Appendix 2
Record players
Record players are obsolete technology, but it is felt that information should remain in the literature describing how to get the best out of the format, not least because inadequate or inept setting up of equipment can cause permanent damage to records, and because it seems likely that the LP will persist for some time yet. Also, there is still valuable material to be found here which is not always available on CD.
The replay stylus motion should describe an arc offset from the vertical by 20°, as shown in Figure A2.1. This will be achieved if the arm height at the pivot is adjusted such that the arm tube is parallel to the surface of the record when the stylus is resting in the groove. The stylus tip should have a cone angle of 55°, as shown in Figure A2.2. The point is rounded such that the tip makes no contact with the bottom of the groove. Stylus geometry is discussed further in Fact File A2.1.
The arm geometry is arranged so that a line drawn through the cartridge body, front to back, forms a tangent to the record groove at a point where the stylus rests in the groove, at two points across the surface of the record: the outer groove and the inner position just before the lead-out groove begins. Figure A2.3 illustrates this. Note that the arm tube is bent to achieve the correct geometry. Alternatively, the arm tube can be straight with the cartridge headshell set at an offset angle which achieves the same result. The arc drawn between the two stylus positions shows the horizontal path of the stylus as it plays the record. Due to the fact that the arm has a fixed pivot, it is not possible for the stylus to be exactly tangential to the groove throughout its entire travel across the record’s surface, but setting up the arm to meet this ideal at the two positions shown gives a good compromise, and a correctly designed and installed arm can give less than ±1° tracking error throughout the whole of the playing surface of the disc.
Figure A2.1 Stylus vertical tracking geometry
Figure A2.2 Stylus cone angle
Alignment protractors are available which facilitate the correct setting up of the arm. These take the form of a rectangular piece of card with a hole towards one end which fits over the centre spindle of the turntable when it is stationary. It has a series of parallel lines marked on it (tangential to the record grooves) and two points corresponding to the outer and inner groove extremes. The stylus is lowered on to these two points in turn and the cartridge and arm are set up so that a line drawn through the cartridge from front to back is parallel to the lines on the protractor.
Fact File A2.1 Stylus profile
Two basic cross-sectional shapes exist for a replay stylus – conical and elliptical, as shown in the diagram. The elliptical profile can be seen to have a smaller contact area with the wall of the groove, and this means that for a given tracking weight (the downforce exerted by the arm on to the record surface) the elliptical profile exerts more force per unit area than does the conical tip. To compensate, elliptical styli have a specified tracking force which is less than that for a conical tip. The smaller contact area of the elliptical tip enables it to track the small, high-frequency components of the signal in the groove walls, which have short wavelengths, more faithfully. This is particularly advantageous towards the end of the side of the record where the groove length per revolution is shorter and therefore the recorded wavelength is shorter for a given frequency. Virtually all high-quality styli have an elliptical profile or esoteric variation of it, although there are still one or two high-quality conical designs around. The cutting stylus is, however, always conical.
Figure A2.3 Ideal lateral tracking is achieved when a line through the head-shell forms a tangent to the groove
The original cutting stylus is driven across the acetate in a straight line towards the centre, using a carriage which does not have a single pivot point like the replay arm, and it can therefore be exactly tangential to the groove all the way across the disc. The cutting lathe is massively engineered to provide an inert, stable platform. There are some designs of record player which mimic this action so that truly zero tracking error is achieved on replay. The engineering difficulties involved in implementing such a technique are probably not justified since a well-designed and well-set-up arm can achieve excellent results using just a single conventional pivot.
A consequence of the pivoted arm is that a side thrust is exerted on the stylus during play which tends to cause it to skate across the surface of the record. This is a simple consequence of the necessary stylus overhang in achieving low tracking error from an arm which is pivoted at one end. Consider Figure A2.4. Initially it can be considered that the record is not rotating and the stylus simply rests in the groove. Consider now what happens when the record rotates in its clockwise direction. The stylus has an immediate tendency to drag across the surface of the record in the arrowed direction towards the pivot rather than along the record groove. The net effect is that the stylus feels a force in a direction towards the centre of the record causing it to bear harder on the inner wall of the groove than on the outer wall. One stereo channel will therefore be tracked more securely than the other, and uneven wear of the groove and stylus will also result. To overcome this, a system of bias compensation or ‘anti-skating’ is employed at the pivot end of the arm which is arranged so that a small outward force is exerted on the arm to counteract its natural inward tendency. This can be implemented in a variety of ways, including a system of magnets; or a small weight and thread led over a pulley which is contrived so as to pull the arm outwards away from the centre of the record; or a system of very light springs. The degree of force which is needed for this bias compensation varies with different stylus tracking forces but it is in the order of one-tenth of that value (see Fact File A2.2).
Figure A2.4 The rotation of the disc can create a force which pulls the arm towards the centre of the disc
Although every cartridge will fit into every headshell apart from one or two special types, one must be aware of certain specifications of both the arm and the cartridge in order to determine whether the two are compatible. In order for the stylus to move about in the groove, the cantilever must be mounted in a suitable suspension system so that it can move to and fro with respect to the stationary cartridge body. This suspension has compliance or springiness and is traditionally specified in (cm/dyne) × 10−6, abbreviated to cu (‘compliance units’). This is a measure of how many centimetres (in practice, fractions of a centimetre!) the stylus will deflect when a force of 1 dyne is exerted on it. A low-compliance cartridge will have a compliance of, say, 8 cu. Highest compliances reach as much as 45 cu. Generally, values of 10–30 cu are encountered, the value being given in the maker’s specification.
Fact File A2.2 Tracking weight
The required weight varies from cartridge to cartridge. A small range of values will be quoted by the manufacturer such as ‘1 gram ± 0.25 grams’ or ‘1–2 grams’ and the exact force must be determined by experiment in conjunction with a test record. Firstly, the arm and cartridge must be exactly balanced out so that the arm floats in free air without the stylus moving either down towards the record surface or upwards away from it, i.e.: zero tracking force. This is generally achieved by moving the counterweight on the end of the arm opposite to the cartridge either closer to or further away from the pivot until an exact balance point is found. The counterweight is usually moved by rotating it along a thread about the arm, or alternatively a separate secondary weight is moved. This should be carried out with bias compensation off.
When a balance has been achieved, a tracking weight should be set to a value in the middle of the cartridge manufacturer’s values. Either the arm itself will have a calibrated tracking force scale, or a separate stylus balance must be used. The bias compensation should then be set at the appropriate value, which again will have either a scaling on the arm itself or an indication in the setting up instructions. A good way to set the bias initially is to lower the stylus towards the play-in groove of a rotating record such that the stylus initially lands mid-way between these widely spaced grooves on an unused part of the surface of the record. Too much bias will cause the arm to move outwards before dropping into the groove. Too little bias will cause the arm to move towards the centre of the record before dropping into the groove. Just the right amount will leave the arm stationary until the relative movement of the groove itself eventually engages the stylus. From there, the optimum tracking and bias forces can then be determined using the test record according to the instructions given. In general, a higher tracking force gives more secure tracking but increases record wear. Too light a tracking force, though, will cause mistracking and damage to the record grooves.
The record groove is an analogue of the sound waves generated by the original sources, and this in itself caused early pioneers serious problems. In early electrical cutting equipment the cutter stylus velocity remained roughly constant with frequency, for a constant input voltage (corresponding to a falling amplitude response with frequency) except at extreme LF where it became of more constant amplitude. Thus, unequalised, low frequencies would cause stylus movements of considerably greater excursion per cycle for a given stylus velocity than at high frequencies. It would be difficult for a pickup stylus and its suspension system inside the cartridge body to handle these relatively large movements, and additionally low frequencies would take up relatively more playing surface or ‘land’ on the record curtailing the maximum playing time. Low-frequency attenuation was therefore used during cutting to restrict stylus excursions.
In modern record cutting, a standard known as RIAA equalisation has been adopted which dictates a recorded velocity response, no matter what the characteristics of the individual cutting head. Electrical equalisation is used to ensure that the recorded velocity corresponds to the curve shown in Figure A2.5(a). A magnetic replay cartridge will have an output voltage proportional to stylus velocity (its unequalised output would rise with frequency for a constant amplitude groove) and thus its output must be electrically equalised according to the curve shown in Figure A2.5(b) in order to obtain a flat voltage–frequency response.
Figure A2.5 RIAA recording and reproducing characteristics
The treble pre- and de-emphasis of the RIAA replay curve have the effect of reducing HF surface noise. An additional recommendation is very low 20 Hz bass cut on replay (time constant 7960 μs) to filter out subsonic rumble and non-programme-related LF disturbance. The cartridge needs to be plugged into an input designed for this specific purpose; the circuitry will perform the above discussed replay equalisation as well as amplification.
The vast majority of cartridges in use are of the moving-magnet type, meaning that the cantilever has small powerful magnets attached which are in close proximity to the output coils. When the stylus moves the cantilever to and fro, the moving magnets induce current in the coils to generate the output. The DC resistance of the coils tends to be several hundred ohms, and the inductance several hundred millihenries (mH). The output impedance is therefore ‘medium’, and rises with frequency due to the inductance. The electrical output level depends upon the velocity with which the stylus moves, and thus for a groove cut with constant deviation the output of the cartridge would rise with frequency at 6 dB per octave. The velocity of the stylus movements relative to an unmodulated groove is conveniently measured in cm s−1, and typical output levels of moving magnet cartridges are in the order of 1 mVcm−1s−1.
The average music programme produces cartridge outputs of several millivolts, and an upper limit of 40 or 50 mV will occasionally be encountered at mid frequencies. Due to the RIAA recording curve, the output will be less at low frequencies but not necessarily all that much more at high frequencies owing to the falling power content of music with rising frequency. A standard input impedance of an RIAA input of 47 k has been adopted, and around 40 dB of gain (× 100) is needed at mid frequencies to bring the signal up to line level.
Another type of cartridge which is much less often encountered but has a strong presence in high-quality audio circles is the moving-coil cartridge. Here, the cantilever is attached to the coils rather than the magnets, the latter being stationary inside the cartridge body. These cartridges tend to give much lower outputs than their moving-magnet counterparts because of the need to keep coil mass low by using a small number of turns, and they have a very low output impedance (a few ohms up to around a hundred) and negligible inductance. They require 20–30 dB more gain than moving magnets do, and this is often provided by a separate head amplifier or step-up transformer, although many high-quality hi-fi amplifiers provide a moving-coil input facility. The impedance of such inputs is around 100 ohms or so.
Owing to the inductive nature of the output impedance of a moving-magnet cartridge, it is sensitive to the capacitance present in the connecting leads and also that present in the amplifier input itself. This total capacitance appears effectively in parallel with the cartridge output, and thus forms a resonant circuit with the cartridge’s inductance. It is the high-frequency performance of the cartridge which is affected by this mechanism, and the total capacitance must be adjusted so as to give the best performance. Too little capacitance causes a frequency response which tends to droop several decibels above about 5 kHz, sharply rising again to a 2–3 dB peak with respect to the mid band at around 18–20 kHz, which is the result of the resonant frequency of the stylus/record interface, the exact value depending upon the stylus tip mass. Adding some more capacitance lifts the 5–10 kHz trough and also curtails the tip mass resonant peak to smooth out the frequency response. Too much capacitance causes attenuation of the highest frequencies giving a dull sound.
Around 300–400 pF total is the usual range of capacitance to be tried. Adding capacitance is most conveniently carried out using special in-line plugs containing small capacitors for this purpose. Assume that 200 pF is present already, and try an extra 100 pF, then 200 pF. Alternatively, small polystyrene capacitors can be purchased and soldered between signal and earth wires of the leads inside the plugs or sockets. Sometimes solder tags are present in the base of the record player itself which are convenient. NEVER solder anything to the cartridge pins. The cartridge can very easily be damaged by doing this.
Moving-coil cartridges have a very low output impedance and negligible inductance, and their frequency response is therefore not affected by capacitance.
The effective mass of the pickup arm (Figure A2.6), which is the inertial mass of the arm felt by the stylus, coupled with the cartridge’s suspension compliance, together form a resonant system, the frequency of which must be contrived such that it is low enough not to fall within the audio band but high enough to avoid coinciding with record warp frequencies and other LF disturbances which would continually excite the resonance causing insecure tracking and even groove jumping. Occasionally, large, slow excursions of the cones of the speaker woofers can be observed when the record is being played, which is the result of non-programme-related, very low-frequency output which results from an ill-matched arm/cartridge combination.
A value of 10–12 Hz is suitable, and a simple formula exists which enables the frequency to be calculated for a given combination of arm and cartridge:
f = 1000/(2π√(MC))
where f = resonant frequency in hertz, M = effective mass of the arm + mass of the cartridge + mass of hardware (nuts, bolts, washers) in grams, C = compliance of the cartridge in compliance units.
For example, consider a cartridge weighing 6 g, having a compliance of 25 cu; and an arm of effective mass 20 g, additional hardware a further 1 g. The resonant frequency will therefore be 6.2 Hz. This value is below the optimum, and such a combination could give an unsuitable performance due to this resonance being excited by mechanical vibrations such as people walking across the floor, record warps, and vibrations emanating from the turntable main bearing. Additionally, the ‘soft’ compliance of the cartridge will have difficulty in coping with the high effective mass of the arm, and the stylus will be continually changing its position in the groove somewhat as the arm’s high inertia tends to flex the cartridge’s suspension and dominate its performance.
Figure A2.6 The SME Series V pickup arm. (Reprinted with permission of the Society of Manufacturing Engineers, USA)
If the same cartridge in an arm having an effective mass of 8 g is considered, then f = 8.4 Hz. This is quite close to the ideal, and would be acceptable. It illustrates well the need for low-mass arms when high compliances are encountered. The resonance tends to be high Q, and this sharp resonance underlines the need to get the frequency into the optimum range. Several arms provide damping in the form of a paddle, attached to the arm, which moves in a viscous fluid, or some alternative arrangement. This tends to reduce the amplitude of the resonance somewhat, which helps to stabilise the performance. Damping cannot, however, be used to overcome the effects of a non-optimum resonant frequency, which must still be carefully chosen.
The idea of reading a record groove with a laser beam rather than a stylus has been mooted for quite some time, and in 1990 a player using such a technique finally appeared. It is a very attractive proposition for record libraries due to the fact that record wear becomes a thing of the past. However, a laser beam does not push particles of dust aside as does a stylus, and the commercial system needs to be fed with discs which are almost surgically clean, otherwise signal drop-outs occur. Two entirely separate laser beams are in fact used, one reading the information from each wall of the groove. Error concealment circuitry is built in, which suppresses the effects of scratches. Towards the centre of the record, short wavelengths occupy a proportionately smaller area of the groove than at the perimeter of the disc, and the width of the laser beam means that difficulties in reading these high-amplitude HF signals occur. The frequency response of the player therefore droops by around 10 dB if the highest frequencies towards the end of the side are of a high amplitude.
CD-type features are offered such as pause, track repeat and track search, which are very useful. The player does not suffer from traditional record player ills such as LF arm/cartridge resonance, rumble, and wow and flutter. Costing not much under £10 000, the player obviously has a limited appeal, the professional user being the main potential customer.
AES (1981) Disk Recording – An Anthology, Vols 1 and 2. Audio Engineering Society
BS 7063. British Standards Office
Earl, J. (1973) Pickups and Loudspeakers. Fountain Press
Roys, H. E. (1978) ed. Disk Recording and Reproduction. Dowden, Hutchinson and Ross