Originally appeared in the January/February 1989 issue of Car Stereo Review magazine.
Amplifiers, speakers, cassette decks, tuners, and CD players. These are some of the more familiar components used in the reproduction of sound. And while all of these components play an important part in this process, the quality of the sound you hear will only be as good as the weakest link in the system. For this reason, you must select every component for your auto sound system carefully. Probably the most neglected, yet critical component in this selection process, is the crossover network.
The need for a crossover can be attributed to the inability of most drivers to reproduce the entire musical spectrum smoothly and efficiently. Because of this, most speaker systems are comprised of several drivers, each of which is dedicated to the reproduction of a specific range of frequencies.
These frequencies are routed to the appropriate driver by a crossover network. This is a device that divides the audio spectrum into two or more frequency bands. By doing this, we can insure that each driver will receive only the frequencies that it can reproduce properly. Also, the crossover network prevents potentially damaging low-frequency energy from reaching the midrange or tweeter drivers.
This frequency division process is accomplished by using high-pass and low-pass filters. A high-pass filter passes frequencies above a pre-determined frequency and attenuates those below. This pre-determined frequency is called the crossover point. Conversely, a low-pass filter passes frequencies below the crossover point and attenuates those above. By cascading high-pass and low-pass filters, a band-pass filter can be realized. Filters of this type attenuate all frequencies outside of the passband created by the high and low-pass filters.
The rate at which the attenuating process takes place is dependent upon the slope of the crossover. Typical slopes for auto sound systems are 6, 12, and 18 dB per octave, with 12 being the most common. Higher numbers indicate a much steeper cutoff rate while the lower numbers indicate a more gradual rolloff.
Currently, there are two varieties of crossovers available; passive and active. The biggest distinction between them is where they are located in the audio signal path.
Passive crossovers are located between the amplifier and the drivers, and for this reason, are sometimes referred to as high level networks. These networks are comprised of inductors, capacitors, or some combination of each.
An inductor, also known as a "coil" or "choke", is a coil of wire which may or may not have an iron or ferrite core. Inductors with an iron core are called "iron core inductors" while those without are called "air core inductors". The basic characteristics of both types is the same. As the frequency passing through an inductor increases, so does the inductive reactance of the coil. This rise in impedance at higher frequencies allows us to use the inductor as a filter that passes low frequencies but chokes off high ones.
Capacitors do just the opposite. As frequency decreases, the capacitors reactance increases. That is, capacitors pass high frequencies and filter out low ones. These capacitors are generally non-polarized electrolytics, polypropylene, or Mylar and consist of interleaved layers of foil and insulating material.
One common myth pertaining to passive crossovers is that they "soak up" the power that is not used for each particular driver. While there is some insertion loss, the filtering action actually takes place due to the impedance mismatch created by the network. For example, if we place a capacitor in series with a 4 ohm tweeter, we have created a first order (6 dB/octave) high pass filter. As frequency goes down, the capacitive reactance of the capacitor goes up. At the crossover point, the impedance presented by the capacitor will be equal to the impedance of the tweeter. Since the capacitor is in series with the tweeter, the effective load impedance "seen" by the amplifier is 4 + 4 or 8 ohms. This rise in load impedance causes a 3 dB reduction in output power at the amplifier. As the frequency continues to go down, the effective impedance of the network continues to rise and the output of the amplifier continues to be reduced at a rate determined by the slope of the crossover. In our example, the rolloff rate would occur at 6 dB/octave because it is a first order network.
The number of inductors and capacitors in a network determines its order. For example, if a network consisted of one component it would be a first order network. Two components would comprise a second order network and so on. Final response rolloff rates occur in multiples of 6 dB/octave, according to the order of the filter. Consequently, a first order network would have a slope of 6 dB/octave; a second order network would have a slope of 12 dB/octave; and so on. Figure 1 shows examples of 1st, 2nd, and 3rd order high pass and low pass networks while Figure 2 depicts a chart used for determining the proper values of inductance and capacitance for different crossover frequencies.
Component Values for 12 dB/Octave High & Low Pass Filters
It is generally accepted that 1st order networks should only be used for low-pass filters because a 6 dB/octave rolloff rate will not provide adequate protection for midrange and tweeter drivers. On the other hand, slopes greater than 18 dB/octave tend to suffer from poor transient response, audible ringing, phase disparity among the drivers, and excessive insertion loss.
Insertion loss is a term used to account for the power that is lost when a passive crossover network is used. The most significant power loss occurs in the inductor due to the resistance of the wire. Some inductors can have dc resistances as high as 1 ohm. This can rob almost 20 percent of the power intended for the drivers. Although this may seem high, 20 percent is only about a 1 dB drop in output level.
Resistance in the crossover network also degrades the damping ability of the amplifier. Since the damping factor is the ratio of the load impedance to the output impedance of the amplifier, adding any resistance to the output impedance of the amp will greatly reduce the ability of the amp to control cone movement. The result of this could be bass that lacks definition or sounds muddy.
In order to reduce these problems, you must be very careful when selecting the components for your passive crossover network. The best sounding (and most expensive) capacitors are the Mylar and polypropylene high-voltage variety. These should always be used for high-pass filters because they are series components.
For low-pass filters, any type of capacitor can be used; however, the choice of inductor is more critical since it is the series element. Air core inductors are better sounding because of their lower distortion figures; unfortunately, they tend to have a higher dc resistance than iron core varieties.
By using active networks, many of the problems associated with passive networks can be eliminated. Unlike passive networks, active networks are comprised of solid state electronics that divide the audio spectrum BEFORE amplification. Since these networks contain active components such as transistors and integrated circuits, they are sometimes referred to as "electronic crossovers".
Whenever an electronic crossover is used, multiple power amplifiers are required. Two-way active crossover systems will require two amplifiers, one for each group of frequencies. This is known as "biamplification". Three-way systems (triamplification) would require three amplifiers and so forth.
Performance is just one of the many key features realized by using an electronic crossover. Since the filtering process takes place prior to amplification, there is no insertion loss. Also, since the amplifier is connected directly to the load, there will be a significant improvement in damping. Distortion will also be reduced since the load that the amplifier "sees" will be less reactive. Isolation is another benefit active systems provide. What this signifies, is that changing the impedance of the load will have no effect on the crossover. This is not the case in passive systems where the crossover itself is designed around drivers with specific impedances. If these impedances change, crossover points as well as filter Q and thus response shape and damping will change.
Perhaps the most attractive feature offered by an electronic crossover is its flexibility. Almost all units of this type allow for the continuous selection of crossover points. Some even go as far as permitting the user to select the desired slope and "Q" of the network. These adjustments can usually be made quickly and conveniently with the twist of a knob or the replacement of a module.
So, which type of crossover is best for your system? Well, that depends. Usually the system itself dictates the type of crossover required. However, there are some important questions you should ask yourself before purchasing a crossover.
How much money do you want to spend? Electronic crossovers on the average cost between 100 and dollars. A passive network can range from 5 to 10 dollars if you build it yourself, to as much as 50 dollars for a professionally designed and packaged system. Also, active crossovers require multiple amplifiers which can add significantly to the price of the system.
How much flexibility do you want? Active crossovers, with their continuously variable crossover points and small size, are considerably more flexible than their passive counterparts.
What type of performance are you looking for? The efficiency, isolation, damping, and distortion ratings of active networks are superior to those of passive systems.
After answering these questions, evaluate the system itself. Single amp systems must use a passive network. For systems with two amplifiers, I prefer to use a combination of active and passive filters rather than going strictly passive.
First, use an active crossover to split the audio spectrum at the lowest frequency that the midranges can handle comfortably. The low-pass out of the crossover feeds the woofer amp and the high-pass output feeds the midrange/tweeter amp. Next, wire the output of the midrange/tweeter amp directly to the midranges. Then, insert a 12 dB/octave passive high-pass crossover in series with the tweeters and then connect their inputs in parallel with the midranges. One advantage to this type of arrangement is that large bass power demands will not reduce the power available for the midrange and tweeter drivers. Also, clipping the bass will not damage the mids and tweets as would be the case in a single amp passive system.
For systems with three or more amplifiers, I would recommend going active all the way. At this point, the advantages of using an active system far outweigh cost considerations.
In a triamplified system, the low-pass output of the crossover would feed the woofer amp. The band-pass output would feed the midrange amplifier and the high-pass output would feed the tweeter amp. This type of arrangement will give maximum performance and flexibility since the output level of each amplifier can be tailored to the exact requirements of the drivers.
Of all the variables involved in crossover design, the crossover point is the most ambiguous. One reason for this is that every vehicle is different. Peaks and dips created by the listening cavity need to be taken into account when choosing a crossover point. Other factors affecting the selection process are size of the drivers used, rolloff slope, and the number of crossover points in the system.
For a typical 12 dB/octave crossover network the following rules can provide a starting place for experimenting with various crossover points. Keep in mind, however, that the crossover selection process is very subjective. The goal is to obtain a reliable system that provides a quality of sound that is pleasing to the listener. Because of this, crossover point selection is more of an art than a science.
Never cross the midranges or tweeters at a frequency of less than twice their natural frequency of resonance. For example, if a midrange driver has a resonance of 400 Hz, it should be crossed over at a frequency of 800 Hz or higher.
Theoretically, a woofer should be used to cover only those frequencies where the wavelength of the sound to be reproduced does not exceed the diameter of the driver.
Sub-woofers are generally crossed at 100 Hz and below.
Woofers are generally crossed between 100 - 500 Hz.
Midranges are crossed between 300 - 500 Hz and 3500 - 8000 Hz.
Tweeters are crossed at 5000 Hz and above.
If these guidelines dont help, try starting with a 200 Hz crossover point between woofers and midranges and a 5000 Hz crossover point between midrange and tweeters. Contacting the manufacturer of the drivers used might also provide valuable insight in the selection of a crossover point.
With the right crossover and the correct crossover points, the reliability of the system will be enhanced. Drivers will only reproduce frequencies they were designed for and thus will sound smooth and silky. The end result is a system that is reliable, efficient, and a pleasure to listen to.