Archive-name: car-audio/part2
Rec-audio-car-archive-name: FAQ/part2
Version: 2.3
Last-modified: 5 Jul 94


      3 Components

 This section describes various components that you can have in
 a car audio system, along with common specifications, desirable
 features, some of the best and worst brands, and so on.

 Be aware that there is no standardized testing mechanism in
 place for rating car audio products.  As such, manufacturers
 are open to exaggerating, "fudging", or just plain lying when
 it comes to rating their own products.

    3.1 What do all of those specifications on speakers mean? [JSC,CD]

 "Input sensitivity" is the SPL the driver will produce given
 one watt of power as measured from one meter away given some
 input frequency (usually 1kHz unless otherwise noted on the
 speaker).  Typical sensitivities for car audio speakers are
 around 90dB/Wm.  Some subwoofers and piezo horns claim over
 100dB/Wm.  However, some manufacturers do not use true 1W
 tests, especially on low impedance subwoofers.  Rather, they
 use a constant voltage test which produces more impressive
 sensitivity ratings.

 "Frequency response" in a speaker refers to the range of
 frequencies which the speaker can reproduce within a certain
 power range, usually +/-3dB.

 "Impedance" is the impedance of the driver (see 1.1), typically
 4 ohms, although some subwoofers are 8 ohms, some stock Delco
 speakers are 10 ohms, and some stock Japanese imports are 6
 ohms.

 "Nominal power handling" is the continuous power handling of
 the driver.  This figure tells you how much power you can put
 into the driver for very long periods of time without having to
 worry about breaking the suspension, overheating the voice
 coil, or other nasty things.

 "Peak power handling" is the maximum power handling of the
 driver.  This figure tells you how much power you can put into
 the driver for very brief periods of time without having to
 worry about destroying it.

    3.2 Are component/separates any better than fullrange or coaxials? [JSC]

 Usually, yes.  Using separates allows you to position the
 drivers independently and more carefully, which will give you
 greater control over your imaging.  For rear fill applications,
 however, coaxial speakers will perform fine, as imaging is not
 a primary concern.

    3.3 What are some good (and bad) brands of speakers? [JSC]

 People will emotionally defend their particular brand of
 speakers, so asking what the "best" is is not a good idea.
 Besides, the best speaker is the one which suits the
 application the best.  In general, however, various people have
 claimed excellent experiences with such brands as Boston
 Acoustics, MB Quart, a/d/s/, and Polk.  Also, most people agree
 that you should avoid brands like Sparkomatic and Kraco at all
 costs.

    3.4 What do all of those specifications on amplifiers mean? [JSC,BG]

 "Frequency response" refers to the range of frequencies which
 the amplifier can reproduce within a certain power range,
 usually +/-3dB.

 "Continuous power output" is the power output of the amplifier
 into one channel into a certain load (usually four ohms) below
 a certain distortion level (usually at most 1%THD) at a certain
 frequency (usually 1kHz).  A complete power specification
 should include all of this information, e.g. "20W/ch into 4
 ohms at < 0.03%THD at 1kHz" although this can also be stated as
 (and be assumed equivalent to) "20W/ch at < 0.03%THD".  The
 amplifier should also be able to sustain this power level for
 long periods of time without difficulties such as overheating.

 "Peak power output" is the power output of the amplifier into
 one channel into a certain load (usually four ohms) below a
 certain distortion level (usually much higher than the
 continuous rating level) at a certain frequency (usually
 1kHz).  A complete power specification should include all of
 this information, e.g.  "35W/ch into 4 ohms at < 10.0%THD at
 1kHz" although this can also be stated as (and be assumed
 equivalent to) "35Wch at < 10.0%THD".  Consumer warning: some
 manufacturers will state the "peak power output" rating by
 including the amount of power which can be drawn from
 "headroom", which means power supply capacitors.  They usually
 will not tell you this in the specification, however; indeed,
 they tend to prominently display the figure in big, bold
 letters on the front of the box, such as "MAXIMUM 200W PER
 CHANNEL!!!" when the continuous rating is 15W/ch and the unit
 has a 5A fuse.

 "Damping factor" represents the ratio of the load being driven
 (that is, the speaker - usually four ohms) to the output
 impedance of the amplifier (that is, the output impedance of
 the transistors which drive the speakers).  The lower the
 output impedance, the higher the damping factor.  Higher
 damping factors indicate a greater ability to help control the
 motion of the cone of the speaker which is being driven.  When
 this motion is tightly controlled, a greater transient response
 is evident in the system, which most people refer to as a
 "tight" or "crisp" sound.  Damping factors above 100 are
 generally regarded as good.

 "Signal to Noise" or "S/N" is the ratio, usually expressed in
 decibels, of the amount of true amplified output of the
 amplifier to the amount of extraneous noise injected into the
 signal.  S/N ratios above 90 to 95dB are generally regarded as
 good.

    3.5 What is "bridging"? Can my amp do it? [JSC]

 Bridging refers to taking two channels of an amplifier and
 combining them to turn the amplifier into a one channel
 amplifier.  In normal operation, one wire which goes to a
 speaker from the amplifier is "neutral", that is, the potential
 never changes (with respect to another fixed point, like
 ground).  The other wire is "hot", that is, it carries the
 fluctuating AC speaker signal.  The speaker "sees" a potential
 between these two leads, and so there is a voltage applied to
 the speaker.  When an amplifier is bridged, both leads are
 "hot".  However, one signal must be inverted, or else the
 speaker will never see a potential, as both wires are carrying
 roughly the same signal.  With one signal inverted, the speaker
 will see a signal that is twice as great as one signal alone.
 Thus, if your amplifier does not have a switch or button of
 some sort which inverts one channel, you cannot bridge your
 amplifier (unless you build an external inverter).  With
 respect to power, the commonly accepted definition is that when
 you bridge an amplifier, you add all of the characteristics of
 the bridged channels together.  Thus, if you bridge an
 amplifier that is 50W/ch into 4 ohms at < 0.05%THD, your
 bridged channel is 100W/ch into 8 ohms at < 0.10%THD.
 Therefore, an amplifier which is 2 ohm stable in stereo mode is
 only 4 ohm stable in bridged mono mode, and an amp which is 4
 ohm stable in stereo is only 8 ohm stable in bridged mono.

    3.6 What is "mixed-mono"? Can my amp do it? [JSC]

 Some amplifiers which are both bridgeable and able to drive low
 impedance loads also allow you to use "mixed-mono" mode.  This
 involves driving a pair of speakers in stereo mode as well as
 simultaneously driving a single speaker in bridged mono mode.
 What happens is that you put your amp in bridged mode, which
 inverts one output signal.  You then connect the mono speaker
 as you normally would in bridged mode.  To the channel which is
 not inverted, you connect your stereo speaker as you normally
 would.  To the channel which is inverted, you connect the other
 stereo speaker with its leads reversed (+ to - and - to +)
 since the signal is inverted.

    3.7 What does "two ohm stable" mean? What is a "high-current"
 amplifier? [JSC]

 An x ohm stable amplifier is an amp which is able to
 continuously power loads of x ohms per channel without
 encountering difficulties such as overheating.  Almost all car
 amplifiers are at least four ohm stable.  Some are two ohm
 stable, which means that you could run a pair of four ohm
 speakers in parallel on each channel of the amplifier, and each
 channel of the amp would "see" two ohms.  Some amps are
 referred to as "high-current", which is a buzzword which
 indicates that the amp is able to deliver very large
 (relatively) amounts of current, which usually means that it is
 stable at very low load impedances, such as 1/4 or 1/2 of an
 ohm.  Note that the minimum load rating (such as "two ohm
 stable") is a stereo (per channel) rating.  In bridged mode,
 the total stability is the sum of the individual channels'
 stability (see 3.5).

    3.8 Should I buy a two or four (or more) channel amplifier? [JSC]

 If you only have one line-level set of outputs available, and
 wish to power two sets of speakers from a single amplifier, you
 may be able to save money by purchasing a two channel amplifier
 which is stable to two ohms rather than spending the extra
 money for a four channel amp.  If you do this, however, you
 will be unable to fade between the two sets of speakers
 (without additional hardware), and the damping factor of the
 amplifier will effectively be cut in half.  Also, the amp may
 run hot and require fans to prevent overheating.  If you have
 the money, a four channel amp would be a better choice.  You
 would need to add a dual-amp balancer in order to maintain
 fader capability, however, but it is more efficient than
 building a fader for a two channel amp.  If you wish to power a
 subwoofer or additional speakers as well, you may want to
 purchase a five or six channel amp.

    3.9 What are some good (and bad) brands of amplifiers? [JSC]

 As with speakers, people emotionally defend their amplifier, so
 choosing the best is difficult.  However, some brands stand out
 as being consistently good while others are consistently bad.
 Among the good are HiFonics, Phoenix Gold, a/d/s/, and
 Precision Power.

   3.10 What is a crossover? Why would I need one? [JSC]

 A crossover is a device which filters signals based on
 frequency.  A "high pass" crossover is a filter which allows
 frequencies above a certain point to pass unfiltered; those
 below that same point still get through, but are attenuated
 according to the crossover slope.  A "low pass" crossover is
 just the opposite: the lows pass through, but the highs are
 attenuated.  A "band pass" crossover is a filter that allows a
 certain range of frequencies to pass through while attenuating
 those above and below that range.  There are passive
 crossovers, which are collections of purely passive (unpowered)
 devices - mainly capacitors and inductors and sometimes
 resistors.  There are also active crossovers which are powered
 electrical devices.  Passive crossovers are typically placed
 between the amplifier and the speakers, while active crossovers
 are typically placed between the head unit and the amplifier.
 There are a few passive crossovers on the market which are
 intended for pre-amp use (between the head unit and the
 amplifier), but the cutoff frequencies (also known as the
 "crossover point", defined below) of these devices are not
 typically well-defined since they depend on the input impedance
 of the amplifier, which varies from amplifier to amplifier.

 There are many reasons for using crossovers.  One is to filter
 out deep bass from relatively small drivers.  Another is to
 split the signal in a multi-driver speaker so that the woofer
 gets the bass, the midrange gets the mids, and the tweeter gets
 the highs.

 Crossovers are categorized by their "order" and their
 "crossover point".  The order of the crossover indicates how
 steep the attenuation slope is.  A first order crossover "rolls
 off" the signal at -6dB/octave (that is, quarter power per
 doubling or halving in frequency).  A second order crossover
 has a slope of -12dB/octave; third order is -18dB/octave; etc.
 The crossover point is generally the frequency at which the
 -3dB point of the attenuation slope occurs.  Thus, a first
 order high pass crossover at 200Hz is -3dB down at 200Hz, -9dB
 down at 100Hz, -15dB down at 50Hz, etc.

 It should be noted that the slope (rolloff) of a crossover, as
 defined above, is only an approximation.  This issue will be
 clarified in future revisions of this document.

 The expected impedance of a crossover is important as well.  A
 crossover which is designed as -6dB/octave at 200Hz high pass
 with a 4 ohm driver will not have the same crossover frequency
 with a driver which is not 4 ohms.  With crossovers of order
 higher than one, using the wrong impedance driver will wreak
 havoc with the frequency response.  Don't do it.

   3.11 Should I get an active or a passive crossover? [JSC]

 Active crossovers are more efficient than passive crossovers.
 A typical "insertion loss" (power loss due to use) of a passive
 crossover is around 0.5dB.  Active crossovers have much lower
 insertion losses, if they have any loss at all, since the
 losses can effectively be negated by adjusting the amplifier
 gain.  Also, with some active crossovers, you can continuously
 vary not only the crossover point, but also the slope.  Thus,
 if you wanted to, with some active crossovers you could create
 a high pass filter at 112.3Hz at -18dB/octave, or other such
 things.

 However, active crossovers have their disadvantages as well.
 An active crossover may very well cost more than an equivalent
 number of passive crossovers.  Also, since the active crossover
 has separate outputs for each frequency band that you desire,
 you will need to have separate amplifiers for each frequency
 range.  Furthermore, since an active crossover is by definition
 a powered device, the use of one will raise a system's noise
 floor, while passive crossovers do not insert any additional
 noise into a system.

 Thus, if you have extra money to spend on an active crossover
 and separate amplifiers, and are willing to deal with the
 slightly more complex installation and possible noise problems,
 an active crossover is probably the way to go.  However, if you
 are on a budget and can find a passive crossover with the
 characteristics you desire, go with a passive.

   3.12 How do I build my own passive crossovers? [JSC]

 A first order high pass crossover is simply a capacitor placed
 inline with the driver.  A first order low pass crossover is an
 inductor inline with the driver.  These roles can be reversed
 under certain circumstances: a capacitor in parallel with a
 driver will act as a low pass filter, while an inductor in
 parallel with a driver will act as a high pass filter.
 However, a parallel device should not be the first element in a
 set; for example, using only a capacitor in parallel to a
 driver will cause the amplifier to see a short circuit above
 the cutoff frequency.  Thus, a series device should always be
 the first element in a crossover.

 When like combinations are used, the order increases: a
 crossover in series followed by an inductor in parallel is a
 second order high pass crossover.  An inductor in series
 followed by a capacitor in parallel is a second order low pass
 crossover.

 To calculate the correct values of capacitors and inductors to
 use, you need to know the nominal impedance (Z) of the circuit
 in ohms and the desired crossover point (f) in hertz.  The
 needed capacitance in farads is then 1/(2 x pi x f x Z).  The
 needed inductance in henries is Z/(2 x pi x f).  For example,
 if the desired crossover point is 200Hz for a 4 ohm driver, you
 need a 198.9 x 10^-6 F (or 199uF) capacitor for a high pass
 first order filter, or a 3.18 x 10^-3 H (or 3.18mH) inductor
 for a low pass first order filter.

 To obtain low insertion losses, the inductors should have very
 low resistance, perhaps as low as 0.1 to 0.2 ohms.

 Also, be sure to select capacitors with proper voltage
 ratings.  The maximum voltage in the circuit will be less than
 the square root of the product of the maximum power in the
 circuit and the nominal impedance of the driver.  For example,
 a 4 ohm woofer being given 100W peak will see a maximum voltage
 of sqrt(100*4) = sqrt(400) = 20V.  Make sure that the
 capacitors are bipolar, too, since speaker signals are AC
 signals.  If you cannot find bipolar capacitors, you can use
 two polar capacitors in parallel and in opposite polarity (+ to
 - and - to +).  However, there are some possible problems with
 this approach: the forward voltage rating will probably not be
 equal to the reverse voltage rating, and there could be a
 reverse capacitance as well.  Both problems could adversely
 affect your circuit if you decide to use opposite polarity
 capacitors in parallel.

 To build a second order passive crossover, calculate the same
 initial values for the capacitance and inductance, and then
 decide whether you want a Linkwitz-Riley, Butterworth, or
 Bessel filter.  An L-R filter matches the attenuation slopes so
 that both -3dB points are at the same frequency, so that the
 system response is flat at the crossover frequency.  A
 Butterworth filter matches the slopes so that there is a peak
 at the crossover frequency, and a Bessel filter is in between
 the two.  For an L-R filter, halve the capacitance and double
 the inductance.  For a Butterworth filter, multiply the
 capacitance by 1/sqrt(2) and the inductance by sqrt(2).  For a
 Bessel filter, multiply the capacitance by 1/sqrt(3) and the
 inductance by sqrt(3).

 You should realize, too, that crossovers induce a phase shift
 in the signal of 90 degrees per order.  In a second order
 filter, then, this can be corrected by simply reversing the
 polarity of one of the drivers, since they would otherwise be
 180 degrees out of phase with respect to each other.  In any
 case with any crossover, though, you should always experiment
 with the polarity of the drivers to achieve the best total
 system response.

 As with the definition of crossover slopes, the above
 definition of the phase shift associated with a crossover is
 also an approximation.  This will be addressed in future
 revisions of this document.

   3.13 Should I buy an equalizer? [JSC]

 Equalizers are normally used to fine-tune a system, and should
 be treated as such.  Equalizers should not be purchased to
 boost one band 12dB and to cut another band 12dB and so on -
 excessive equalization is indicative of more serious system
 problems that should not simply be masked with an EQ.  However,
 if you need to do some minor tweaking, an EQ can be a valuable
 tool.  Additionally, some EQs have spectrum analyzers built in,
 which makes for some extra flash in a system.  There are two
 main kinds of EQs available today: dash and trunk.  Dash EQs
 are designed to be installed in the passenger compartment of a
 car, near the head unit.  They typically have the adjustments
 for anywhere from five to eleven (sometimes more) bands on the
 front panel.  Trunk EQs are designed to be adjusted once and
 then stashed away.  These types of EQs usually have many bands
 (sometimes as many as thirty).  Both types sometimes also have
 crossovers built in.

   3.14 What are some good (and bad) brands of equalizers?

   3.15 What do all of those specifications on tape deck head units mean?

   3.16 What are features to look for in a tape deck?

   3.17 What are some good (and bad) brands of tape decks?

   3.18 What are features to look for in a CD head unit?

   3.19 Should I buy a detachable faceplate or pullout CD player?

   3.20 What are some good (and bad) brands of CD head units?

   3.21 Can I use my portable CD player in my car? Won't it skip a lot? [JSC]

 You can use any portable CD player in a car provided that you
 have either an amplifier with line level inputs (preferred) or
 a tape deck.  If you have the former, you can simply buy a 1/8"
 headphone jack to RCA jack adapter and plug your CD player
 directly into your amplifier.  If you have the latter, you can
 purchase a 1/8" headphone jack to cassette adapter and play CDs
 through your tape deck.  The cassette adapters tend to be far
 more convenient; however, there is a significant tradeoff: by
 using cassette adapters, you limit your sound to the frequency
 response of the tape head, which is sometimes as much as an
 entire order of magnitude worse than the raw digital material
 encoded onto the CD itself.

 Portable CD players which were not designed for automotive use
 will tend to skip frequently when used in a car (relatively).
 CD players that are specially designed for automotive use, such
 as the Sony Car Discman, tend to include extra dampening to
 allow the laser to "float" across the bumps and jolts of a
 road.  Some people have indicated success with using regular
 portable CD players in a car when they place the CD player on a
 cushion, such as a thick shirt or even on their thighs.

   3.22 What's that weird motor noise I get with my portable CD player? [JSC]

 Many people report problems while playing CDs from a portable
 CD player into their car audio systems.  The problem, stated
 very simply, has to do with the stepping of the motor requiring
 a varying amount of current and non-isolated power and audio
 signal grounds.  Using a liberal application of capacitors and
 inductors, this voltage variance can be restricted to a window
 of 8.990 to 9.005V for a 9V CD player, yet even the swing
 between these two levels is enough to cause annoyingly loud
 noise on the outputs.  It has been reported that this entire
 problem can be solved by using a true DC-DC inverter at the
 power input to the CD player.

   3.23 What are some good (and bad) brands of portable CD players?

   3.24 What's in store for car audio with respect to MD, DAT and DCC? [HK]

 MiniDisc (MD) seems to have a better future than Digital Audio
 Tape (DAT) or Digital Compact Cassette (DCC) which don't seem
 to have appeal to the public.  Ease of use seems to be an
 important factor and the CD formats allows direct access to
 musical tracks at an instant.  Although MD doesn't match the
 sound quality of the standard CDs it will probably be popular
 since the players have a buffer to eliminate skipping.  DAT
 will remain as a media for ProAudio for recording purposes
 before pressing CDs.

   3.25 Are those FM modulator CD changers any good? What are my other
 options?

   3.26 What are some good (and bad) brands of CD changers?

   3.27 Why do I need a center channel in my car, and how do I do it? [HK,
 JSC]

 If a proper center image isn't achievable via a two channel
 configuration, installation of a center channel can help.
 Since the majority of recordings are done in two channel, a two
 channel system designed correctly should be able to reproduce a
 center image which was captured during recording.  A center
 channel is not simply a summation of the left and right
 channels, like bridging an amplifier; rather, it is an
 extraction of common signals from the left and right channels.
 This usually means the lead vocals, and perhaps one or two
 instruments.  These signals will then be localized to the
 center of the stage, instead of perhaps drifting between the
 left center and right center of the stage.  A signal processor
 is usually required in order to properly create a center
 channel image.  The image should then be sent to a driver in
 the physical center of the front of the car, at an
 amplification level somewhat lower than the rest of the
 speakers.  The correct frequency range and power levels will
 depend on the particular installation, though a good starting
 point is perhaps a pass band of 250-3000Hz at an amplification
 level of half the power of the main speakers (3dB down).

   3.28 Should I buy a sound field processor?

   3.29 What are some good (and bad) brands of signal processors?

Jeffrey S. Curtis                      | Email: curtis@anl.gov
Argonne National Laboratory            | Voice: 708/252-1789
9700 South Cass Avenue, 221-ECT        | Fax:   708/252-5983
Argonne, IL 60439                      |
