The fundamental purpose of practical microphones is to convert sound to electrical signals. In the most common types, the sound vaves cause a diaphragm to vibrate, this vibration is converted by transduction into electrical signals

The way in which a microphone responds to sounds coming from different directions is its 'directional response'. The factor which determines the directional response of a microphone is the way the diaphragm is exposed to sound. There are two basic types and a third which is derived by combining them.

In each type, the diaphragm responds to the difference in pressure between its two sides

'PRESSURE' RESPONSE (Omnidirectional)

The diaphragm is exposed to the open air on one side only, the air on the other side being enclosed by a rigid airtight structure.

As the enclosed side is unaffected by the sound, it is the air pressure on the outside which determines the activating force. If the microphone is small compared with the wavelength of the sounds being received, it will not obstruct the pressure waves coming from the sides or back of the casing and so the response will be the same in all directions.

 An air space which is small enough to have a resonance above the audio frequencies of interest leads to a 'stiff' diaphragm which responds to pressure with a flat frequency response.

'PRESSURE-GRADIENT' RESPONSE (Bipolar)

Any change of sound pressure with distance will give rise to a 'pressure gradient'. The diaphragm is suspended freely so that the air has access to both sides, but the sound wave approaching from one side has to travel an extra distance to reach the other side. The pressure gradient causes a pressure difference between the two sides of the diaphragm and this difference produces the operating force.

There are two pressure gradients which produce a pressure difference:

PHASE GRADIENT
As a wave of the sound passes the diaphragm, it will reach the front of the diaphragm first - then, after travelling around the edge of the diaphragm and any supporting structures, it will reach the back. The difference in pressure at different points on the waveform supplies the difference in pressure across the diaphragm. As the frequency becomes higher, the wavelength becomes shorter and the constant distance from one side of the diaphagram to the other is, in effect, a greater proportion of the shortening waveform. The pressure difference therefore increases with increasing frequency (up to the point where the path difference approaches a half-wavelength).

The resonance of the loosely-suspended diaphragm is below the audio frequencies of interest and the mass of the diaphragm needs a pressure difference which increases with frequency in order to obtain a flat frequency response.

SPHERICAL EXPANSION GRADIENT
When the sound source is compact and close to the microphone, the wavefront is expanding in a spherical fashion. There is a pressure drop due to expansion and, once again, this appears as a pressure difference between the two sides of the diaphragm. In this case, the pressure difference is not wavelength dependent and does not increase with rising frequency - but neither does it decrease with falling frequency. At low frequencies, this effect will predominate over the phase gradient and the diaphragn will respond with increasing amplitude as the frequency falls. Therefore, speaking too close to a velocity microphone results in a bass-heavy sound, 'bass tip-up'.

In either case, if the sound is impinging at right angles to either of the flat sides of the diaphragm, the response will be maximal. If the sound impinges on the edge, there will be no response at all because the path lengths to each side of the diaphragm are identical. The response pattern is therefore bipolar with a sharp null in the plane of the diaphragm. The polarity of the electrical output will depend on which side of the diaphragm the sound is approaching from.

COMBINED RESPONSE (Cardioid)

By combining the pressure-gradient and pressure responses in equal proportions, a sound on one side of the combined microphones will give double the individual outputs. If it comes from the opposite side, the output of the pressure unit will remain unchanged but the output of the pressure-gradient unit will reverse; the result being cancellation and no net output.

This is a directional response which, when plotted as a polar diagram, is 'heart shaped' and is called 'cardioid'.

By varying the relative strengths of the two signals, other polar responses can be obtained but these are less generally useful.

Rather than having two microphones and combining their outputs electrically, one diaphragm can be made directional by allowing limited access to the air on one side and unlimited on the other. Bass tip-up does occur, although to a lesser extent than with the full bipolar response.

Transduction is the process for converting vibration of the diaphragm into electrical signals. Three different electrical effects, capacitance change, electromagnetism and the piezo-electric effect, are commonly used for this and there are several different physical arrangements for making use of them.

CAPACITANCE CHANGE TRANSDUCTION

Capacitor (or condenser) mics use the principle that the moving diaphragm of the microphone is made as one plate of a capacitor and a fixed back-plate forms the other. When sound vibration moves the diaphragm, it causes variations in the spacing between the plates and, hence, the capacitance.

There are at two practical methods of converting that capacitance change into electrical signals.

1) FREQUENCY CHANGE

The capacitance is made to be part of a circuit which determines the frequency of an oscillator. Changes in capacitance modulate this frequency which is converted to a varying audio signal by a detector ('discriminator') circuit.

This gives the lowest background noise of all the methods, it is expensive and only used in absolute top-quality mics ~ £800+

It is vulnerable to distortion, in extremes of temperature, when the oscillator and detector drift by different amounts and get out of line with each other.

2) VOLTAGE CHANGE

A fixed charge is applied to the plates. When capacitance varies, if the charge on the plates cannot escape, the voltage between the plates is obliged to change. There have been big improvements in the ways of applying the charge and measuring the changes of voltage since the principle was first used.

When high voltages were readily available from valve circuitry, a supply of about 300 volts would be connected through a very high value resistor ~ 1,000,000,000 ohms or more. The very high impedance of the grid circuit of a specially-made cathode-follower valve then picked up the voltage variations. The valve had to sit inside the microphone head to avoid problems caused by even relatively short wiring.

This was no longer as easy to impliment when low voltage transistor equipment came into use, so an alternative method of supplying and retaining charge was sought. The electrostatic equivalent of the permanent magnet is the 'electret', a kind of permanent charge. This made the high voltage supply unnecessary, just a circle of pre-charged plastic film was needed. A Field Effect Transistor, with virtually infinite input impedance and low output impedance complets the job and a 1.5 volt battery is all that is needed to power it.

ELECTROMAGNETIC TRANSDUCTION

There are two ways in which microphones are commonly constructed to use the electromagnetic effects for transduction

1) MOVINC COIL

The diaphragm carries a lightweight coil of wire which is arranged in a strong magnetic field. When it moves, it cuts the flux and a voltage is induced in it. In physical form it is like a loudspeaker, but works in reverse. By varying the thickness and number of turns of wire, the impedance can be chosen (within limits) to suit a particular application.

2) RIBBON

The diaphragm is reduced to a thin strip of lightweight aluminium ribbon suspended between (and edge-on to) the poles of a powerful magnet. When it vibrates, it cuts the lines of magnetic flux and a voltage is induced between its ends. The voltage is very small but the source impedance is extremely low allowing the use of a transformer (usually inbuilt) to obtain the required voltage and output impedance.

PIEZO-ELECTRIC TRANSDUCTION

The diaphragm is connected to a crystal of material which generates an electrical voltage when it is stressed. It has a high output, but at fairly high impedance. Most commercial examples were not intended for quality work; they purported to have an omnidirectional response but this was open to doubt.

The high-voltage CAPACITOR microphone can provide omnidirectional, bidirectional or cardioid responses, according to construction. Some very expensive studio types could be electrically remote-controlled to provide any of these, with a smooth transition between them.

The first ELECTRET CAPACITOR microphones on the market were omnidirectional, but cardioid response units are now dominant.

The RIBBON is usually a bidirectional microphone (but there was one version produced which obtained a cardioid response by partially enclosing one side).

The MOVING COIL has been available as either omnidirectional or cardioid for many years (some cheap ones could turn out to be either, depending on frequency!). It was the 'work horse' of the industry until recently, now the electret has taken over.

The CRYSTAL was a convenient and cheap 'consumer' quality microphone, but it is rarely used nowadays because it has acquired a reputation for poor frequency response and unhelpful directional properties.

Impedance is the relationship of signal voltage to signal current.

Voltage and current are a measure of the power of the signal being transferred from the from microphone to the amplifier and, in the days when good quality amplification equipment was large and expensive, it was important to transfer as much of the signal power as possible, keeping price and size within limits as well as optimising the signal-to-noise ratio.

To maximise power transfer, the output impedance of the microphone needs to be the same as the input impedance of the amplifier. With improved amplification and more sensitive microphones this is no longer as important. Provided certain compromises can be tolerated, even a considerable mis-match may be found to give acceptable results.

In the past, a standard impedance was proposed to ensure compatibility. Unfortunately the conflicting professional and commercial requirements resulted many different 'standards' and a return to confusion.

Nowadays 600 ohms is the nominal impedance of many professional microphone outputs and amplifier inputs and is a satisfactory standard to work to.

Impedance is an inherent property of the particular transduction process used in a microphone, but it does not follow that this will be the output impedance at the plug, because it may have been modified by additional components.

The CAPACITOR and ELECTRET transducers are very high impedance units: 1,000,000,000ohms or more. This is impractical to send down even a few inches of cable, so an impedance matching amplifier is built into the microphone head. The resulting output impedance is usually in the range 300 ohms to 1000 ohms.

The MOVING COIL transducer can be wound to give impedances in the range 50 to 1000 ohms. The sensitivity is moderate and, for best results, impedance matching should still be employed. Some capsules give a distorted frequency response if mis-matched.

Sometimes a transformer is incorporated to allow matching to different amplifier impednces.

RIBBON units are very low impedance indeed: a fraction of an ohm. A transformer is always incorporated as part of the assembly and usually matches to the 30 - 600 ohm range.

PIEZO-ELECTRIC 'crystal' microphones (and gramophone pickups) are about 1,000,000 ohms. It is possible to use short lengths of cable for these but signal loss due to cable capacitance, noise generated by movement of the cable itself and, sometimes, unintentional pickup of radio signals and mains hum limit the practicality of this system to a few feet. The popularity of these microphones in the early days of home tape recorders was mainly due to the relatively large output signal and the ease of matching it directly into the high impedance of a valve input stage.

TRANSFORMERS can be used to match microphones of low (50 ohms) impedance to medium (600 ohms) impedance and vice-versa. They are also available for low or medium impedance to high impedance matching, but rarely the reverse. Unless well-designed, and hence expensive, transformers are prone to hum pickup, distortion and other vices.

They do have the virtue, when used in the input of the amplifying equipment, of allowing complete electrical isolation between the microphone signal wires and the internal 'earth' of the equipment. This avoids any possibility of 'hum loops' and other sources of unwanted input signals which bedevil amateurishly designed consumer equipment.

Very often a cheap microphone, around £10 - £15, can give excellent results if you don't mind the fragility. Sometimes, if the casing has been shaped so that it doesn't interfere with the sound, it is possible to change the electret capsule and get studio quality for an extra £2.50. (This is usually a matter of chance because cheap mics are designed to look good, not to work well). Sometime the supplied capsule is already a high enough quality one and may not need changing.

If the casing is large enough to obstruct the sound, the effect will vary according to wavelength and a very ragged frequency response will result in all positions except dead-ahead. This is also a defect of many older so-called omnidirectional mics and a few of the cheaper modern ones. It is unavoidable in moving-coil microphones because of the size of the magnet and diaphragm, but the effect will have been minimised by careful design in the better quality models.

With P.A., it is important not to have any sharp peaks in the forward response which would promote acoustic feedback. Also important is the frequency response in the back hemisphere (if it is a cardioid, which P.A. mics usually purport to be). Any peaks in that region will pick up unwanted sound from the room and the loudspeakers; and will guarantee absolutely uncontrollable feedback at moderate volume settings. Several commercially available microphones, including an expensive radio mic, are appalling in this respect; the battery compartment forming a resonant tube behind the capsule.

These days, for a given price, the electret system will give better sound quality than the moving-coil. The disadvantage of an electret is the need to power it, with problems of noisy battery connections, switch 'clump' noises and a flat battery every time you go to use it because someone (not yourself, of course) put it away switched-on.

WIRING

Screened cable is essential for all present-day microphone circuits (there were some systems in the past which could sometimes work without it). To avoid spurious pickup of mains and radio signals, the very best results will be obtained by using 'balanced line'; this is expensive and increasingly rare. An acceptable alternative is the 'twin core + screen' system although one of the cores eventually goes to the same earth as the screen - a bit like the Live, Neutral and Earth of the mains. This has proven adequate for most situations but necessitates at least three connections in the plug. It rules-out the use of single-point jacks and 'phono' (RCA) plugs, as if they didn't have enough faults to rule them out anyway.

This leaves DIN and XLR as the two main contenders. DIN is cheaper, smaller and lighter, but is rarely found on microphones nowadays. XLR connectors have become the modern de-facto standard and are particularly robust. This will probably mean re-wiring any cheap mics, as they invariably come supplied with a single-core cable and single-point jack &emdash; but a little experience of working with that arrangement should soon convince the user that it is worth any amount of trouble to change it.

Some mics have a small transformer coupled to a three position impedance switch (Low - Off - High). Invariable the switch gets mis-operated and causes havoc, also the high impedance setting is usually unnecessary, so the feature is best removed or disabled.

The switch can be usefully retained and rewired so that it gives: Off - Power on - Sound on, which prevents the loud 'clump' when the power is switched on and off an electret capsule.

Poppy Records website: http://www.poppyrecords.co.uk