a new turntable arm design

A New Turntable-Arm Design EDGAR VILLCHUR"

Incidental to a description of a new product in the realm of record- playing equipment is this thorough analysis of the physical principles and the geometry involved in the design of an arm and turntable.

IN TWO PARTS-PART 1

T HE INTERDEPENDENCE OF TURNTABLE

AND TONEARM in a record player is analogous to that between a speaker

and its enclosure. The correct mounting of the tonearm depends on its relation- ship to the turntable in exact position- ing (both vertical and horizontal) and in the necessity for common mechanical isolation from the motor, from external stimuli, and from acoustic feedback.

Deficiencies in record player performance result from unwanted relative motion between the arm and platen. The design approach described here therefore considers the arm-turntable as a single mechanism.

Requirements of a Good Record Player

A record player that is excellent in performance will not noticeably intrude its characteristics into the signal a t all. Record player design is advanced enough for this goal to be a realistic one.

The following qualitative criteria of performance were considered most significant when we undertook the design of the AR turntable:

1. Rumble to be inaudible on program material

2. Wow and flutter1 (including vertical warp wow) to be inaudible on program material

3. Speed to be accurate, and stable with time, with changes of line voltage or of mechanical load

4. Arm to be capable of tracing warped and/or off-center records

5. Tracking-error distortion to be low

6. Isolation from mechanical shock and from acoustic feedback to bc adequate for even severe conditions

7. Convenience in handling, rugged- ness, and safety to be provided

* d c o ~ ~ s t i c Research, Inc., 24 Thorndike St., Cambridge 41, Nass.

1 In the American Standard Acoustical Terminology "flutter" refers to any deviation in frequency; "wow" is a colloquial term used to describe those deviations ~vhicli are relatively slow in rate and rec- ognizable as pitch fluctuations.

Quantitative Standards on rumble, flutter, and speed accuracy were estitb- lished in 1953 by the NAB, then called NARTB, for broadcast equipment. These Standards call for rumble a t least 35 db below 1.4 cm/sec peak a t 100 cps (equivalent to 6.3 cm/sec a t 1000 cps on the RIAA curve), flutter and wow no greater than 0.2 per cent peak in reproducing turntables and no greater than 0.1 per cent peak in recording turntables, and a speed accuracy of r 0.3 per cent (21 dots/min. drift on a standard 216-dot 33%-rpm stroboscope card).

Unfortunately these Standards do not provide adequate guideposts for describ- iilg a completely self-effacing record player, in that they do not take into ac- cou& the frequenEy destribution of the rumble or the predominant rates of the flutter. While an over-all rumble level of 35 db below the NAB reference is a very low figure when measured as speci- fied by the NAB, it is an incomplete specification. Rumble a t the higher bass frequencies needs to bc considerably lower in level than the NAB Standard to be inaudible, because of increased hearing sensitivity in this range, while subsonic rumble much higher in level than the Standard will offend neither the ear nor the amplifier.2 The NAB document specifically points out that its rumble measurement reflects only an electrical effect, not aural annoyance value.

Flutter whose rate is in the 3-cps region is of the order of three times as noticeable as flutter a t 30 cps, and twice as noticeable as flutter a t 0.5 cps3 An ASA Standard of 19544 established a

2 While rumble that is subsonic in frequency is inaudible, it cannot be ignored because it can load the amplifier. The order of amplitude involved in rumble which is anywhere near the NAB standard, however, is so low as to rule out any possibility of loading the amplifier significantly. Rumble which is no more than 20 db below reference, for example, will only drive the amplifier t o deliver an additional 1 per cent of power.

3 F. A. Comerci, "Perceptibility of flutter in speech and Music," JSMPTE, June, 1955.

unit called "flutter index" which takes into account the subjective influence of flutter rate, but no scale of values is suggested.

On the basis of this incomplete state of Standards relative to record players, it was decided to design a unit which would meet NAB specifications, and would in addition stand up to more stringent measurement criteria that were weighted for subjective sensitivity to rumble frequency and to flutter rate. Such weighted measurement data werr available through the use of a flutter and rumble meter, referrcd to later on.

The NAB Standards call for a rumblc- level meter with the same ballistic characteristics as a VU meter; this speed of needle response has been found to cor- respond well to aural effects. Both rumble and flutter measurements can be "improved" by employing a more highly damped meter movement. Another method of showing a dramatically lower rumble figure is to make the measurements with the pickup working into a flat preamplifier rather than one with the normal RIAA equalization. With a flat preamplifier, the rumble measurement relative to a 7-cm/sec 1000-cp? test signal will be approximately 19 db lower than the corresponding NAB figure. Such a measurement does not rep- resent the actual level of the rumble under operating conditions, but it appears to be in use by one testing or- ganization.

Solutions to the problems of turntables and arms appear to the writer to br a matter of correct design rather than of expensive processes and ma- terials. Since the record-player ar t is well advanced, the AR design borrows liberally from good engineering practice existing a t present, combined with some new approaches.

Mounting

Aberrations in record plnycr per- formn.nce must result from any relxtivc

4 "American Standard Method f o ~ Ds- term,zng Flwtter Content of So~cnd Record- ers and Reproducers," 257.1-1954.

AUDIO SEPTEMBER, 1962 19

<

Foam grommet

Mounting suspension (one of three)

Fig. 1. Suspension system of the arm-turntable assembly. The tonearm pivot and platen bearing are connected by a steel I-beam, which i s suspended from the top plate by three springs. Note that neither the tonearm nor the platen touches the top

plate at any point.

notion between the arm and platen. A player can tolerate all sorts of violent motion so long as the arm and platen do not move relative to each other. Therefore the main platen bearing and the tonearm pivot are ideally mounted as rigidly as possible with respect to each other. The AR design uses a steel I-beam as a basic frame on which these two parts are mounted, with a crossbar to form a T-shaped mounting structure. The rigidity achieved could be equalled by a very heavy steel plate, but a t much greater expense and inconvenience. As i t is, the entire frame can be suspended with springs from a top plate that is both light and inexpensive.

This system decouples the arm and platen from the motors, and also shock- niounts the arm-platen assembly as a unit against external mechanical shock and acoustic feedback. It is possible to deal a moderate hammer blow directly to the top plate or to stamp on the floor violently without making the needle jump grooves.

Balance is an important element in shock mounting. With an unbalanced assembly there would be a greater tendency for the entire unit to rock, that is, to move rotationally. The inertia a t the different points of support would not be equal, and one suspended point of the assembly would move farther than another in response to the same external force. Therefore the points of spring suspension are located so that they are equidistant from the center of gravity of the complete assembly, and are spaced on equal arcs.

The lower the resonant frequency of the shock-mounted assembly the better

the isolation from external forces. As the frequency of a stimuating force is lowered below resonance, the mounting suspensions act increasingly as though they were rigid. The practical figure chosen for mounting resonance was 3.5 cps. I n this range the resonant frequency of a mounted turntable assembly can be determined simply by giving the spindle a push and counting the rate of free oscillation. Unfortunately many record players are mounted with a much higher resonant frequency, often so high that the frequency cannot be determined visually. Performance of such record players can be improved considerably by a more wmpliant mounting. Stability in the presence of floor shocks is increased greatly, and perhaps more important, acoustic feedback is reduced or elimi- nated.

The importance of the latter is often not recognized. Acoustic feedback is sometimes responsible for a seemingly inexplicable "boomy" quality of repro- duction or for an apparent increase of rumble, created by incipient low-fre-

Fig. 2. Three types of static balance: (A) unstable, (B) stable, a n d (C)

neutral.

quency oscillation. The floor and room structure provide a path of positive feedback from the speaker back to the pickup; proper shock-mounting of the player assembly opens the feedback loop.

A qualitative method of testing for acoustic feedback from speakers to the record is to place the needle on a record a t rest, and then to observe how fa r the volume and bass-boost controls can be advanced without creating feedback. While the information gained by such a test is entirely relative, the adequacy of a particular installation can be eval- uated roughly by turning the volume and bass controls to the maximum position in which they would be used in that installation, placing the needle on the still record, and stamping lightly on the floor near the turntable or tapping the table surface on which the turntable rests. Any tendency to feedback will be evident in a train of low-frequency os- cillations following the mechanical ex- citation.

The positions to which volume and bass controls can be advanced is a function of pickup output, preamplifier sensitivity a t low frequencies, and room conditions. I n a typical installation the AR turntable does not show feedback instability with the volume control a t maximum and the bass controls fairly well advanced.

A simple method of choosing mounting springs is to use the formula F = (10/D) ", where F is desired resonant frequency and D is the static deflection of the spring in inches when loaded with its mass. If one wanted to shock-mount an assembly weighing x pounds with three springs, for example each spring would be tested with an x/3-lb. load. Fo r a resonant frequency or 3.5 cps the static deflection would have to be 13/16 in., which is to say that the length of the loaded spring a t rest would be changed from its unloaded length (either stretched or compressed) by this amount.

Figure 1 illustrates the T-frame suspension system of the AR turntable. Foam washers that help center the spring also provide a light amount of damping, which is desirable. Note that the platen bearing, T-frame, and tone-

PIVOT - A - B -

C

20 AUDIO SEPTEMBER, 1962

arm pivot will move as a unit if excited through their mounting springs. Shock- mounting either the tonearm or the platen independently would tend to increase rather than decrease relative motion between the two.

The H. H. Scott turntable and the more recent Stromberg-Carlson unit used a similar design approach in sus- pending a rigid arm-platen assembly from the top plate.

Tonearm Balance

There are three kinds of static balance-unstable, stable, and neutral. There is also dynamic balance, which has no significance for tonearms and which will be discussed a little later.

Unstable balance exists when the line connecting the centers of gravity of two sides of a balanced system passes above the pivot, as in A of Fig. 2. If either side is tipped it will continue to move. This is undesirable for tonearms, because the stylus force will not be the same in all arm positions, and will vary with warped records. It is easier for the needle to skip when it hits a bump.

Stable balance, useful in scales, exists when the line between centers of gravity passes below the pivot, as in B of Fig. 2. In this case there is a tendency for the system to return to the horizontal if it is tipped. The stylus force of a tonearm in stable halance is again different for different arm positions. As in the case of the unstable arm, a warped record will creatc instantaneous changes in stylus force, although in the opposite direction. When an arm in stable balance is lifted by an uneven record surface, stylus force will increase, creating a tendency for the needle to dig into the record.

Neutral balance, most desirable for tonearms, is illustrated in C of Fig. 2. Here the line between centers of gravity pnsses through the pivot. The system pill be in equilibrium a t any angle, and stylus force does not vary with vertical motion of the arm.

Neutral balance for the horizontal plane can be provided in various ways, but the principle is the same; the line between centers of gravity of the forward and rear sections of the arm must pass through the pivot for horizontal motion.5 Five different methods of keeping the horizontal pivot on this line are currently employed (see Fig. 3) :

(A) The counterweight and cartridge shell are connected by a straight arm, which must then enter the offset cartridge shell obliquely (Gray, GE, Strom- berg-carlson).

(I3) The counterweight and cartridge shell are in line as in (a) , but in order to allow the arm to enter the shell at a right angie the arm is given a double curve (ESL, Weathers, Ortofon, AR).

5 I am indebted to John McConnell of EST, for first calling my attention to the significance of balance in the horizontal plane.

tering cartridge sheli bff-cenjer, (b) arm with reverse curve, (c,) offset counter. Vertical WarpWow

weight, (c,) additional "outrigger" A11 records must be considered as hav- weight, and (c,) offset horizontal pivot. ing normal warp which creates vertical

'

(C) A single-curve arm enters the a& motion. EV& a record that fulfills cartridge shell at a right angle, and the NAB Standards of Good Engineering center of gravity of the counterweight Practice is allowed a 1/16-in. vertical is shifted in the opposite direction either \Tarp, and obviously all records do not, through the Of the counterweight particularly after use, meet this stand- (C,, Dynaco), by an additional adjustable side weight (C,, Grado, Relt-o- Icut), or by moving the horizontal When the cartridge moves up and pivot inward from the vertical pivot (C,, down along an arc the needle must move Empire). back and forth along the groove, as Stable balance in the vertical plane illustrated in Fig. 4. The instantaneous

does not need to be perfect. The quanti- relative speed between needle and groove tative effect of different degrees of un- will therefore be changed. I t makes no stable balance is the given by the follow- difference, of course, whether the platen ing examples. I f the pivot is l/z in. be- has actually changed speed or the needle low the line connecting the arm centers moved along the groove; the audible and of gravity, the net force of gravity on measured wow is the same. the pickup, with typical arm parameters, I t can be seen from Fig. 4 that there would vary roughly 1% grams for every nil1 be minimum longitudinal needle mo- inch the pickup was lifted above the tion when the pivot controlling vertical record surface. I n a reoord with a 3/16- motion is as far back as possible, and in. warp (about as much as can exist in when the height of this pivot brings a playable reoord), the instantaneous it as close to the plane of the record stylus force would vary by 3/s gram. If surface as possible (more precisely, to the pivot were only Ys in. from the the average plane of the warped sur-

Fig. 4, Geometry of vertical warp wow. When the needle is swung up and down by vertical warp, it must also move horizontally (distance d) along the groove. p = pivot

height, w = warp height, I =arm length.

height needed for neutral balance the variation in stylus force would be cut irl four, to 3/32 gram.

It was decided to position the AR pivot exactly in line between centers of gravity because there was an additional advantage involved. The AR arm is so designated that if it is dropped from a point several inches above the record its rate of fall is slowed up by a damping mechanism (released when the arm touches the record). An unvarying downward force on the pickup to a distance of three or four inches ahove the record was therefore desirable, to keep the rate of fall constant.

It is no more expensive to design an arm with neutral balance than one which does not have it. There is an allied problem, however, which affects the

AUDIO 0 SEPTEMBER, 1962 21

Fig. 3. Five ways of achieving neutral the form of this design-vertical warp horizontal balance: la) straiaht arm en- wow.

face). Warp wow as high as 1 per cent has been reported; a little geometry applied to Fig. 4 will show that such a high percentage can be achieved by using a very high pivot and a very short arm for vertical motion.

The vertical pivot-to-needle length of the AR arm is 9 in. With an arm of this length, and a hypothetical pivot set 1/2 in. above the record surface, longitudinal displacement of the needle along the groove (d in Fig. 4) for a normal 1/16- in. warp would be .0032 in. Assuming that i t takes 0.15 seconds for the needle to travel this distance-representing a not atypical bump in the record cover- ing about 60 degrees of arc-the wow in- troduced would be an undesirable 0.1 per cent. This is to say that the instantaneous needle-groove velocity (19.2 in/sec a t a 5-in. radius) would be changed by 0.1 per cent, just as if the turntable had been slowed or speeded up by that amount. This much wow from a single cause is significant in relation to the NAB limits for wow stem- ming from all causes of 0.1 per cent for recording and 0.2 per cent for reproducing turntables. On the other hand, the pivot could not be lowered from this height without creating unstable balance.

The way out of this dilemma was to lower both pivot and counterweight. The line between centers of gravity of the forward 'znd rear sections of the arm, using a typical lightweight cartridge, tken passes through a pivot y4 in. above the reeord surface. This is the pivot height actually used in the AR turntable. Warp wow under the conditions described above is approximately 0.05 per cent.

The rear section of the arm slopes a t an angle of 1.5 deg. to the horizontal, the slope of the line between centers of gravity. This is so that neutral balance will not be disturbed a t any adjustment of the counterweight.

A warped reeord tends to exhibit more vertical motion when it is supported a t its center than when it rests on its outer surface. The outer section of the AR turntable mat is therefore raised slightly.

Dynamic Balance

The te~-ln "dynamic balance" has been described by writers in the field as re- ferring to: a) the use of a counterweight, b) the use of a statically balanced arm with stylus force supplied by a spring, or c) static neutral balance i i ~ both horizontal and vertical planes.

i Dynamic balance as a standard term in physics has a more specific meaning.

Dynamic balance is a condition of balance in which forces created by rotation of the arm about the pivot do not upset the balance conditions that exist a t rest. Static balance has to do with the forces of gravity, while dynamic balance has to do with iuertial forces. An

example of combined static balance and dynamic imbalance given by the Inter- natidwal Dictionary of Physics amd Elec- tronics ( D . Van Nostrand CO., 1956) is shown in Pig. 5, and this system illustrates the type of dynamic unbalance that can occur in a tonearm.

The system of Fig. 5 is in neutral balance statically, and will remain in equilibrium and a t rest when left in any position. When the system is rotated, however, the centrifugal force of each weight creates a force perpendicular to its own arm. Two opposite forces are thus exerted on the horizontal pivot shaft a t two different points, creating a "couple" that tries to turn the shaft about an axis perpendicular to the plane of the weights.

Applying this concept to tonearms it will be seen that dynamic balance requires that the line between centers of gravity on each side of the arm not only passes through the pivot but is perpendicular to the axes of rotation. Offsetting the counterweight horizontally as is done

Fig. 5. Example of combined static balance and dynamic unbalance given by the International Dictionary of Physics and Electronics, 1956. Courtesy D. Van

Nostrand Co.

in some current designs, or vertically as is done in the AR arm, or using side weights, loses dynamic balance.

Dynamic balance was deliberately ignored in the AR arm, as i t is in most current arms, in order to lower the counterweight and make it possible to use a lower pivot for vertical motion without losing neutral static balance. This means that when the arm is moved horizontally by an off-center record, tiny vertical forces will be created because of the application of centrifugal forces at two different points on the axis of the pivot.

The magnitude of this vertical force, assuming the worst possible conditions of record eccentricity and dynamic uu- balance, is measured in thousandths of a gram and has no significance for reeord players, even with stylus forces of a gram or less.

Tonearm Mass

When the tonearm is set into motion vertically by record warp, or horizontally by reeord eccentricity, inertia becomes a controlling element. The greater

the inertia a t the end of the arm the more the stylus will alternately dig into the groove and try to leave it.e

This inertia must not be confused with stylus force, even when the stylus force is provided by unbalancing the counterweight. I n order to set the arm into motion the inertia of both front and rear sections must be overcome. The momentum of both cartridge and counterweight i s in opposition to the required reversal of arm motion when record warp changes slope, or when an eccentric center hole swings the arm back and forth.

Thus the effect of the inertial mass of the counterweight must be added to, not substracted from, the total inertial mass in both horizontal and vertical planes. Clearly the arm should be as light as possible for maximum stability. I t might seem contrary to common sense that a heavy arm will not keep the needle in the groove as well as a light one in the face of record warp and eceentricity, but it is true. No matter what the stylus force, once a heavy arm is set into motion either sideways or upwards, the chances of the needle leaving the groove are greater than in the case of a light arm.

I n previous years too light an arm involved the danger of allowing too hig:~ a resonant frequency of the needle-suspension/arm-mass system, with an attendant bass-response peak in the audible range. Modern cartridges have such highly compliant needle suspensions, however, that it is possible to work for minimum arm mass and still keep the resonant frequency in the subsonic range. Further, in the case of the AR turntable the drive motor has a fundamental frequency of 6% cps (400, r ~ m ) , and the arm resonanee, which is iu the 10-eps region with typical stereo cartridges, should not fall too close in frequency.

The first prototype of the AR turntable employed an arm with an aluminum cartridge shell, which turned out to be heavier than desired. By changing to an acrylic plastic the weight was reduced by more than half, to 7 grams. This is a double saving, since the required mass of the counterweight is also reduced.

The tracing capabilities of an arm (ability to maintain proper needle- groove contact) may be tested by sub- jecting the arm to severe adverse eon- ditions such as those provided by a badly warped record. Some years ago C. Q. McProud suggested a tracing test that consisted of playing a 45-rpm reeord placed eccentrically on the turntable, that is, with the spindle against

6 R. E. Carlson, "Resonance, tracking, and distortion," JAES, V. 2, No. 3, July, 1954. Mr. Carlson uses the concept of "equivalent mass" referred to the stylus tip in discussing inertial effects.

22 AUDIO SEPTEMBER, 1962

one edge of the large inner hole. The arm is swung back and forth-creating, of course, severe mow due to the changes in path length-but the needle is espected to maintain contact with the groove.

The hIcProud test was conceived in terms of the needle forces then current, ~vhich were generally in the six- to eight- gram rnngc. As the stylus force is reduced the tendency for the needle to be thrown from the groove in the McProud test is correspondingly increased. The inertial force tending to throw the arm sidewa?.s and the horizontal-pivot bearing friction remain the same for a given arm, while the downward force holding the needle in the groove is reduced.

The McProuci test a t six grams vertical force would not be a stringent one for modern arms, but it becomes increasingly difficult as the stylus force is reduced. The AR arm begins to fail the JfcPrond test in the range between 1/2 and '?/4 gram. The exact value depends on the mass of the cartridge, which con- tributes to the total inertia of the arm.

Counterbalance: Spring vs. Weight

Thcre are three methods currently used for providing vertical stylus force: a ) , the use of a spring to cancel out all but the desired mass of the cartridge and arm, with no coiinterweight, as in the Bogen; b) , the use of balanced masses, with a spring applying the desired force, as in the ESL, Grado, Shure (universal), Empire, Dynaco, and so on; and c), the use of an underbalanccd counterweight to provide stylus force, aa in the Pickering, Weathers, SME, H. H. Scott, Shure (integrated), Pritchard ADC, and AR.

There is a rationale to each of these methods. The counterweightless, all- spring system provides the lowest inertia, but the greatest sensitivity to external mechanical shock and to tilt of the turntable because of the extreme unbalance of masses. The balanced-mass, spring-loaded system provides the greatest stability under conditions of im- perfect leveling and exposure to external mechanical stimuli, other things being equal.

Once ncutral balance was established in the AR arm, stylus force had to be provided, either by adding a spring or by unbalancing the counterweight. The latter method was chosen, mainly for simplicity and reliability. ( I t must be understood that the advantage of neutral balance, in keeping stylus force constant a t different arm positions, remains.) It was felt that the problems solved by the balanced-mass, spring- loaded system did not exist in the AR turntable, although this system could certainly provide advantages in other applications. The AR turntable is already isolated to an unusual degree from

Fig. 6. (a) Tonearm damping release mechanism. The arm is damped if dropped, but has 34" free vertical play while playing a record. (b) Side view of damping re-

lease mechanism.

external mechanical shock by a balanced system. So far as tilt is concerned this suspension system provides a degree of self -leveling sufficient for the normal slopes of floors and furniture.

On the other hand simple mass coun- terbalancing presented the positive advantage of reliability of setting. There is no reason for the stylus force, once set for a particular cartridge, to change.

Tonearm Pivot

The first prototype of the AR arm employed damped pivots in both vertical and horizontal planes. The advantages sought included suppression to tonearm resonance and convenience in handling.

While the damped pivot7 did provide these advantages when it was designed, its use with modcrn high-compliance stereo cartridges presented new problems. I f enough damping were used to brake the prototype AR arm adequately when dropped, the impedance of the arm pivot while playing records was increased to an undersirably high value. The needle suspension, rather than the total arm, tend to yield to record warp and eccentricity, and the arm tended to "hang up" on warped records.

The advantage of a damped arm drop appeared very attractive, however. There are few people who have not a t some time fumbled a tonearm and allowed it to slip from thcir fingers onto the record.

A design was worked out (for which patent application has been made) in which the arm drop is damped to the desired degree-the arm takes several seconds to reach the record when dropped. As soon as the needle reaches its destination, however, the damped bearing surfaces are disengaged from the arm by a simple dcvicc, illustrated in Fig. 6.

When the arm is dropped the front edge of the stopper hole presses against the stopper pin, which is smaller in diameter than the hole. This stopper pin G rigidly mounted on a porous bronze shaft impregnated with silicone fluid. The shaft is forced to rotate inside a

7 W. S. Bachn~nn, The tlpplicxtion of damping to phonograph reproducer arms. Proc. I.R.E., 40, 2, Feb., 1952, 133-13'7.

Delrin slecve, and the silicone provides the necessary drag.

When the needle comes to rest a t the lowest point on the record there is no longer any force between the stopper hole and the pin. Then, when a warped section of the record lifts the arm, the stopper hole and pin are free of each other and no force is exerted on the damped bronze shaft. The arm lifts on its operational pivot, consisting of pol- ished steel conical set screws turning - iu Delrin bearings.

The arm is allowed this freedom over a vertical distance of 3/s in. at the stylus. When the arm is lifted by hand the back edge of the stopper hole engages the pin and forces the bronze shaft to turn again, resetting the damping mechanism. Horizontal arm motion remains un- damped a t all times.

The release of the damping mcchnn- ism during record play makes it possible to use the necessary degrre of damping for the arm drop, without having to consider the effect on tracing records. In order to keep the same velocity of drop a t different adjustments of the counterweight (for different cartridges) the absolute amount of viscous friction must be adjustable. This is accomplished by changing the pressure on a spring- loaded washer resting against part of the silicone-soaked bronze shaft. The pressure is varied by turning the tonearm spindle relative to the arm itself.

Tracking Error

The equations for minimum tracking error distortion are well known.8, Tracking-error index is a function of tracking error divided by record-groove radius, since the same value of tracking error creates increasing distortion a t the inner grooves. Offset angle and overhang are calculated on the basis of the maximum and minimum groove radii the arm is designed to play. I t must be noted in passing that an arm designed

(Continzted on page 69)

8 B. B. Bauer, "Tracking angle in phonograph pickups," Electronics, March, 1945, p. 110.

9 J. D. Seagrave, "Minimizing pickup tracking error," Audiocraft, December 1956.

AUDIO SEPTEMBER, 1962

distance of the AR arm was made adjustable. The turntable is factory-set for a cartridge with a %-in. distance between mounting centers and needle tip. For any other cartridge dimensions the effective arm length is adjusted and tested with a plastic template that fits over the spindle. When the needle rests in the small template hole the overhang is correct--0.688 in.

A 9-in. arm (pivot-to-needle) with minimum tracking-error index (and therefore minimum tracking distortion) whose design is calculated for record radii of 5.7-in. outside and 2.4-in. inside, will show a maximum tracking error index of 0 . 3 2 deg. per inch. This means that, a t any groove radius, the maximum tracking error that will ever appear will be 0.32 deg. multiplied by the radius. At any given radius the actual tracking error may, of course, be less. The maximum tracking error over the entire record occurs a t the outer- most groove, and is 1.8 deg. Minor differences in the value of maximum tracking error for a correctly designed 9-in. arm will exist if the arm is designed for a slightly different range of groove radii. However, a conventionally pivoted arm 9 in. or less which is rated as having a maximum tracking error significantly lower than 1.8 deg. reflects either im- proper design (achieving a lower figure for maximum tracking error a t the expense of higher distortion a t the inner grooves) or a copywriter's misguided enthusiasm.

To say that the AR arm design provides the theoretical minimum of tracking-error index for a 9-in. length is merely reiterating the validity of equations that have long been known and ac- cepted. The statement applies to any ot,her arm designed from the same equations. There still remains the mechanical embodiment of the theory, in keeping the overhang and offset angles accurate in actual production. One measure taken here for this accuracy is to form the aluminum tubing with a permanent die; another is to provide the plastic overhang adjustment template referred to.

Electrical Capacitance

Some modern cartridges require low capacitance across their output terminals. The shielded cable supplied with the AR turntable has a capacitance of 21 pf/ft., and the length has been pur- posely kept down to 4.5 feet. Within the arm itself the cables are not shielded until they leave the rear housing in order to keep the capacitance down. The curved section of the arm, being aluminum tubing, acts as the shield. Total capacitance between the cartridge shell and the amplifier input is approximately 135 pf per channel, suitable for any commercial cartridge.

TO BE CONTINUED

the manual turntable you can play automatically A superbly engineered unit developed especially counterbalanced tone arm and 'feather-touch' for use in high quality music systems for faith- pushbutton automatic controls. Model 10H with fully reproducing high fidelity stereo and mono- hysteresis-synchronous motor, $99.50; Model 10 phonic recordings. Plays single record manually, with heavy-duty induction motor, $89.50 (prices automatically or continuously, and up to 10 rec- are less cartridge and base). Descriptive litera- ords, automatically. Operates a t r-------------------------- ture at your dealer, or write to four speeds. Has a heavy die-cast, ! ENJAM N ! Benjamin Electronic Sound Gorp., 12-inch balanced turntable, mass- ! MIRACORD ! 80 Swalm Street, Westbury, N.Y.

L--------------------------A

CIRCLE 69A

388 GIANT-SIZED

-It's Better Than Ever It's from the I

Check these exclusive Lafayette benefits \I Satisfaction guaranteed or money refunded. Free 30-Day Home Trial.

V Largest Stock Selection V Exclusive Lafayette Kits V lowest Prices V S ~ e c i a l Hi-Fi System Prices VNO MONEY DOWN. UP TO 24 MONTHS TO PAY

1 ' SAYETTn LAFAYETTE MAIL O R D E R & S A L E S CENTER WL a a 111 JERICHO TURNPIKE, SYOSSET, L. I.. N. Y.

JAMAICA NEW YORK NEWARK BRONX PARAMUS P L A I N F I E L D SCARSDALE New York 1 New York ( New Jersey I New York ) New Jersey I BE:N I New Jersey I New York

~ ~ ~ ~ ~ ' , , , ~ , , , ~ , , , , ~ ~ ' ~ - - - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ? . ~ ~ ~ ~ ~ ~ ~ ~ ~ J LAPAYETTE RADIO Mail the coupon TODAY for

- . A , P . . O X I , s t I . N.Y. Your FREE lga lafaYette Catal0~ I I I

Name .-.., ............. L- - A I Address. . i t Zone - ........... State ........... !

CIRCLE 69B

AUDIO SEPTEMBER, 1962

A New Turntable-Arm Design

EDGAR VILLCHUR"

Incidental to a description of a new product in the realm of record- playing equipment is this thorough analysis of the physical principles and the geometry involved in the design of an arm and turntable.

IN TWO PARTS-PART 2

Two types of motors are in general use for turntables. These are the induction motor (4-pole for the better units) and the hysteresis synchronous motor. The induction motor can be made with high starting and running torque- about three times higher than that of a hysteresis motor of similar size and rating. A disadvantage of the induction motor is that its rotor speed is less than that of the synchronous rotating field, a phenomenon referred to as rotor "slip." There must be a relative velocity between the rotor and the rotating magnetic field, or no current and no magnetic poles will be induced in the rotor.

The exact amount of the slip, which determines the motor speed, can vary with changes of line voltage, mechanical load, and temperature. Measures taken to counteract this speed drift, however, can result in stable induction motors.

The rotor of a hysteresis synchronous motor, unloaded mechanically, turns a t the same speed as that of the rotating field. The rotor, being a t rest with respect to the field, sees fixed poles which induce the rotor poles, and the motor works on these directly induced poles. I ts speed in the unloaded condition is thus purely a function of the number of stator poles and of line frequency, un- affected by line voltage or motor temperature.

When the mechanical load of a hysteresis synchronous motor is increased sufficiently, however, the motor will slow down and work a t sub-synchronous speeds on new sets of rotor poles, in spite of the relatively high retentivity of the rotor material. To overcome this difficulty hysteresis motors are typically made heavy, with the attendant increase in cost.

One type of hysteresis synchronous motor that has been used in recent years

* Acoustic Research, Inc., 34 Thorndike St . , Cambridge 41, Mass.

Fig. 7. Motor drive system of AR turntable.

is the small motor normally intended for clocks and other time-regulation applications. The motors of this type that have appeared in turntables have some definite advantages, over and above low cost, over previous motor designs. The central feature from which these advantages stem is a low rotational speed. This makes it more convenient to use a single-step drive system, that is, a system which employs a single belt or friction wheel from the motor pulley to the turntable platen, with no idler wheels 01. other intermediary coupling devices. These motors turn a t 600 rpm, and the speed reduction ratio for 33% rpm is only 18 to 1. With the standard 1800- rpm 4-pole induction or hysteresis motors, the speed reduction ratio would be 54 to 1. A single-stage system carries with it the advantages of simplicity, reliability, reduced coupling slippage, and

the use of a minimum number of moving parts.

The low speed of these motors also helps in providing reduced motor vibration. What small vibration there is includes a component that is subsonic in frequency and inaudible as rumble. As has already been pointed out, rumble a t a given level that is very low in frequency has greatly reduced annoyance value.

The outstanding disadvantage of the time-regulation motors that have been used is their low torque, requiring light platens. In order to overcome this disadvantage partially, and to provide greater stability of turntable speed with changes of load (caused by differences of stylus force, and so on) tmo such motors are sometimes used.

The motors of this type currently used

34 AUDIO OCTOBER, 1962

in commercial turntables are rated by the motor manufacturer as having 30 in.-oz. of running torque referred to 1 rpm, which means that the two-motor units have 60 in.-oz. I n the case of the AR turntable a heavier platen is used -the total platen weight is 3.3 1bs.- and more torque was desired.

A third type of motor was investi- geted, and found highly satisfactory for turntable use from several points of view. This is the permanent magnet synchronous motor. The rotor of this synchronous motor, as the name implies, de- rives its magnetic field from permanent magnets. The recent development of ceramic magnets with high retentivity has helped considerably in PM motor design.

adequate for its reproducing application.

One especially desirable characteristic of the permanent magnet synchronous motor is that it does not slow down gradually with increasing load before stalling, as does the hysteresis synchronous motor. Unlike the hysteresis rotor the permanent 111agnet rotor cannot shift poles. It maintains its correct and exact speed within its torque capabilities-it cannot run off-speed unless the line frequency is changed-and then stalls out abruptly.

The PM synchronous motor is inferior to the hysteresis synchronous motor in the smoothness of velocity within one revolution. The filtering requirements of a permanent-magnet-motor drive sys-

Fig. 8. The com- p le te A R turn-

table.

The PM motor cannot work a t any speed except synchronous speed, since its poles are perlnanently fixed, and it has high synchronous torque. The motor selected provides about 150 in.-02. of running torque, even though its physical size is small. Unfortunately this motor has very poor starting torque, and it is not unidirectional, that is, it is as willing to go clockwise as counter- clockwise, depending on the direction in which it is first urged. Mechanical "no- back" devices, which allow rotation in only one direction, tend to be noisy.

The configuration that is used in the AR turntable for the motor system is shown in Fig. 7. The drive motor to the left is of the permanent magnet synchronous type, an 18-pole, 400-rpm unit. The starting motor on the right is the same hysteresis synchronous 600-rpm type currently used in other turntables.

Although the latter motor adds about 30 in.-oz. of running torque to the system, its real function is to start the PM motor, and in the right direction. The total torque of 180 in.-oz., which can be translated to a little over 5 in.-oz. a t the platen itself, is still not tremendous and will not saw wood, but it is entirely

tcm, compared to those for a hysteresis- motor drive system, are therefore greater.

Useful Turntable Torque

I t is possible for a turntable to be driven by a large amount of power, y!t be unable to maintain stable speed m the face of only a slight increase in mechanical load. The test of useful turntable torque is tho force required to slow the platen beyond the NAB limit on speed inaccuracy, not the force required to stop the platen.

The willingness of a turntable to run a little slow with only a slight increase ic mechanical load often bears little re- lationship to the size or power of the motor. This speed inconstancy may be a function of the type rather than the size of the motor, or it may be due to coupling slippage between the motor and platen.

Useful turntable torque cannot be tested by trying to stop the platen by hand. A turntable which can be stopped easily with a finger may keep its speed within the NAB 0.3-per-cent limit in the face of a greater load change than that

tolerated by a turntable which requires much greater effort to stop by hand.

One method of determining useful torque is to see how much additional weight can be added to the pickup, while it is playing the outside grooves of a record, before the speed is changed 0.3 per cent. Such a speed change repre- sents a drift of 21 lines/min. on a standard stroboscope card. The lines must be counted while holding a pencil or some other point near them, as i t is easy to be misled by merely glancing a t the card.

Placing a U. S. penny on the pickup shell adds 3 grams; a nickel adds 5 grams. The significance of this added test load is indicated by the fact that the difference in drag imposed by a pickup when playing the outside grooves of a record, and the drag when the pickup is on the inside grooves, is typically equivalent to about 2 added grams of stylus force. The drag of a "Dust Bug" is equivalent to about 6 extra grams. A heavily recorded passage may slow the turntable to the equivalent of a gram or two extra stylus force. The above data will vary somewhat with the friction of different records and needles, but it gives the general order of equivalent values.

The carrying of a few extra grams by the pickup may seem too light a trial for a turntable, but there are units that will not pass the nickel and/or penny test. The degree of sensitivity to increase of mechanical load of a group of five current commercial turntables, includin~ the AR turntable, was measured an; compared by this kethod. The turntables were not a random sampling of available makes, but were selected to illustrate particular types. Three of the turntables-#1, #2, and the AR turn- ta.ble--kept well within the NAB 0.3 per cent speed limit when the pickup was loaded by an extra 5 grams from an initial 2%-gram needle force. Turn- table #1 had a heavy-duty hysteresis motor and a very heavy platen. Turntable #2 was of the "light" type, in both motor torque and platen weight. Com- pared to these two the AR turntable would be classified as medium-weight.

The other two turntables-#4 and #5-were slowed down more than 0.3 per cent when the pickup was loaded by a 3-gram penny, and were slowed down considerably more when the pickup was loaded by an extra 5 grams. #4 had a heavy motor and platen, and #5 would be classed as a lightweight.

It is of special interest that i t takes some effort to stop turntable #4 by hand, while the AR turntable and turntable #2 can be stopped with relatively light pressure from a finger. Turntable #4 nonetheless has less useful torque than either of the latter. #4 will con-

36 AUDIO OCTOBER, 1962

t ime to revolve, at an unusably slow speed, under mechanical loads which would cause the AR turntable or turntable #2 to stall out.

I t should be clear that the index of useful turntable torque is neither the "heavy" or "light" classification of the turntable, nor the force required to stop the platen by hand.

Speed Accuracy

The AR turntable will slow down several tenths of a per cent if sufficient increase of mechanical load is applied to the pickup in spite of the fact that the permanent magnet synchronous motor is incapable of running a t reduced speed without stalling. Exali~ination of the rotor under conditions of increased mechanical load show that it is still revolving a t exact synchronous speed. The slowing down of the platen is caused entirely by belt slippage.

The accuracy of the diameters of the belt surface of the platen and of the drive pulley are kept to a sufficiently close tolerance that the speed from one turntable to another is always within 0.16 per cent (sixteen hundredths of one per cent). This alone does not ensure speed accuracy, because of the element of belt slippage. The actual turntable speed is partly determined by the friction between the belt and the drive surface.

A calculation of the diameter ratio between the belt surfaces of the pulley and of the platen would show that the pulley diameter is slightly smaller than the exact speed ratio indicates. With the correct calculated value the turntable will run too fast.

I t may seem illogical that belt slippage will cause a turntable to run fast, yet this is what happens in the typical belt-drive system. The drive pulley re- volves in jerks from one motor pole to the next, and the instantaneous speed of the pulley is a t times greater than its av,erage speed. The belt filters out the roughness and picks off an effective speed which turns out to be slightly higher than the average.

Flutter

There appears to be unanimous agree- ment among academic authorities that raw flutter readings (per cent frequency deviation) have limited meaning until they are qualified by information on flutter rates.1° All of the writer's ex- perience in development work on the AR turntable bears out this conclusion.

10 Op. cit., JSMPTE and 257.1-1954.

The statement that turntable X has flutter of 0.1 per cent and that turntable Y has flutter of 0.2 per cent does not reveal which of the two turntables has better flutter performance. If the 0.1 per cent deviation occurs a t a flutter rate of 3 cps, and the 0.2 per cent deviation occurs a t other rates, the turntable with the higher raw flutter may have a lower flutter index. Since its flutter has less nuisance value it obviously deserves a superior flutter rating.

I n spite of the above considerations, turntable performance continues to be described in terms of per cent raw flutter. The meaning of this number is easily understood, but it may have little aural significance by itself.

Although the AR turntable does con- form in flutter performance to the 1953 NAB Standards of Good Engineering Practice for broadcast equipment, which are in terms of raw flutter, (the flutter nleasurelnents yielded results closer to the recording rather than the reproducing figure), the significant production line testing a t the AR plant is per- forlned by a weighted flutter meter whose indications take into consideration the instantaneous flutter rates, and are actually in ternis of an aurally significant flutter index. This flutter index is a unit solnewhat similar to that suggested .by the American Standard11 referred to previously.

Rumble

As in the case of flutter, rumble continues to be described inadequately, in terms of raw amplitude below a given reference level without consideration for the frequency distribution of the rumble energy. The frequencies in the rumble signal may be subsonic, in the 30-cps range, or as high as 120 cps. The difference in perceptibility between 10-cps rumble and 30-cps rumble is tremendous. ( I t is assumed that the rumble amplitude is low enough so that it does not significantly load the amplifier.)

The AR turntable meets the NAB rumble standard, but more significant information in production line testing is provided by a rumble meter which is weighted according to the frequency characteristics of hearing sensitivity.

Both of these testing functions-providing weighted flutter and weighted rumble readings-are performed by a Dataservice Corp. flutter and rumble meter, model FM-3.

Correlation between the test information given by this meter and standard flutter meters is relatively low. Correla-

11 Zbid.

tion with simple listening to a 3000-cps flutter record is excellent. The latter is a particularly good test method when several turntables are being compared, providing information of much greater significance than that of a raw flutter reading. No turntable or professional tape machine to the writer's knowledge, is capable of reproducing a 3000-cps pure tone in a reverberant room in such a way that flutter cannot be readily heard, but it is relatively easy to com- pare the flutter effect of different reproducing machines. I t should be re- membered that very slow flutter corresponding to one change per revolution at 33y3 rpm (0.55 cps) is much less an- noying in music than flutter whose rate centers around 3 cps.

The flutter index in the AR turntable is kept to low values through the use of the simplest possible one-step drive system, mounting of the tone arm for mini- niunl warp wow, ~nachined drive pulley and platen, machined bearings, and pos- sibly most important of all, a precision- iliachined belt (machined after freez- ing). Most of these features are conlmon in turntables in which professional performance standards are sought.

Correlation between the weighted rumble information provided by the DSC meter and NAB ratings is also low. Correlation with listening tests in which different turntables are compared-using the same "quiet groove" record, pickup, and reproducing equiplllent-is excellent. Two turntables with the same NAB rumble rating may be lniles apart on a runible listening test. The weighted rumble in the AR turntable is kept to low values through the use of low-speed ~iiotors, compliant belt drive, isolation of the rigid platen-arm assembly from the motors, and a ir~achined platen bearing.

Safety and Convenience

The AR turntable is supplied with its wooden base and transparent plastic dust cover included. If this turntable were sold without the base, Underwrit- er's Laboratories approval, which has been applied for, would not be available. UL approval requires that the moving parts and electrical terminals be safe from prying fingers. The individual electrical components used in the AR turntable, from motors to power cord, already have separate UL approval.

A dust cover seems a logical and in- trinsic part of a record player. The convenience from the housekeeping point of view is obvious, but more important, the turntable mat is protected from dust which can transfer to the record. Fig. 8 illustrates the complete unit. E

AUDIO OCTOBER, 1962

a new turntable arm design

Documents