Section 1 - Driver guidelines
Tuesday, 09-Sep-2008, 20:56:20 GMT
Last modified: 25-Mar-2007, 19:49:50 GMT








Go to….

Motors
Parameters
Tweeters (including exotic types)
Midranges
Horns
Mid/Bass
Woofers
Subwoofers
Exotic wide-range drivers
Exotic woofers and subwoofers
Full range



Driver parameter guidelines:

Before we discuss specific types of drivers, it is first appropriate to discuss specific parameters and how they are specified. It is beyond the scope of the LDSG to explain the technical meaning and significance of the various driver parameters. The bibliography in Appendix H has further information on measuring and using Thiele-Small and other parameters in speaker system design. In the meantime, you can read up on the basics of driver parameters at the following sites:

Eminence Loudspeakers The Subwoofer DIY Page v1.1 Adire Audio Tech Central
How A Loudspeaker Works Definitions Driver Orientation
Understanding Speaker Data DUMAX Parameters Dual Voice Coil Woofers
Other sites: True Audio Tech Topics Klippel Measurement System
JBL Technical libary HowStuffWorks.com


There are two standard parameters which are typically subject to some ambiguity:

  1. In comparing speaker parameters, be wary of the published Xmax data! Xmax is a problematic parameter. It's supposed to be the maximum linear excursion (usually expressed in millimeters, as is all the data shown in the LDSG). But there are complications just in the technical definition of this. Once the marketing departments get involved, things get even more muddled.

    The first problem is how to express it. The two methods are peak-to-peak (p-p) and one-way (+/-). Therefore, a driver which can move 5 mm either side of its rest position with reasonable linearity can be claimed to have an Xmax of either 5 or 10 mm. Both are true, except the first is a +/- spec and the second is p-p. When a vendor doesn't specify Xmax at all, the correct value can often be determined by taking the absolute value of the VC height minus the magnet gap height. This will be the p-p value. To get the +/- value, simply divide by two.

    The second problem is even more fundamental - what is meant by "linear"? No driver will ever be perfectly linear, so the spec therefore requires that the one measuring it defines the maximum allowable distortion. Unfortunately, JBL is the only driver manufacturer I've seen which publishes this (they use 10%). The more reputable hi-fi vendors simply use the point at the knee of the displacement vs. distortion curve. Some vendors, driven by marketing pressure to claim the highest possible value, define really excessive levels of distortion as acceptable. In the hypothetical example above, we can change both the +/- and p-p values by simply defining that 20% distortion is "reasonable".

    A final issue is that while many drivers are limited by their motors, some are actually excursion limited by their suspensions. Worse still, a specific driver may be limited by one factor in one direction and the other factor in the other direction. For this reason, the DUMAX system was developed (which, unfortunately, no manufacturers use in their published specs) which specifies two sets of +/- data, +Xsus, -Xsus, +Xmag, and -Xmag, representing the suspension and motor system maximum linear excursions in each direction, respectively. In the DUMAX system, the published Xmax value would be the smallest of these four parameters. In recent years, an even more precise driver test and measurement system has become available from Klippel in Germany. The Klippel system uncovers flaws in drivers which DUMAX sometimes overlooks.

    In the LDSG, I have tried to normalize the advertised values of Xmax to +/- excursions only. This is the most accepted way of measuring Xmax and is most likely to be correct when I have no choice but to simply quote the manufacturer's data because they doesn't specify the measurement method nor list the height of the gap or the VC. As noted, there is not sufficient data to attempt a DUMAX or Klippel characterization of each driver.

  2. The other ambiguous specification is sensitivity. Some vendors specify it as the sound pressure level (SPL) measured at 1 meter from the driver when driven by 1 watt. Other vendors specify it as the SPL measured at 1 meter from the driver when driven by 2.83 volts. In the case of 8 ohm drivers, these numbers will be identical. For drivers which deviate from 8 ohms, the numbers will be different.

    In the LDSG, I have normalized all sensitivity values to dB/2.83V/m. The reason for this is that most modern amplifiers are constant voltage devices - i.e. they have a low output impedance. This is why an amplifier rated at 100 watts into 8 ohms, will usually also be rated at 200 watts into 4 ohms. Therefore, using the voltage method of rating sensitivity facilitates system design since it's easier to calculate the SPL of the various drivers in a system with differing sensitivities when driven by a constant voltage source - i.e. a typical amplifier.

In the future, I plan to integrate a more complete parameter database into the LDSG. For the time being, the parameters listed in the comparison charts (Sections 7-12) are necessarily incomplete in order to fit them into an easily browsable document. However, other parameters may be derived from those given, e.g…

  • Qes = Fs / EBP
  • Qms = 1 / ((1 / Qts) - (1 / Qes)))


Driver guidelines - Tweeters:

Prior to ca. 1960, the predominant tweeter designs were either cones or horns. Cones were generally found in lower priced consumer gear and horns were more often found in audiophile systems. Since that time, dome tweeters have pretty much supplanted both. Cone tweeters have always suffered from problems with beaming and resonances, while horns have always been expensive to manufacture and tricky to get right. Domes, on the other hand, can be as cheap and easy to work with as cones, while still challenging horns on sound quality (but not efficiency).

Early domes tended to be made of hard plastics or soft treated fabric. Current technology uses a variety of materials as noted in Appendix B above.

For DIY projects, most of my respondents recommend sticking to 1" (25 mm.) dome tweeters and avoiding the 3/4" units altogether. I've generally (more on this qualified statement in a moment) followed this advice because it's easy to see why. There are few things that a smaller dome tweeter can do that a 1" dome can't. Add to that the facts that a smaller dome will almost certainly have lower power handling and have to be crossed over at a higher frequency. Still, I have included a select few smaller domes for those who want the slight edge they offer in extended and off-axis response. It's still fairly safe to recommend against using one in anything less than a 3-way system (although even this has been done successfully in a handful of commercial systems).

Having said this, I have to note that there exists a small minority of 3/4" dome tweeters that turn this advice on its ear! Most significantly, the line of 3/4" domes from Hiquphon are among the best dome tweeters in the world and certainly an outstanding value.

A fairly recent addition to the technology of dome tweeters is the use of ferrofluid. Ferrofluid isn't a single material, but a class of some 50 magnetically and thermally conductive liquids of varying viscosities which are used to fill the gap around a tweeter's voice coil, thus providing heat dissipation (i.e. greater power handling), mechanical damping, and lowered mechanical resonance. Still, some manufacturers (Scan-Speak and Hiquphon) have moved away from using ferrofluid in their top-of-the-line tweeters, claiming improved transient response and transparency.

In addition to the ubiquitous domes, there are still several other technologies worth considering:

  1. Ribbon tweeters have been around for quite a while, but have always remained out of the mainstream. A classic ribbon is characterized by very low impedance (usually requiring a transformer) and DC resistance, low sensitivity, and a dipolar radiation pattern - all characteristics which can challenge a speaker designer's abilities. Done correctly, however, ribbons can present a degree of sparkling transparency that other speaker types (except electrostatics - see below) simply can't match. Leaf tweeters are quite similar except they are typically monopole radiators. The following comments apply to both.

    Originally, all ribbons used a thin conductive ribbon (think of it as an unwound voice coil) to driver the air directly. A pure metal ribbon like this achieves its sonic excellence by virtue of the ribbon being only a fraction of the mass of a conventional electrodynamic tweeter's dome/cone and VC assembly. While such a design is still in use on the most expensive ribbon tweeters, it is also inherently quite fragile and can be easily burned out with too much power or an inadequate crossover. Fortunately, the ribbon be made quite easy and economical to replace.

    Considerably more rugged and less expensive to manufacture are a new generation of ribbon designs using planar (see below) technology. Instead of a fragile metal ribbon suspended in thin air between the magnet poles, these new ribbon/planar hybrid designs deposit a metal ribbon conductor on a thin film of some dimensionally stable plastic, such as KaptonTM or polyester. This technology is especially appropriate for some of the large wide range ribbon drivers.
     small "real" ribbon tweeter (Raven)   large wide range ribbon/planar hybrid (BGC)  "real" leaf tweeter (Philips)

  2. Planar dynamic speakers use the same technology as the well-regarded Magneplanar loudspeakers. They offer some of the same advantages as ribbons but without the same low impedance. They also typically don't offer quite the same degree of transparency, although they can often improve on more conventional dynamic tweeters. Also, like electrostatics (see below), their design lends itself to larger designs, up to and including full-range speaker systems.
     commercial full-range planar (Eminent Technology)   small planar tweeter (Hi-Vi) 
  3. Electrostatic loudspeakers (ESL's) are still quite popular for use as wide- and full-range drivers in top-end audiophile applications. They offer all of the sonic benefits of ribbons in an inherently simpler design. The downside is their size (a shortcoming shared with other planar drivers) and the requirement for level-shifting transformers and high-voltage power supplies.
     large ESL (Audiostatic) 
  4. Piezoelectric tweeters have been around for a couple of decades. There's nothing inherently inferior about a piezoelectric tweeter, but most available units have been designed for mid-fi consumer applications. A well-designed piezoelectric tweeter can easily challenge or exceed the performance of a dome. The only problems are that: 1) units which are affordable tend to not be very good, and vise-versa, and 2) since piezoelectric tweeters appear electrically as capacitors, crossover design can be critical.
  5. The Linaeum tweeter is a proprietary design which is now available to DIY'ers only from Radio Shack. Like a ribbon, it is a dipole radiator, designed to be mounted atop the main speaker enclosure(s). Some monopole versions have also been produced. At approx. $40, this has to rate as a best buy! Note, however, that all design rights to the Linaeum were sold to Aura a few years back, which has stopped further development on the technology. The Radio Shack units represent an engineering dead end with reported consistency problems. Some of the original founders of Linaeum have formed a new company, so further developments may be forthcoming from another source.
  6. The Heil Air Motion Transformer (AMT) is a ribbon-like tweeter developed by Dr. Oskar Heil in the early 70's. Difficult to describe or visualize, the AMT uses a loosely pleated plastic "diaphragm" on which is bonded a serpentine voice coil. Placed in a strong magnetic field, sound is produced by the alternate expansion and contraction of the pleats. Like the other tweeter technologies discussed above, this is inherently a dipole driver although monopole versions are available. The AMT was used with great effect in systems of the era built by ESS, but fell into decline in the 80's, although ESS still builds AMT systems. Eton has recently revived the technology in its ER4 monopole AMT driver.
Finally, there are a very few mid/tweeters that are worthy of serious consideration. Check these out in Section 3 covering wide/full-range drivers. These are especially valuable if you want to build a two-way system with a mid/bass driver larger than 6.5" (170 mm). Some of the more notable are made by Audax, Cabasse, E.J. Jordan, and Manger.


Driver guidelines - Midrange drivers:

Midrange drivers, once quite common, have become a fairly specialized niche since they're only used in 3- and 4-way systems. Midrange drivers have become a hot topic of debate, with two mutually exclusive, occasionally religious, schools of thought:

  1. Most DIY'ers use conventionally designed 3-4" (7-10 cm.) cone-type drivers. The advantage in using a small mid/bass driver is to allow the use of more gentle crossover slopes in the critical 100-500 Hz region. Another undeniable argument in favor of cone-type midranges is the maturity of the technology along with the direct applicability of lessons learned in the highly active mid/bass market.
  2. 2-3" (50-75 mm.) domes went through a flurry of activity a while back, and still seem to get some attention in new designs. The problem is that few manufacturers can really get them right. The presence of a separate spider in the case of a cone adds design flexibility (pun intended). Some notable manufacturers who seem to have found the formula are Accuton (ceramic domes), ATC, and Morel, all discussed below. In addition some people also recommended both fabric and metal midrange domes from LPG and Vifa.

Whatever the individual driver technology, midrange drivers fall into one of three competing design camps:

  1. Full midrange - drivers designed to cover the spectrum of musically interesting frequencies (including the human voice) in one driver. These are typically found in 3-way systems where only the lowest and highest frequencies are trusted to other drivers. Full midrange drivers are virtually identical to mid/bass drivers, except that they're designed to cross over to the woofer at a frequency that would be unacceptably high for a mid/bass driver.
  2. Upper midrange - drivers which split the job of reproducing the critical mid frequencies with another driver, typically a woofer or mid bass. Drivers in this category typically cross over in the 500 Hz to 1 kHz range. You'll find these in many 4-way designs. Many audio purists and DIY'ers alike agree that such designs are tricky to get right. Even then, often the sound will be inferior to a well-designed two- or three-way system where one high-quality driver is entrusted with the critical mid frequencies.
  3. LF tweeters - typically 1.5.5" dome drivers which share the HF duties with a super tweeter of some sort - usually a 3/4" dome or ribbon. These typically cross over at 1 kHz or higher and would be considered tweeters except that they lack the top octave. The same comments and caveats as noted above for upper midrange drivers apply equally here.

Finally, there are a very few mid/tweeters that are worthy of serious consideration. Check these out in Section 3 covering wide/full-range drivers. These are especially valuable if you want to build a two-way system with a mid/bass driver larger than 6.5" (170 mm). Some of the more notable are made by Audax, Cabasse, E.J. Jordan, and Manger.


Driver guidelines - Horns

Driver guidelines - Midrange and tweeter horns:

As an option to direct radiators, many people prefer systems designed around horns. While bass horns are possible (e.g. Klipschorn), they often require woodworking skills beyond the capabilities of the average DIY'er. Also, bass horns are usually implemented using an otherwise standard woofer.

Therefore, most DIY'ers selecting horn drivers will be concerned about units covered by the midrange and tweeter units in a conventional direct radiator system.

Horns are significantly different from direct radiators in the following ways:

  1. Horns are usually built by combining a physical horn (called variously the horn or horn/lens) and a compression driver. The compression driver is essentially similar to most direct radiators, consisting of a motor, diaphragm, and horn coupling. Unlike direct radiators, most horn drivers have small (in comparison to direct radiators) throats, regardless of the frequency range they must cover.

    As an exception to the discussion above, some horn units are sold only (or primarily) as complete integrated units, this easing the selection process.

  2. One critical concern with horns is the cutoff frequency. The low frequency limit of any horn assembly is determined by the size and geometry of the horn, and is pretty much independent of the driver's FR. It's therefore incumbent on the horn system designed to match the cutoff frequencies of the horns with the desired frequency coverage. At the end of the design process, you should also check to make sure the driver you've selected can also reach the low frequency limit set by the horn. If not, you can simply select a different driver while retaining the horn.
  3. Another critical concern with horns is dispersion. With direct radiators, all drivers act as a point source until they reach the lower frequencies when spatial loading comes into play. Horns have very rigid directivity which is defined by the geometry of the horn itself. The critical spec when choosing a horn/lens is the directivity, usually defined as two dispersion angles - e.g. 120 x 40 degrees - which specify the horizontal and vertical dispersion angles of the particular design. One clue provided by many manufacturers is the description of a horn as "long throw" (i.e. narrow dispersion) vs. "short throw" (i.e. wide dispersion).

    Matching dispersion characteristics will determine how a system will sound. Obviously a horn which restricts its sound in a tight beam will match a wider dispersion horn only at one fairly critical listening distance. Used in a home hi-fi system, where room interactions effect the sound, the situation can only get worse. My recommendation is therefore to use wide dispersion (i.e. short-throw) horns for 2-way designs for home use.

    If using more than one horn, things get more difficult. It may not be possible to match the dispersion patterns to assure consistent sound within the listening area. Furthermore, available horn tweeters typically have fairly narrow dispersion when compared to the options available in midrange horns. Even if everything else is right, the listening room can still mess everything up. For these reasons, most people building 3-way horn systems, will cross over the tweeter quite high (at 5 kHz or higher) to minimize apparent mismatches. This works out well, since the dimensions of a purpose-built horn tweeter usually dictate a quite high cutoff frequency.

Driver guidelines - Horn/lens and driver combinations:

When assembling your own horn system from a compression driver and a matching horn/lens assembly, the responsibility for the final sound shifts away from the manufaturer to the builder. The best driver in the world won't sound good with a poorly designed or manufatured horn - and vise versa. It is beyond the scope of the LDSG to try to identify all such successful combinations, so in Section 11, you will find a list of vendors and specific parts which have been used successfully in hi-fi- horn systems.


Driver guidelines - Mid/Bass drivers:

This is the largest population of currently available drivers, reflecting the trend toward 2-way systems with separate subwoofers. This is also an area where you'll find the greatest variety of "new and improved" cone materials technology. The two most ubiquitous materials, paper and polypropylene, have been joined by a large variety of exotic materials. The goal is to provide rigid cones, free from breakup, with a high degree of internal damping. Some of these are good and some are not so good, so let's start this discussion with some basic guidelines:

  1. Vendors like to hype their drivers with new and/or fancy cone materials. Be suspicious of the "highest-tech" drivers! Often, the manufacturers still have some secondary effects of the new technology they haven't quite worked out yet. In most cases, these drivers will require more time and expertise in the design of a suitably equalized crossover than the average DIY'er can invest. Refer to Appendix B for a thorough discussion of cone materials. It's worth repeating that materials which you may wish avoid include metals (especially magnesium), fiberglass, and Kevlar (among non-proprietary materials), along with HD-Aerogel, Hexacone, and Polykevlar (among proprietary materials).

    The most common characteristic of high-tech cone materials is extreme stiffness combined with light weight, as discussed in Appendix B. This can make a driver with excellent transient response and extremely low distortion in the passband. Where you have to look out is above the passband. Extremely stiff cone materials, most notoriously Kevlar and metal (aluminum and magnesium), usually exhibit undamped resonances and breakup modes in the upper stopband. This typically requires a notch filter somewhere in the system to completely tame. Even outside the passband, the severe phase distortion introduced by a notch filter can have audible consequences for the passband response.

    A secondary consequence of high-tech cone materials is that all crossovers may have to be 3rd order (18 dB/octave) or greater at the high end in order to help tame the resonances and breakup modes. This runs contrary to the maxim that you should always use the lowest-order crossover possible in order to better manage phase and group delay. Generally speaking, the only cone materials which can effectively use first- or second-order crossovers are based on paper or polypropylene.

  2. No matter what the box design or alignment, probably the smallest diameter driver you'll want to consider for a full-range system (or even a satellite system designed to be used with a subwoofer) is 5" (13 cm.) You can successfully use 4" (11 cm.) drivers in a D'Appolito/MTM configuration, but you may still run into problems with low values of Xmax (max. linear cone excursion) and Pmax (power handling).
  3. Speaker parameters are useful in designing systems and deciding on box types and alignments, but most important are the Frequency Response (FR) and impedance curves. A driver with nasty humps, valleys, or ripples in its passband will probably still evince symptoms of their presence after a system has been designed around it. Also important are "waterfall" plots, families of curves which depict both the time and frequency response in three dimensions. Waterfall plots will readily show undamped resonances and other design and/or material shortcomings.


Driver guidelines - Woofers:

Most of the comments above for mid/bass drivers apply equally to woofers. What differentiates the two are the selection of cone materials and the driver selection criteria. There are very few woofers which use cones made of anything except paper or filled PP. Also, as you go lower in frequency, room response tends to overwhelm the driver's free-air FR, which becomes less important than how much air it can move - mathematically the product of the effective diameter, Sd, and Xmax, previously discussed.

This brings us back to an issue related to the Xmax issue… Be aware that as we get into woofers and subwoofers, more American and autosound manufacturers (more on this below) are represented, and these are the folks least likely to publish adequate technical specs. This makes it quite difficult to verify that the published value of Xmax (if any!) is a correct +/- value.


Driver guidelines - Subwoofers:

First of all, the proper definition of a subwoofer is speaker system which only needs to reproduce frequencies in the lowest octave or two. Since human hearing and acoustics cause us to be unable to localize frequencies lower than 40-80 Hz, a stereo or multi-channel sound system still only needs a single subwoofer to create realistic sound at these frequencies. Whether a single subwoofer can realistically reproduce the required sound pressure levels (SPL's) involved is another question.

For the purpose of the LDSG, a subwoofer is defined as a woofer whose FR tops out at no more than 200 Hz. This is based on two factors which do have a relationship to the actual definition of subwoofers:

  1. This frequency range correlates to the FR of woofers of 12" or more in diameter. Drivers crossed over higher than this really should be restricted to 10" or less in diameter. Once you pass the 10" mark, the ability of such large and massive cones to reproduce the higher frequencies is severely compromised.
  2. With the addition of some of the autosound vendors (more on this below), the woofer list grew too large to comfortably fit on a single page. By dividing it in two between the entries for the 11" and 12" drivers, the comparative listings each will fit on a single page for easy reference.

Otherwise, all woofer comments above apply. Another differentiating factor is an interesting phenomenon in low frequency drivers that, as woofer diameters increase, the primary sources of supply shift from Europe to North America and the Pacific Rim.

Although recommended drivers ranging up to 22" in diameter are listed in the LDSG, you reach a point of diminishing returns once you get much past 12". As a rule of thumb, if you need to move more air than a well-designed 12" driver can do, you're often better off - both sonically and financially - to use several smaller drivers rather than one huge driver.

Another trend in subwoofers is the use of car audio drivers in home systems. While few people would think of using 5" coaxial speakers designed for automobile door mounting in a serious home system (although a select few are quite suitable for center or surround channel use), many folks turn to the vendors of the sort of 2-4 ohm, tub-thumping bass drivers used to convert cars into rolling boom boxes. This is not inherently a bad idea, but caution should be exercised. Automotive subwoofers used in a home environment generally have inferior sonic and engineering characteristics when compared to units designed for home use. Not only is the nature of the environment entirely different, the design goals are different. Engineering is the management of compromises. Automotive and commercial sound speakers emphasize high SPL's (i.e. efficiency and high power handling) along with mechanical ruggedness, often at the expense of fidelity. In researching these drivers, a pattern began to emerge where the autosound subwoofers which could compete technically cost as much as, or more than, those designed for home use. Those which did offer attractive pricing ("bang for the buck") often suffered in a purely technical comparison. Also, be aware that the Xmax measurement/specification problems noted above are particularly severe when dealing with autosound subwoofer vendors.

Another, more fundamental problem with autosound woofers is that they may simply not work well in home Hi-Fi applications. Computer modeling of some of the most impressive woofers and subwoofers in the autosound market reveals that they begin rolling off around 100 Hz. While this is clearly unacceptable for home use, it works well in a car because the "cabin gain" reinforces the bottom octaves. Conversely, the typical Hi-Fi woofer feeding into such a small volume as an automobile cabin would show a seriously exaggerated low end. Compounding all this is that autosound woofers are often designed to be used in "assisted" alignments, where countouring in the amplifier corrects for speaker FR anomalies.

This brings up the last point… Detailed technical specs are certainly harder to obtain from the autosound vendors. Most will have published Thiele- Small parameters, but few actually publish FR plots and waterfall plots are non-existent. Many often don't even publish sensitivity ratings. If you do consider using an autosound subwoofer, get all the information you can and then plug it into the recommended drivers list at the end of the LDSG to see how it compares to the other drivers listed. Only then can you make an informed decision whether it's something you really want to include in your design.

Also, when browsing autosound vendor specs, be aware that many highly-touted autosound drivers are really nothing more than custom-branded drivers from more traditional vendors at really inflated prices. For example, Diamond Audio Technologies "Hex" series are identical in all technical respects with the standard Eton "Hexacone" series of drivers. The only apparent difference is the name and the premium prices.

At one time, the LDSG included an entire section listing autosound drivers which are suitable for home hi-fi applications. As time has passed, their number has diminished markedly. So with the first 2003 release, all autosound drivers are simply included with those from other vendors.

All of the foregoing is obviously dependent on pricing. Autosound drivers are among the most heavily discounted. Unfortunately, street prices are highly volatile, so I can only include best guesses as to what an "average" price might be. In some cases, I don't even try and simply quote list prices. If you're interested in an autosound driver, the only way to judge its value is to compare it with other drivers costing the same as you can get it at the moment. One interesting aspect of selecting an autosound driver for a home hi-fi application is that the best suited drivers are typically one of the autosound vendor's lower cost lines. The reason for this is simply that the more expensive drivers are custom designed for the autosound market and often exhibit the severe LF roll off characteristic of a driver sdesigned for cbain gain reinforcement. On the other hand, a vendor's less expensive drivers are more likely to be an off-the-shelf OEM unit not specifically designed for autosound.


Driver guidelines - Exotic wide-range drivers:

Most of the cutting edge in driver technology appears in new high-tech designs for wide-range drivers. These are usually designed to cover the range from ~200 :Hz to the limits of audibility. The most significant technologies on the market are the so called "bending wave" drivers. These look deceptively like other dynamic drivers, but operate on completely different principles. In a bending wave driver, the "cone" is non-rigid, designed to flex. When driven by a more or less conventional voice coil. Rather than acting like a piston, the cone vibrates as the driven wave propagates to the edge, much like dropping a pebble in a lake. The air is moved not by the gross motion of the cone, but by the peaks and valleys of the transverse wave as it propagates from the voice coil to the edge termination. As might be expected, the method of edge termination becomes critical to avoid standing waves in the cone material.

There are three principle types of bending wave transducers in common use today:

  1. The Walsh driver, originated by Ohm Acoustics speakers, is available primarily to DIY'ers from German Physik's line of "DDD" transducers. Walsh drivers use steep cone angles and are mounted vertically, creating a cylindrical wave front. Walsh drivers can cover the full musical range, but become quite inefficient when called on to reproduce low bass. The DDD drivers are not currently included here since they cost over $2,000 each and DIY information on them is scarce. Replacement Walsh drivers used in the Ohm Acoustic systems are available from Orange County Speaker Repair.  DDD driver (German Physiks) 
  2. Somewhat similar are the Manger MSW (Manger Schallwandler Wideband) wide-range drivers which cover ~150 Hz to 20 kHz. MSW drivers are thin disks, approx. 8" in diameter, using a flat diaphragm of proprietary clear plastic construction rather than a more conventional cone. The voice coil attaches to the diaphragm midway between the center and outside edge. High frequencies are mostly generated in the area inside the voice coil attachment, while lower frequencies are mostly generated in the outer area. MSW drivers are easily recognized by their large diameter and unique 9-pointed star edge absorption treatment.  MSW driver (Manger) 
  3. The newest bending wave technology comes from a company called NXT. NXT drivers are akin to MSW drivers in their general physical layout - a coil driving a flexible diaphragm - but their operation is more complex. Rather than relying on concentric ripples as the MSW does, the NXT panel vibrates in multiple modes. This is somewhat hard to visualize, so for a more detailed explanation, visit NXT's web site.

    Currently, the only NXT drivers available to the DIY'er are the Peerless SoundPanels, but these are not really hi-fi, being designed for use in suspended ceilings.. Another way to get a small NXT panel is to tear apart some Benwin multimedia speakers.


Driver guidelines - Exotic woofers and subs:

Some of the cutting edge in driver technology appears in some new high-tech designs for low-frequency applications. Even more significantly, some of these drivers are available for DIY use. The two newest technologies are:
  1. Rotary transducers have proven very effective in reproducing incredibly low bass at high SPL's. The operating principle is as simple as it is obvious. Since a conventional speaker is simply a linear DC motor attached to a cone, why not use some sort of traditional rotary DC motor with some sort of rotary-to-linear translation mechanism? The upside of this is that you can potentially achieve excursions that would be impossible with linear motors. The downside is that the translation mechanisms tend to be difficult to design, noisy, and failure-prone. Still, such drivers do exist. The best known such system is the "ContraBass", produced by ServoDrive, and designed by Thomas Danley. The ContraBass uses a conventional rotary motor along with a belt-drive motion translator to drive dual cones. Capable of delivering 114 dB at 16 Hz, the ContraBass is used wherever truly awesome amounts of low bass are required. Some of these include Disney World, Earthquake and Kong at Universal Studios, The Mirage Volcano (Las Vegas), Buccaneer Bay at Treasure Island (Las Vegas), and the Mt. St. Helens Visitor Center. A couple of years ago, ContraBass kits were available for a limited time to DIYspeaker list members.

    The only other rotary driver ever available for DIY was the "Cyclone", another Danley design, produced by Phoenix Gold (PG). Unfortunately, it's no longer available. An odd looking driver resembling a ducted fan, PG tried to sell it into the autosound market, where it was eminently unsuitable.

  2. Recently a startup company named CoDrive patented a new single-motor, dual-cone technology for a proprietary line of drivers. Even more than with the rotary transducers, the CoDrive(TM) drivers are difficult to describe.  view 1   view 2  Visit CoDrive's web site for more information on their 12" and 15" drivers.

    Both of CoDrive's drivers are available for DIY use. Although they were released almost two years ago, they're not currently listed in the LDSG because I simply haven't had enough feedback on them, nor detailed information about them.


Driver guidelines - Full-range and wide-range drivers:

Not generally of interest to many U.S. DIY'ers (other than for autosound applications), there is considerable international interest in drivers which can do it all. Whether your interest is building a single driver reference system or multimedia (computer audio) speakers, there's something for you out there. Unfortunately, these drivers are the hardest to spec well. Their complex construction often leads to anomalies in their FR curves. The most common technologies are:

  1. Conventional drivers can be made to cover an extremely wide range by careful engineering. The problems are simply compounded, though. The optimum cone material and profile to achieve good LF response won't usually be optimal for good HF response, and vise-versa. To avoid IM distortion, the HF radiator should make much smaller movements, unaffected by the displacement of the LF radiator. As a result, these are mostly limited to small (4" or less) diameter multimedia drivers.
  2. Dual-cone drivers are more or less conventional drivers with two cones, a small cone and the main cone. The small cone (traditionally called a "whizzer" cone) is attached rigidly to the voice coil. The main cone is also attached at the voice coil, but with some built-in compliance. The compliance can be part of the cone, the adhesive, or something else entirely. The theory is that at high frequencies, the woofer cone is decoupled and only the whizzer cone is driven. This doesn't address the issue of LF IM distortion in the whizzer cone. Sometimes, a dual cone driver will only cover the bass and midrange, relying on a coaxially mounted tweeter (see below).
  3. Coaxial and triaxial drivers use more or less conventional technology stacked together. A separate physical tweeter is mounted in the middle of the woofer cone. Even more complex are triaxial drivers with three separate motor assemblies shoe horned into the middle of the main cone. Aside from mechanical complexity, a disadvantage of co/tri-axial drivers is the fact that the HF units physically block the radiation of the LF unit.
  4. Coincident drivers are similar to coaxial drivers, except that the tweeter motor is recessed behind the LF motor, enabling a smooth transition from a tweeter horn seamlessly into the profile of the woofer cone - the woofer cone effectively becomes part of the tweeter horn. Several British firms (Tannoy and, more recently, KEF) have made quite credible drivers using coincident technology. Done right, this is probably the best non-exotic technology for wide/full-range drivers.
  5. Exotic drivers abound in this market segment. Which is one factor making direct comparisons of drivers more difficult. Perhaps the most credible are the bending wave drivers previously described.

The best guideline for full-range drivers is good word-of-mouth. In the LDSG, I have collected data on all full-range drivers which have been recommended by experienced DIY'ers. One final technical note is that these drivers typically are not very large in diameter and the bass cones do not have a large Xmax. Vented enclosures or large horns are the preferred cabinet alignments.

Also, note the distinction between full-range drivers and wide-range drivers. A wide-range driver is similar in concept to the ubiquitous mid/bass drivers except it covers a wider range. Wide-range drivers generally come in two flavors:

  1. LF units designed to be used with a supertweeter. These generally are designed handle all of the bass, plus the treble up to 4-10 kHz.
  2. HF units designed to be used with a (sub)woofer. These generally are designed to handle all of the treble, plus the bass down to around 100-200 Hz.

One additional note is that most dedicated DIY vendors of full/wide-range drivers have little or no North American distribution network. U.S. and Canadian DIY'ers may have to order these from the country of origin (usually in Europe). Where this is the case, and due to fluctuations in currency exchange rates, pricing is only approximate at best. Where North American distribution is available, the vendors typically are not dedicated to full/wide-range drivers, and the individual drivers are likely to be cataloged as autosound drivers.




©1998-2007 by Bob Stout, all rights reserved

Tested with: Get Firefox Get MSIE .

Support: Rapid PHP