Magnetic sound recording

The first complete magnetic recorder (wire on a drum) was constructed in 1898 by Poulsen (Denmark). In 1938 the first recorders using plastic tape with an iron oxide layer were used. Since then, magnetic sound recording has become the most popular way of storing broadcast programmes for transmission at any time. Early attempts with temporary disc recording and photographic recording proved unsuitable for broadcast applications.

Although digital recording systems have gained increasing importance, the analogue magnetic recording is still in wide use and will probably remain so for several more years.

Mechanical, magnetic and electronic principles are involved in tape recording. All of them have to be studied to fully understand the entire system.

During maintenance, the mechanical alignments are always done first. Therefore also this course will start with the consideration of tape recorders' mechanics.

Traditionally, synchronous ac motors were used. They were "locked" to the mains frequency. The motor windings can be switched to provide two tape speeds. The motor may have the armature outside or be of standard design with an additional flywheel on the shaft below the motor body. The shaft of the motor extends upwards as capstan.

Disadvantages:

Other makes (e.g. Studer) use an asynchronous ac motor, which is then speed controlled. In all such systems, a generator (or 'tachometer') is attached to the capstan. The frequency from this generator is proportional to the capstan speed. This frequency is checked against a standard (reference frequency). If the two frequencies are not equal, the capstan speed is re-adjusted. In such a closed loop system (servo system) the speed is checked and adjusted all the time, so maintaining it correct and constant.

Such systems make the speed externally adjustable if required and independent of the mains frequency.

The capstan drive may be direct, e.g. as a ac motor (as in Studer models), or as a DC motor (as in Nagra); or indirect via belt drive with a DC motor (as in Telefunken M15 and M15A).

Capstan speed control circuits are always based on the same principle: A FREQUENCY is derived from the capstan speed. This frequency (f_tacho) is COMPARED with a fixed, precisely known REFERENCE FREQUENCY (f_ref). The circuit, which compares these two frequencies, produces a signal, which indicates the difference between the capstan frequency and the reference frequency. This is called the ERROR SIGNAL. The error signal controls the SERVO AMPLIFIER for the capstan motor. The whole system represents a SERVO SYSTEM, which is actually a closed loop NFB system.

The principle of the capstan servo system.

To compare the capstan and the reference frequency, two different principles must be distinguished:
- the phase comparator and
- the frequency comparator.

The tachometer system has to provide the information of the actual capstan speed to the servo system. In all cases the system generates a signal, of which the FREQUENCY carries the information.

Different ways are used to generate this tachometer frequency. In most cases the signal is produced by induction, so a coil is involved. The "armature" has a number of poles, which defines, together with the capstan speed, the frequency of the generated signal. The poles may consist of slots or teeth on a disc mounted on the capstan shaft (e.g. Nagra), or on the outside of the motor (e.g. Studer).

The teeth or slots on the rim are cut to close tolerances. Also, the wheel must run without any eccentricity, otherwise frequency variations will be produced. The tachometer-head is DC-magnetized. The teeth passing produce field changes, the field changes induce a voltage of a frequency, representing the capstan speed (f_tacho).

Nagra recorders use a toothed flywheel with a single tachometer head.

To make the frequency of the tachometer output less dependent on the eccentricity, one can employ two tachometer heads.

A system using two tachometer heads 180° apart. Head no.l can be displaced a bit for optimum performance. The potentiometer is set for equal voltages from the heads. (Used in Studer A80, A81).

An alternative method for the tacho generator:
the capstan (motor) turns the inner ring against an outer fixed ring, both with equal number of teeth. The outer ring includes a coil, which is DC current magnetized and acts as pick-up (tachometer ring). (Used in Studer A7000, B67 and Telefunken M15, M15A.)

Phase comparators produce an output signal which depends on the PHASE RELATIONSHIP between the input frequency and a reference frequency. This means in fact, that the two frequencies must be on average absolutely equal. A servo system, using phase comparators normally maintains a phase difference of approximately 90°.

If such a system is used in oscillators, it is called a PHASE LOCKED LOOP (PLL) system.

The principle of the phase comparator.

Different circuits can be used as a phase comparator. We mainly distinguish analog and digital principles.

The circuit used as the phase comparator is the same as for the modulator, demodulator or mixer.

The analog phase comparator using a circuit known as phase discriminator.

This circuit produces an output signal, which is known from the modulation:

In the phase comparator the second term (f₁-f₂) is used. The first term (f₁+f₂) is a relatively high frequency, which can be eliminated by filtering. When both frequencies are equal, the term (f₁-f₂) will become 0. Thus the output signal is a DC voltage. The magnitude of this voltage depends on the phase relationship between f₁ and f₂.

When the two frequencies are different, an ac signal is produced at the output.

The typical output signal of a phase comparator as function of the phase difference between f₁ and f₂.

The phase discriminator produces an ac signal as long as the two frequencies are not yet synchronized. This represents no useful error signal. Therefore, in practice additional means are required to produce a useful signal during the speeding up of the capstan motor.

Different principles are used here. There are also ICs specially available for servo systems.

The principle will be explained here by using an ordinary EXOR gate:

The simplest form of a digital phase comparator can be an EXOR-gate.

The relationship between the input signals and the output signal is shown in the following diagram:

The output signal of an EXOR-gate fed with two signals of different phase.

Only the part between 0° and 180° can be used. Also, this phase comparator is normally operated near 90°.

The disadvantage of this simple circuit is that it does not provide a useful signal, while the capstan motor does not yet run at synchronized speed. Therefore, in practice, more complex integrated circuits or special ICs for servo controls are used.

These circuits produce a signal which depends on the frequency of the input signal. The input signal will therefore be the tacho-frequency, while the output voltage will be used as the error signal.

Block diagram of the frequency voltage converter.

The frequency voltage converter therefore receives only one input frequency. The reference frequency is represented by some circuit elements of the converter. The reference frequency does not physically appear in the circuit.

The frequency voltage converter will only produce some variation of its output voltage if there is a variation of the input frequency. Therefore, a servo system, using a frequency voltage converter, will allow small variations of the capstan speed, while a system using a phase comparator is able to maintain the capstan speed absolutely constant.

Again, analog and digital solutions are possible for the frequency comparator.

The circuit can be understood as a phase comparator with a tuned circuit, which can be considered as the reference frequency source.

Principle circuit of an analog frequency voltage converter.

In this circuit L and C represent the reference frequency.

The circuit has the following characteristic:

Relationship between input frequency and output voltage for the frequency voltage converter.

This principle is used in the Nagra 3 and Telefunken M15.

These circuits use a monoflop, producing a pulse of defined length.

The length of the pulse represents the reference to the circuit. Its accuracy will define the accuracy of the capstan speed.

Block diagram of a monoflop and time diagrams for an input signal of increasing frequency.

The timer IC 555 can be used for the following application:

Monoflop with timer-IC 555. The circuit differentiates the input signal to produce short pulses, and integrates the output voltage to produce a pure DC signal.

This method is used in the Studer A77, B77 and Nagra 4.

As most tape machines run at different tape speeds, the capstan speed must be controllable. In principle there are two ways to achieve this:

For vari-speed applications these parameters are manipulated continiously. Method 1 is not suitable for this variation. In this case, the vari-speed action uses method 2, while the standard capstan speeds are selected by method 1.

In most modern tape machines DC motors are used.
This gives the following advantages:

- Tape speed independent of main frequency,
- Battery operation possible,
- High accuracy achievable with servo systems,
- Easy variation of tape speed,
- Good efficiency,
- Low running noise,
- Low magnetic stray fields.

In order to produce a rotating movement, commutation is required to convert the DC into a current, which changes its polarity according to the position of the rotor.

Traditionally, this is done by using a commutator ring on the rotor and stationary brushes.

The principle of the motor: the current through the rotating coil produces a mechanical force in the stationary magnetic field. In order to continue the rotation, the current in the coil has to change the direction if the rotor field is in line with the stator field.

Commutators on the rotor supply the rotor coils with a constantly changing current, which provides a constant force on the rotor. The current is provided to the rotor by brushes.

The mechanical commutation of DC-motors has the following disadvantages:

Because DC motors with commutators are cheap, they are, despite there disadvantages, today still widely used in simple cassette recorders.

Electronic commutation avoids all the disadvantages of the mechanical commutation. It is therefore entirely used in all modern professional tape machines and in more sophisticated consumer equipment.

Electronic commutation makes use of the hall generator to detect the position of the rotor. It also controls the stator current via a current amplifier, so that a rotating magnetic field is achieved in the stator windings.

Principle of action:

The rotor is a permanent magnet with one or many pairs of poles. A hall generator is mounted to the stator to detect the position of the rotor. The signal from the hall generator controls a transistor amplifier, which passes current through the stator winding so that a constant torque is achieved to the rotor. If the rotor rotates at constant speed, the current through the stator winding will have a sine shape and will have a frequency which is synchronous to the speed of rotation of the rotor.

The principle of the burshless DC motor and the commutating circuit.

There are many different arrangements for the rotors, stators hall commutators and electronic circuits of such motors. Some of them have the commutation circuit integrated in the motors (Ampex 440), some have it externally (EMT, STUDER, Telefunken). Some use the commutation transistors directly as servo amplifiers (EMT, Studer A807).

All these motors have the same advantages:

Very stringent demands are put on the mechanical behaviour of a professional tape machine:

The pressure roller is used to press the tape to the capstan to force it to run at the circumferential speed of the capstan. The pressure roller must meet the capstan at such an angle, that the tape will first touch the capstan and later the pressure roller. This is necessary to avoid eccentricity of the pressure roller (rubber) affecting the tape speed.

The arrangement of pinch roller and capstan.
The tape, coming from the heads, must first meet the capstan, then the pressure roller.

The pressure roller is pressed against the capstan by a spring- loaded solenoid.The pressure is adjustable, the correct value is quoted in the service manual. Capstan and roller tend to collect tape particles and dirt. Both must be cleaned regularly (e.g. daily) with alcohol.

Special care is essential when lubricating: remove all traces of oil from roller and capstan.

The capstan must be vertical to the tape, the pressure roller must run true parallel with the capstan. In some machines this is adjustable. The pressure roller position can be checked by watching the tape to run off straight from the capstan without any tendency to twist upwards or downwards. Mis-adjustment will cause the tape to become curled and damaged.

If the axis of the capstan and pressure roller are not parallel, the tape will be twisted and pulled out of track.

When pressure and/or alignment are not correct, TAPE SLIP will increase. This will also increase the WOW AND FLUTTER and can be checked with suitable instruments. Note that slip and wow and flutter also depend on other mechanical parameters of the tape machine like dirt, tape tension and the alignment of the brakes.

The capstan (with its motor and/or flywheel) is precision engineered to very close tolerances.

Take care not to knock it. A damaged capstan has to be replaced.

Magnetic sound recording