig. 1; and
Fig. 14 shows a multiplying discriminator which
may be used in the servo system of Fig. 13.
Detailed description
Fig. 1 illustrates the general structure of a
disk recording system being an embodiment of the
present invention. A master disk 1 is supported on a
turntable 2 which in turn is mounted on a shaft 3.
That shaft is the inner element of a turntable bearing
4 (or is an extension of the inner element). The
CA 02271551 1999-OS-26
22
turntable bearing 4, which carries the turntable 2, is
mounted on a bracket 5 held by arms 6,7 which arms 6,7
are supported by the shaft 8 of a main bearing 9. The
r
turntable bearing 4 and the main bearing 9 are, in
this embodiment, rotary air bearings, and their axes
are vertical.
Furthermore, a recording head 10 is mounted so as
to be located above the turntable 2 and hence above
the disk 1. In this embodiment, the recording head 10
is fixed, and the relative motion in the horizontal
plane of the disk 1 and recording head 10 is achieved
by rotation of the turntable 2 and movement of the
arms 6,7.
The rotation of the turntable 2, about shaft 3,
is governed by the turntable bearing 4, and the
movement of the arms 6,7 on the shaft 8 is governed by
the main bearing 9 . The arms 6, 7 are positioned so
that the recording head 10 will follow an arc-shaped
path relative to the axis of the turntable bearing 4,
extending between the centre and the periphery of that
disk 1.
The rotation of the turntable 2 on the shaft 3 is
driven by a first motor 11 ( and the movement of the
arms 6,7 on the shaft 8 are driven by a second motor
12. In this embodiment, both motors are
electronically commutated motors with permanent-magnet
rotors; the rotors are mounted directly on the shafts
3,8 so that no separate bearings are required.
CA 02271551 1999-OS-26
23
Also mounted on the shaft 3 is a slotted strobe
13, the rotation of which is sensed by an optical
sensor 14. The sensor 14 generates two outputs which
are square-wave signals in a quadrature relationship.
These outputs are used in a servo system for
controlling the rotation speed of the turntable 2.
The main bearing 9 is supported by a vertical
bracket 15 which in turn is supported from the
underside of a horizontal mounting plate 16 by a
bracket 17. Mounted on the upper side- of mounting
plate 16 is a damper unit 18 which will be described
in more detail later. Also mounted above or within
the concave-topped element 33 from a bracket 19a is a
position sensing unit 19 which will also be described
in more detail later.
As has previously been mentioned, the present
invention requires that the relative movement of the
head and the turntable bearing be damped. In an
embodiment where this relative motion is linear, this
may be achieved by a linear dashpot, such as shown in
Fig. 2.
Such a dashpot may be provided in the case where
there is linear motion, for example where a linear air
bearing is used, by the use of a piston moving in a
fluid-filled cylinder. It is important in the present
invention that a11 friction be avoided; therefore a
conventional seal cannot be used to prevent leakage of
fluid around the shaft of the piston, and instead (if
CA 02271551 1999-OS-26
24
the motion is to be horizontal) an arrangement
corresponding to that of Fig. 2 is best used,
incorporating shrouding means to prevent dust from
falling into the fluid.
Thus, Fig. 2 shows that a piston 20 is mounted in
a container 21, which container 21 is filled with
viscous fluid 22. The piston has a bracket 23 which
extends out of the container 21 and is fixed to the
object whose movement is to be damped. The dust cover
24 is fixed to the bracket, covering the opening 25 in
the container through which the bracket 23 extends, to
prevent the ingress of dust.
Preferably, however, the dashpot used in the
present invention is a rotary one, and is used in
connection with a rotary air bearing, the axis of
which is vertical.
Fig. 3a shows the construction of a dashpot which
consists of interleaved thin annulae.
In Fig. 3a, an annular mounting 30 supports a
fixed damping element 31. The fixed damping element
31 surrounds a shaft 32 (which may e.g. correspond to
shaft 8 in Fig. 1 so that the dashpot corresponds to
the damper unit 18 of Fig. 1) and a moving damping
element 33 is fixed to that shaft. The fixed damping
element 31 is secured to the mounting 30 by a clamping
member 34, which is pressed down by springs (not
shown).
As shown in more detail in Fig. 3b static vanes
CA 02271551 1999-OS-26
35 are provided in the fixed damping element 31, which
are interleaved with moving vanes 36 secured to the
moving damping element 33. A viscous fluid 37 fills
the resulting dashpot.
5 Figs . 4a and 4b illustrate an alternative dashpot
arrangement, having interleaved concentric cylindrical
sections. Components of Fig. 4a and 4b corresponding
to those of Fig. 3a and 3b are indicated by the same
reference numerals. As can be seen from Fig. 4b,
10 however, the static cylindrical sections 38 are
vertical, as are the moving vanes 39. Again, if this
dashpot is used in the embodiment of Fig. 1, the shaft
32 may correspond to the shaft 8 so that the dashpot
forms the damper unit 18.
15 The cylindrical geometry of Figs. 4a and 4b is
preferred because is allows freedom of adjustment of
the movable elements in a vertical direction and
because it is easier to fill the structure with fluid
without causing bubbles. Radial gaps may be left at
20 intervals in one set of cylindrical sections, in order
to allow the fluid to distribute itself evenly. A
very high viscosity is required to control the very
slow speeds of motion required for example in compact
disk recording, but suitable fluids are readily
25 obtainable, for example the fluids sold under the
brand .name "Hyvis" by British Petroleum Ltd. and
available in a range of viscosities. In order to fill
the structure with fluid, the viscosity of the fluid
CA 02271551 1999-OS-26
26
may be reduced by heating it. A further feature of
the use of a rotary bearing in conjunction with a
rotary dashpot according to this embodiment of the
present invention is that rapid changes to the
position of the movable elements may be effected by
raising the clamp 34 to free the entire dashpot
assembly so that the outer dashpot element 31 rotates
within its mountings 30.
The driving force to produce the relative motion
of moving and fixed parts may, as stated above, be
generated~by a direct-drive electric motor using the
moving-coil, moving-magnet, induction or hysteresis
principles. In accordance with the principles
expressed above the motor must not contribute any
friction to the relative motion; this is readily
achieved if the moving part of the motor has no
bearings of its own but is mounted directly on the
shaft 8 of the rotary air bearing. If commutation of
the drive current is required (as in conventional
moving-coil d.c. motors) then it should be generated
electronically by any of a variety of known means,
rather than by mechanical switching.
Alternatively, it has been found possible to
exert the necessary torque on the moving system by
means of a spring, for example a coil spring whose
further end is secured, for example, to the output
shaft of a reduction gearbox, the input shaft of which
is driven by a conventional d.c motor. It is apparent
CA 02271551 1999-OS-26
27
that even if the said output shaft is held in a fixed
position, the spring will exert a torque which changes
only slowly as the moving parts of the air bearing and
damper assembly rotate; thus it is in the nature of
this arrangement to produce a substantially constant
relative motion by purely passive means, and any
adjustments to the speed of relative motion can be
performed by driving the gearbox in a relatively crude
manner.
Thus, Fig. 5 shows that the shaft 8 of the main
bearing 9 in Fig. 1 is connected to a spring 40, which
in turn is connected to the output shaft 41 of a
gearbox 42, which gearbox 42 connects the shaft 41 to
a motor 43. It is found advantageous to place the
motor/gearbox assembly in an inner servo loop having
a relatively short response time. In order to make
this possible, the shaft 41 also extends to a rotary
potentiometer 44. Thus, the output signal VP of the
potentiometer 44 corresponds to the position of the
shaft 41.
Figs. 6a and 6b illustrate an alternative
arrangement, in which the strain in the spring itself
is sensed in order to drive the inner servo loop.
Components which correspond to those of Fig. 5 are
indicated by the same reference numerals. In Fig. 6a,
the spring 45 has a flat section, and a strain gauge
46 is mounted thereon. As is shown in more detail in
Fig. 6b, the strain gauge is connected via flexible
CA 02271551 1999-OS-26
28
connections 47 to an output amplifier 48, which
generates an output signal VP.
Fig. 7 shows the inner servo loop referred to
above. In Fig. 7, the gearbox shaft position sensor
(from Fig. 5) or the spring torque sensor (from Figs.
6a and 6b) is shown at 50. The output signal VP from
this sensor is fed to a differential amplifier 51.
That differential amplifier 51 also receives a control
voltage VQ and generates an output 52 which drives the
gearbox motor 43. Thus this inner servo loop causes
the sensor output Vp to follow the control voltage VQ.
The signal VQ is generated by a further servo loop
(which will be discussed in more detail later) which
has a slow response, so that the position of the
gearbox output shaft, and hence the torsion in the
spring and hence the speed of motion of the movable
elements (namely the main air bearing rotor, the
movable part of the dashpot and the complete turntable
bearing assembly), is controlled by comparing an
external control voltage V~ with a voltage VS
representative of the speed of motion and obtained
from position sensing means for example as described
below.
Turning now to the means of sensing the position
of the movable parts, which may form the sensor unit
19 in Fig. 1, it is proposed that a capacitive
electrical sensing device is provided as in Fig. 8a
and 8b. In the arrangement illustrated, the shaft 8
CA 02271551 1999-OS-26
29
of the main bearing 9 carries an insulating disk-
shaped rotor 60. This rotor carries on both its
surfaces conducting sectors 61 occupying part of each
surface. This rotor is mounted between insulating
plates 62,63 attached to the fixed framework, each of
which carries on its inward-facing surface conducting
sectors (A,B) separated by narrow gaps. The number of
such sectors depends on the angular range of rotation
which it is desired to sense, but in the arrangement
illustrated there are eight sectors allowing a maximum
movement of 45°. The rotor 60 carries on each face
four sectors 61, each subtending 45° at the centre and
equally spaced apart so that the spaces between them
also subtend 45°. The conducting sectors 61 on the
two faces of the rotor 60 are aligned together and are
a11 electrically connected to a conducting cylinder 65
and a flexible lead 64. The conducting sectors on
the two fixed plates 62, 63 are aligned in opposite-
facing pairs which are electrically connected, and in
addition alternate sectors on each plate are
electrically connected, so that there are in a11 two
sets of alternate sectors (A and B) each sharing a
common electrical connection on the two fixed plates
62,63.
Alternative, but functionally equivalent methods
of construction will of course be apparent to those
skilled in the art, for example to form either the
fixed conducting sectors, or the moving ones, or both,
CA 02271551 1999-OS-26
from metallic material in ways other than by forming
them on the surface of an insulating material, and in
particular for a11 the moving sectors to form one
self-supporting piece of metal.
5 Then equal and opposite alternating electrical
voltages (VA and VH) are applied to the two sets of
fixed sectors. Preferably, these voltages are square
waves of amplitude for example 30 volts peak-to-peak.
As a result of capacitance between the conducting
10 sectors on the rotating disk 60 and the sectors on the
fixed plates 62,63, an alternating voltage appears on
the conducting sectors on the disk 60, of waveform
similar to that of the voltages applied to the sectors
of the fixed plates 62,63 and of magnitude and phase
15 depending on the azimuthal position of the said
conducting sectors on the rotating disk 60 in relation
to the sectors on the fixed plates 62,63. This
voltage may be buffered (and optionally amplified) by,
for example, an operational amplifier of low gain and
20 high input impedance, connected by the flexible lead
64, and forms the input to a phase-sensitive detector
(essentially a multiplier followed by a low-pass
filter) whose reference input is the waveform applied
to one set of the f fixed sectors as shown in Fig . 9 .
25 From this phase-sensitive detector a d. c. output Ve, is
obtained, which is representative of the orientation
( a ) of the moving sectors relative to the fixed sectors.
It may readily be shown that the d.c. volt,age Ve
CA 02271551 1999-OS-26
31
It may readily be shown that the d.c. voltage Ve
is substantially proportional to
C~e: _ Ca.
CA + CH
where C" and CB are the capacitances between the
movable sectors and one or other set of fixed sectors
respectively. If the disk and the fixed plates are
flat and parallel at a11 times then these capacitances
are in turn proportional to the areas of overlap
between the movable sectors and the respective fixed
sectors. If the sectors have circular inner and outer
edges, then these areas are linear functions of the
azimuthal angle 8, and CA + CB is a constant. It can
then be seen that the d.c. voltage Ve is a
substantially linear function of 8, except near the
points ( 8A and 88 ) at which the rotating sectors are
aligned with one or other set of fixed sectors. Fig.
10 shows the dependence of Ve on 8 for the case where
the sectors subtend 45° at the centre. Owing to its
symmetrical construction, the sensing device of Figs.
8a and 8b is tolerant of situations where the rotor 60
is not exactly equidistant from, or parallel to, the
fixed plates 62,63.
Clearly this rotation sensing device may -be
modified in a number of ways obvious to those skilled
-- in the art. For example, the number of sectors may be
increased so as to subtend less than 45°; the
sensitivity, i.e. the rate of change of Ve with A, is
then increased but the linear working range (88-8")
CA 02271551 1999-OS-26
32
becomes less.
When this rotation sensing means is applied to
the embodiment of the invention shown in Fig. 1 so as
to measure the azimuthal position of the swinging
turntable bearing assembly in its arclike motion, it
may be objected that, although the voltage Ve is a
substantially linear function of 8, it is not thereby
a linear function of the distance of the recording
head from the centre of the master disk. The latter
distance may be expressed as 2R~ sin (6~A_ ),
2
where R~ is the radius of the motion of the turntable
bearing assembly~on its swinging arm and 6o is the
value of 8 when the recording head lies over the
centre of the master disk (assuming that the system is
aligned so that there is such a position), and
sin ( 6-6" )
2
is not a linear function of 8. However, a geometry
may be chosen for which the linearity is adequately
good for the present purpose. For example, if R~ = 200
mm and it is desired to record compact disks with a
radius of 60 mm, then (A-6o) = 17° maximum, and at this
maximum excursion the voltage Ve differs by only 0.4$
from the value linearly extrapolated from small values
of ( 8-60 ) . A correction for this error may be made
electronically if necessary.
In general it will be found desirable to position
the rotation sensing means such that a zero value of
CA 02271551 1999-OS-26
33
Ve does not correspond to zero recorded radius ( i . e.
the position where the recording head lies directly
over the axis of the turntable). A voltage VR which
does so correspond may be obtained by adding a
suitably chosen fixed reference voltage ( VpFFSET ) as
shown in Fig. 9.
It is found experimentally that a very good
signal-to-noise ratio may be obtained with this
arrangement, and in particular that it is possible to
obtain electronically a second output voltage VS which
is adequately representative of the rate of change of
Ve (in other words, representative of the speed of
radial motion, dR of the recording head relative to
dt
the master disk) to be used in a servo loop to control
the radial motion. The second output voltage VS may be
obtained by a differentiating circuit as shown in Fig.
11. As discussed above in connection with Ve, there is
a small radius-dependent error in Vs (owing to the
geometry of the swinging arm in the embodiment of Fig
1), which may be corrected electronically if
necessary.
The differentiating behaviour of this circuit is
primarily determined by C1 and R1. The optional
components RZ and CZ each serve to limit the high-
frequency response of the circuit by introducing
first-order low-pass responses with time constants RzCl
and R1C2, thus reducing the high frequency noise which
CA 02271551 1999-OS-26
34
appears in the output. It has been found possible to
control the said speed of radial motion in a system
for recording compact disks in the manner herein
described, notwithstanding that the said speed is only
of the order of a few microns per second.
Fig. 12 shows the structure of a suitable servo
amplifier for controlling the speed of the radial
motion. The voltage VS representative of the speed of
radial motion is compared with a control voltage V~
representative of the desired speed, and the
difference (VS-V~) forms the input to an amplifier.
The output VQ of this amplifier controls the speed of
radial motion, for example by controlling the torque
applied to a rotary bearing system damped by a
dashpot, for example by a direct-drive electric motor
or alternatively by rotating the one end of a coil
spring whose other end is attached to the movable
element. The action of the servo system is such as to
tend to maintain VS equal to V~. The servo amplifier
shown in Fig. 12 has an integrating response, and by
adjusting the values of the resistors r and the
capacitor C the overall response time of the servo
system may be set to a suitable value, for example 5
to 10 seconds, such a value being found to be a good
compromise between a long response time causing the
system to be too slow in settling down and a short
response time allowing too much noise from the
rotation sensor to be impressed onto the radial
CA 02271551 1999-OS-26
motion.
Fig. 13 shows a servo system for controlling the
disk recording system of Fig. 1. In Fig. 13, a radius
..
sensor 108 generates both a radius voltage VR and a
5 radial velocity voltage VS. The sensor 108 thus may
operate on the basis of information from the position
sensing unit 19 of Fig. 1 which has been discussed in
detail with reference to Figs. 8, 9, 10 and 11. In a
similar way, the rotation of the turntable 2 in Fig.
10 1 is monitored by the turntable strobe 13 and sensor
14, which generates a pulse train signal 102 (ideally
a square wave ) representative of c~ and a delayed pulse
train signal 104. These signals 102,104 are both
passed to two multiplying discriminators 110,112. The
15 first discriminator 110 also receives the radius
voltage signal VR from the radius sensor 108. From VR
and the signals 102, 104, the discriminator 110
generates a voltage V1 which is proportional to VRa.~.
That voltage V1 is compared with a reference voltage Vo
20 representative of the desired linear velocity. The
comparison is carried out by a servo amplifier 111,
and this generates an output acting as the DC drive to
the motor 11 which rotates the turntable 2.
In a similar way, the second discriminator 112
25 receives the signals from the sensor 14, and also a
reference voltage 113. The discriminator 112 is used
to derive from the said pulse train a voltage V
proportional to u~ alone, which is compared with the
CA 02271551 1999-OS-26
36
voltage Vs which is representative of and substantially
proportional to dR .
dt
The difference between the voltages V~ and VS drives a
servo amplifier 114 whose output causes radial motion.
Thus the action of this second servo loop is such as
to tend to hold VS equal to V~. If the sensitivity of
the discriminator circuit is suitably chosen in
proportion to the desired track pitch P, then V~ can be
made representative of the quantity Pte, whereupon
2n
the action of this second servo loop is such as to
hold the radial velocity dR equal to the desired value
dt
P~
2It
In the arrangement of Fig. 13, suitable
adjustment of the gains and frequency responses of the
servo amplifiers 111 and 114 will normally be
necessary, in a manner known to persons skilled in the
art, to ensure a suitably accurate overall response
consistent with stability. In particular the
amplifier 114 may have the characteristics discussed
above in connection with Fig. 12.
Fig. 14 illustrates a structure which may be used
for the discriminator 110. The structure of
discriminator 112 may be similar, with the reference
voltage 113 replacing the signal VR.
'In Fig. 14, an electronic switch 101 is
controlled by the pulse train 102 from the sensor 14.
The switch 101 generates a voltage signal 103 which
CA 02271551 1999-OS-26
37
alternates between the radius voltage VR and ground, at
the frequency of the pulse train 102. The voltage
transitions of the waveform of the signal 103 cause
current pulses to flow through a capacitor 105 to a
second electronic switch 106, which is controlled by
the delayed strobe signal 104 from the turntable
strobe. That delay should be less than half a cycle.
The delayed strobe signal 104 may be generated by a
second optical sensor acting on the strobe disk 13,
but may alternatively be generated by electronically
delaying the strobe signal 102. As a result, the
delayed strobe signal 104 gates the current pulses
arriving from the capacitor 105, that gating being
alternately between ground and the input of an
amplifier 107. Thus, a mean voltage V1 appears at the
output of the amplifier 107 which is proportional both
to VR and to the repetition rate of the strobe signal
102. Hence, V1 is proportional to VRw.
The capacitor 115 serves to attenuate the
transitions in the waveform of the voltage V1, which
are due to the current pulses arriving from the
capacitor 105.
The discriminator 112 may be the same as that of
Fig. 14 except that in place of VR it uses a reference
2_5 voltage 113 which does not vary with radius, but which
may be set in proportion to the desired track pitch P
so as to obtain an output voltage representative
of _Pc~~ .
2n
