Loudspeaker Basics 101- Part 1

Genelec 1038A – Active Speaker

This is the first of three posts that will take a basic look at:

1. Speaker drivers and crossovers
2. Speaker enclosures and
3. Room placement

Introduction

Every day we listen to sound emanating from radios, TV’s, phones, computers and of course our Hi-Fi and HT systems. They all have one thing in common, at least one speaker driver that converts electrical energy into sound pressure energy. So how do they do this?

Electrical energy is converted to sound energy in one of three ways:

1. By generating a magnetic field that interacts with another static magnetic field
2. By creating an electric field that interacts with another static electric field
3. By modulating a plasma arc; the Ionophone or plasma speaker.

The fist two types on energy conversion are the most common with type one being by far the widest used and the lowest cost. I shall not be reviewing the third category in any detail as it is not used in modern day audio environments. Many companies throughout the world have made plasma speakers like Fane in the UK in 1965, but they no longer have a place in home audio and can have some unpleasant side effects on a rooms atmosphere.

Image Credit: Fane Acoustics

Image Credit: Fane Acoustics

The conversion of electrical energy to sound pressure energy must be linear in order to accurately represent the original electrical signal and produce the minimum amount of distortion. Electrical ‘sound’ signals vary tremendously in both amplitude (volume) and rate (frequency). Typically voltages range from microvolts to 100 volts and frequencies from 10Hz to 100KHz, all must be handled by the speaker driver system.

As will be seen such wide ranges of electrical and frequency performance cannot be easily handled by single speaker drivers. This gives rise to the need for multiple speaker drivers to efficiently and accurately reproduce the electrical signal as a sound pressure wave.

NOTE: When discussing sound pressure levels (SPL) the range from the threshold of hearing to those typically encountered in our daily lives is very large. The SI unit of sound pressure is the Pascal (Pa). However, this unit is rarely used when discussing everyday sound pressures. In practice it is measured using a logarithmic unit called the Decibel (dB). This number is a ratio between the level you are hearing and an established known reference, in this case 2×10-5 Pa that represents 0dB, the threshold of hearing. Using decibels is much more convenient and easier to understand than Pascals. For example a human voice at 3 feet would typically read 70dB on a sound pressure meter.

Equal Loudness Curves

In most studio, home and cinema environments sound pressure levels will vary between 20dB and 115dB. Continuous levels above approximately 85dB have been shown to eventually damage hearing and levels above 120dB will cause pain and permanent ear damage.

It is worth remembering that for every 6dB increase in SPL the loudness doubles and for every 6dB decrease the loudness halves. We can therefore see that at a typical listening level of 70dB for a 1KHz tone, for a 20Hz tone to sound as loud it must be more than 40dB louder or the SPL is 100 times higher. This is important when you consider the power that each type of speaker driver is required to handle.

Speaker Driver Designs

The Moving Coil Driver

This is the most common design for a speaker driver. It is based on the discovery in 1820 that found by passing electrical energy through a coil, or even a part of a coil, a magnetic field is generated that is proportional to both the amount of current and the number of turns that make the coil up. If that coil is placed within a fixed magnetic field it will try to react to that fixed field as the polarity and magnitude of the generated field varies. Just like two fixed magnets attracting or repelling each other.

Attaching this coil (voice coil) to a suspended diaphragm causes the air in contact with the diaphragm to be compressed and rarified as the varying magnetic force from the voice coil causes the diaphragm to move forwards and backwards.

The Basic Moving Coil Driver

Major components:

1. The chassis supports the entire driver structure allowing it to be attached to the desired enclosure.
2. The voice coil is made from either aluminum or copper wire wound on a heat resistive former. The arrangement of these coils within the magnetic field affects the performance of the unit.
3. The fixed magnet strength and pole piece shape is an integral part of the design and the speaker’s performance.
4. The diaphragm or cone is made of paper or a light metal like Aluminum or Beryllium or sandwiched polystyrene. It is supported at its opening using corrugated paper/metal or a rubber roll surround. This supporting surround must allow for the movement of the cone while keeping it centered.
5. The spyder maintains the position of the voice coil to be correctly centered within the fixed magnetic field while at the same time providing minimum restriction to the cones movement.

SVS PB12NSD Subwoofer

Genelec 1038 Tweeter

Todays modern driver designs employ many special features to control the movement of and damp the cone/diaphragm in order to support large (and small) accurate linear excursions and low distortion.

Advantages

The enormous flexibility and fundamental low cost of the design of this driver makes it by far the most popular and widely used. The design can easily be modified to cover the entire audible frequency spectrum from below 10Hz to well above 25KHz, and handle enormous powers as high as several KW. Each driver design being carefully tailored to meet the frequency and power requirements over a limited number of octaves.

Disadvantages

The movement of the cone sets up vibrations and resonances in its material that need to be controlled in order to reduce unwanted frequencies and ‘ringing’. There are many designs of these cones from simple stiff paper having complex ripped patterns and shaped flares to metal and metal sandwiched polystyrene. The whole purpose of these designs is to stiffen the cone while at the same time keeping its mass low, so as to stop/damp the unwanted resonances and/or to cause only certain parts of the cone to radiate particular frequencies.

The Ribbon Driver

A special form of the moving coil design is the ribbon or Planar magnetic device.

The Ribbon Driver

Decca Kelly DK30 Ribbon

It can be imagined that moving the mass of a diaphragm/cone takes energy to set it in motion AND to stop it travelling. This energy comes from the signal and is absorbed by the amplifier causing distortion or non-linear movement of the cone. So the lighter the speaker cone the less energy that has to be lost controlling its motion and the lower the distortion. This is the basis of the ribbon speaker that uses a small and very light piece of corrugated aluminum placed in a VERY high magnetic field. Unfortunately, this basic design has a number of technical issues that needed to be overcome:

1. Typical power amplifiers cannot drive the very low resistance (impedance) of the ribbon so an input transformer is required.
2. The sound pressure generated by a small ribbons movement needs to be correctly coupled to the surrounding air so a horn is generally required.
3. Until recently even large ribbon speakers could not handle frequencies much below 100Hz.
4. They have restricted power handling capacity due to the fragile nature of the ribbon.
5. They cannot generate very high sound pressure levels, again due to the fragile nature of the ribbon and its restricted movement.
6. The original designs required large and powerful magnets to make up for the low magnetic field produced by the ‘single turn’ ribbon voice coil. These magnets are very expensive. Modern large ribbon panels use lower cost strip magnets.

The original horn loaded ribbons were designed for frequencies above 1KHz so did not require high powers to create sufficient sound pressure levels for typical home settings.

Well established horn loaded designs where originally created by Decca, the DK30 and London models.

The DK30 Horn Assembly

Despite the above technical issues those original designs have been expanded upon and full-range ribbon speakers without horns are now available by companies like Magnepann who manufacture both large full range and dedicated bass ribbon speakers or ESS who manufacture the Heil Air Motion Transformers. These so called quasi ribbons use materials like mylar for the diaphragm with the aluminum ribbon bonded to it. This provides a huge moving surface area and allows the designs to go as low as 20Hz. However, high SPL’s at low frequencies are still not possible when compared to moving coil designs. Ribbons cannot be used as sub-woofers.

Due to the extremely low moving mass and uniformity of drive, these quasi ribbons provide very open, transparent and un-colored sound, together with an acceptable bass response and a short term mid-band SPL as high as 110+dB from these large panels.

Image Credit: Magnepan -MG30.7

Other manufacturers would include; Alsyvox, WisdomAudio, and Eminent Technology.

The Electrostatic Driver

Like charges attract, unlike charges repel, just like north and south magnetic poles. This driver operates by placing a very high voltage between two electrically conductive fixed panels called stators and a flexible conductive membrane(s). The electrical sound signal is then added to this high voltage causing the level of charge on the membrane to vary allowing it to be attracted to and repelled from the fixed stators. As in the case of the ribbon, special electronics are required within the speaker to allow the power amplifier to drive it. Also the moving membrane design and the way it is driven has a significant effect on the overall performance of the speaker.

We saw that the mass of the moving diaphragm is one of the parameters that significantly effects the performance of a loud speaker. Electrostatic loudspeakers have very low moving mass providing excellent high and mid frequency response while at the same time can have a large surface area allowing for acceptable bass extension. Due to the very small movement of their diaphragms they do not suffer from Doppler distortion and their membrane designs can be optimized in order to handle most of the audio range and provide an acceptable sound dispersion into the listening area.

The Electrostatic Driver

Major components:

• Stator – fixed perforated panels or grids that provide the polarizing static charge. The SPL has to pass through these panels.
• Membrane – a flexible conducting membrane that supports the electrostatic charge and generates the SPL.
• Very high voltage DC power supply – provides a very high voltage to create a polarizing static charge between the stators and membrane.
• Input transformer – matches the impedance of the membrane to that required by the audio power amplifier.

Advantages:

1. Levels of distortion one to two orders of magnitude lower than conventional cone drivers in a speaker cabinet.
2. The extremely light weight of the diaphragm which is uniformly driven across its whole surface.
3. Excellent frequency response (both in amplitude and phase) because the principle of generating the force and pressure is almost free from resonances.

Disadvantages:

1. Lack of low bass response, due to phase cancellation from a lack of enclosure – not shared by all designs. The bass rolloff 3db frequency occurs when the narrowest panel dimension equals a quarter wavelength of the radiated frequency for dipole radiators, so for a Quad ESL-63, which is 0.66 meters wide, this occurs at around 130 Hz. (see post 2)
2. There is also the difficult physical challenge of reproducing high SPL’s for low frequencies with a vibrating taut film having little excursion amplitude. However, as most diaphragms have a very large surface area compared to cone drivers, only small amplitude excursions are required to put relatively large amounts of acoustic energy out. While SPL of bass is lacking it can be of better quality (tighter and with less coloration) than that of cone systems.  Nevertheless maximum bass levels cannot be augmented because they are ultimately limited by the membrane’s maximum permissible excursion before it comes too close to the high-voltage stators, which may produce electrical arcing and burn holes through it. Recent technically more advanced solutions for this lack of bass includes the use of large, curved panels (Sound-Lab, MartinLogan), electrostatic subwoofer panels (Audiostatic, Quad), and long-throw electrostatic elements allowing large diaphragm excursions (Audiostatic).
3. They are fundamentally directional limiting the area of the ‘sweet spot’ where proper stereo imaging can be heard.
4. They have a sensitivity to ambient humidity levels, a tendency to attract dust, insects and conductive particles and the diaphragms can gradually deteriorate over time.
5. They need protection measures to physically isolate their high voltage parts from accidental contact with humans and pets.

The most popular original and current designs were created by Quad Electronics who are now joined by Martin Logan, Sound-Lab and Audiostatic as the recognized leaders in the design of this speaker technology. They provide a quality of sound and acoustic performance that is very similar to large ribbon speakers, being very open and uncolored with very tight well controlled bass.

Driver Size

Your hearing requires very different SPL’s at low and high frequencies for them to sound equally loud. Frequencies of 40Hz and lower require very high SPL’s compared to frequencies from 2KHz and up (see equal loudness curves above). These high SPL’s require a lot of air displacement implying large diameter cones and large amounts of diaphragm/cone movement. Large low frequency high power driver designs are NOT suitable to handle high frequencies where the rapid cone movement of such a large and relatively heavy cone would be very demanding.

There are many technical challenges associated with trying to make a single driver handle a full range of audio frequencies (20Hz-20Khz). The simplest problem to understand would be that of Doppler distortion. If a diaphragm/cone is moving in and out at a low frequency of 20Hz and a high frequency of 2KHz is also fed to it, the 2KHz signal will be modulated by the 20Hz signal in just the same way as you hear an ambulance siren higher in frequency as it approaches you and then drops in frequency as it passes you. Clearly an unsatisfactory performance.

In order to get the best efficiency, performance and lowest distortions from these drivers their designs are optimized over a restricted number of octaves. The design approach is  different for moving coil and electrostatic/ribbon speaker systems.

This typically means the use of at least two or more drivers for moving coil designs:

1. A sub-woofer for flow frequencies 10Hz -80Hz (8″ to 15″ diameter) – usually a separate standalone enclosure handling 80Hz and lower for ALL speaker driver types.
2. A bass and lower mid-range unit – 40Hz to 2.5KHz (5″ to 12″ diameter)
3. An upper mid-range and high frequency unit – 2.5KHz to 20KHz (0.5″ to 2″ diameter)

In the case of electrostatic and ribbon drivers, multiple membrane ‘sections’ are required to cover the entire frequency range. Ribbon bass units are often added to support the lower octaves of ribbon speakers and moving coil sub woofers are still required to efficiently support frequencies below 100 Hz for both types of driver.

Frequently items 2 and 3 above are broken down further for all designs (see Genelec 1038 below) to become:

1. A bass unit -20Hz to 300Hz
2. A mid-range unit – 300Hz to 3KHz
3. A high frequency unit – 3KHz to 30KHz

The final frequency range over which these drivers are selected to operate and there size, are defined by the manufactures design requirements and power handling required.

The size of the moving coil driver or membrane panel usually indicates the amount of power that it can handle and consequently the highest SPL that it can generate. Put simply, the larger the driver (and the greater its excursion) the more power it can handle.

Membrane speakers like electrostatic and ribbons need to be very large in order to compete with the SPL levels and power handling capacities that moving coil speakers can generate. As such they are mostly used in residential settings.

My original pair of Maudant-Short MS600 speakers used a long throw Celestion Studio 12” bass/mid-range with a rubber-roll surround that handled from 20Hz to 2.5KHz and a Decca Kelly horn loaded DK30 ribbon that handled 2.5KHz to 25KHz. Maximum power approx. 50 watts.

My current 1038 Genelecs use a 12” bass, a 5” mid-range and a 1” HF unit. The mid-range and HF units use small waveguides to improve their efficiency and smooth out their frequency responses. Each speaker contains three Class D power amplifiers (1×400 watt & 2×120 watt), the active crossover filters and equalization.

Genelec 1038

1038 Internal Block Diagram

NOTE: Loudspeaker driver efficiency can be significantly increased by adding horns to the front of them, as in the case of the ribbon or many other high and mid-frequency speaker designs.  These horns can provide significant increases in efficiency allowing relatively small speakers not only to go lower but also produce relatively high SPL’s for low input powers. For most practical applications at home the huge horns that support bass frequencies are not practical. See part 2 for further details on horns.

Speaker Crossovers

So now we have a range of drivers that can very accurately reproduce the entire audible frequency range and more. The issue is that all the frequencies are contained on one pair of cables so how do we split them up to feed each dedicated driver or part of the driver system?

This is the job of the crossover system. This set of frequency filters may be:

1. Passive – by far the most popular and least expensive having no active electronics like transistors, integrated circuits or valves.
2. Active – for higher end speaker systems and sub-woofers – lots of electronics.

There are three basic types of frequency filter characteristic that makes up the crossover design:

1. Low pass – these allow frequencies below a cut-off frequency to pass
2. High pass – these allow frequencies above a certain cut-off frequency to pass
3. Bandpass – these allow frequencies between two cut-off frequencies to pass

1st Order (6db/Octave) Passive Filter Designs

Active Filter Designs – Note that no inductors are required.

The design of these filters shall not be reviewed here. One of the design challenges with passive filters is that the speakers and their enclosures are complex loads, and passive filters have to take account of that varying load. Active filters are fundamentally easier to design but they require additional power amplifiers, typically one for each driver, causing a significant increase in cost and system complexity.

Simplified Passive & Active Crossover Designs

Both solutions can provide excellent performance, but active crossovers will generally win out due to the ability to fine tune the crossovers frequency and time domain performance, better driver control by the amplifier and in the case of sub-woofers DSP processing to optimize the control of the sub driver.

There is a third hybrid solution that combines the above two approaches often referred to as bi-amping or even tri-amping. This is when each driver has its own separate passive crossover filter and its own power amplifier.

A Simple Bi-Amped System

The advantages of this configuration is that:

1. As each power amplifier is restricted to handling a much narrower range of frequencies, it can be optimized for them, creating less harmonic and intermodulation distortion.
2. The load presented to each amplifier tends to be more predictable and allows for better control of the driver.

Some of the more expensive speakers have additional speaker terminals that allows the user to separate each drivers input so that they can be driven individually.

This approach was how I originally drove my modified Maudant Short MS600 speakers until I designed a fully active crossover and removed the passive components. A Quad 405 current dumper drove the bass units and a pair of Leak TL50+ valve power amps drove the ribbons. Both approaches providing varying levels of improvement in the bass, mid-range and treble with tighter, more punchy and cleaner bass, a more open mid-range and a slightly more detailed HF with better imaging when compared to the original combined passive arrangement driven by just a Quad 405 or pair of Leak TL50+ power amplifiers.

Next

In part 2 we shall see why drivers need to be attached to a baffle or mounted inside a speaker enclosure, the different enclosure designs and how they affect frequency response and power handling.

See part 2 – Speaker Cabinet Design here.

See part 3 – Speaker Room Placement here.

See here for my 2 part post on Acoustics 101