Speaker Project One

I am obsessed with music. And thus I am also obsessed with instruments, music making equipment, and audio reproduction equipment. I am drawn to and fascinated by the beauty present in the science, design and function of these and similar things -- a Nagra field recorder, a Buchla 200 Series modular synthesizer, a stereo pair of Quad 57 speakers, or a Larrivée acoustic guitar.

In this post I will share my experience designing a three way stereo loudspeaker system, one which I am perpetually exploring, though I am nearing the completion of what I consider to be phase one of this project. I have assembled speakers from kits in the past, primarily using Fostex full range drivers. Prior to that my only hands on experience with building audio gear was extensively modifying a pair of servo amplifiers designed to drive Acoustat X electrostatic panels. For the sake of simplicity, I will refer to these speakers as Project One.

A caveat, I am bad at math. In addition, I do not own any high end test and measurement equipment aside from an oscilloscope and a multimeter. I intend to invest in such gear at some point, though until this point I have been using my ears, and intuition, both of which I feel are strong, to guide my explorations. In the mean time these two websites provide measurement data of many drivers:    https://www.dibirama.altervista.org   ||   https://hificompass.com/en

I am not intending to brag in mentioning the following. I am quite sensitive to sound, I have perfect pitch and thus am musically gifted, I can hear to 20khz at 36 years old, and as a result of having the auditory response that I do music is a primary obsession of mine. One outcome of this is that I have spent the past couple years testing tens of thousands of dollars worth of speaker drivers in order to conceive of a speaker which achieves the utmost performance, using contemporary and relatively affordable parts. These are a few speakers which I have owned or auditioned and am fond of, which serve as a sonic reference for what I'm building, without elaborating more: Quad 57s, Acoustat Xs, Apogee Studio Grands, Tannoy 12" Monitor Silvers,  Fostex 8" full range drivers mated with ribbon tweeters, Pure Audio Project open baffle speakers, Sigfreid Linkwitz's personal pair of LX521s, Meyersound Blue Horns, Genelec 8351s, Magico M6s, Kharma EV1s, Stax earspeakers, and the list goes on. 

When designing Project One, I had originally planned to build a two way speaker featuring a pair of 15" woofers per channel and a 2" or 1.4" compression drivers mated to large horns covering the rest of the range. I had aspirations to eventually settle on a premium driver like a TAD TD4001s for the horn section. Upon testing contemporary gear against older designs, though not the TADs, I realized that there are substantial advantages to certain new drivers compared with certain old drivers, and in some instances cheap drivers compared with expensive drivers, for my application -- near to midfield hifi audio reproduction; not to be confused with high output stadium or theater sound reinforcement. My exploration began with testing drivers intended for the latter application only to eventually determine that lower power handling drivers sound better.

The first drivers which I tested, at the recommendation of Al at USSpeaker, were a pair of Beyma 12XA30ND 12" coaxial drivers featuring a compression tweeter. After receiving them I hooked them up to a passive 1200hz Beyma crossover and had a listen. Not half bad to be honest, and actually quite nice sounding drivers. If one wants to build a simple speaker that is performative and robust, the 12XA30ND is a very good choice and an outstanding value. That said naiveté and curiosity got the best of me and I was quickly on to the next pair of drivers. Shortly thereafter I had established a habit of testing new drivers on a weekly basis. It has been now two years... the drivers tested include those from Fostex, Seas, Satori, Visaton, Beyma, Faital, BMS, Scanspeak, Ciare, Audax, Supravox, 18Sound, and Mark Audio among others.

Rather than provide an 80 page explanation of all thoughts associated with this endeavor, I will list each driver that I tested and provide a concise impression of each. This list is divided into three driver types -- compression drivers and tweeters, mid woofers, and bass woofers. Within each section I put the preferred and most unique drivers at the top of the list and the more blah drivers at the bottom. I will interject at certain points with brief summaries of my findings relative to specific driver types. 

For the compression drivers I have had a variety of horns on hand including wood biradial Yuichi horns, and various 2", 1.4" and 1" aluminum and plastic horns, and my favorite a pair of Ciare PR614 1.4" horns. For the woofers I tested the large woofers in an open baffle enclosure which I fabricated, and many of the smaller drivers in a raw unmounted state. Simply, I find that open baffle designs mitigate both cone and cabinet resonances / standing waves to a far greater degree than closed and ported cabinets. Eliminating resonances is priority number one when designing speakers. Fortunately some designs like open baffles and dipole planar speakers are inherently non-resonant if implemented properly.

For Project One, I custom fabricated various swappable baffles with several different hole diameter cutouts in order to test other drivers besides the 15" woofers that I have on hand. Here's a screenshot of some baffle driver layouts which I mocked up in Illustrator: 

For crossovers I started with passive off the shelf crossovers from Beyma and BMS. Early in my testing I purchased a Sublime Acoustic K231 three way active analog crossover which is what I use today. This device of exceptional quality and value has proven to be indispensable. All drivers are actively amplified using Boxem Arthur Purifi Class D amps for the woofers and mids ~500w per channel depending on driver impedance load, and a Topping LA90 50w Class AB amp for the HF section. I intend to get more granular with a digital crossover, eq, mics and possibly additional test equipment. Here are my findings:

Compression Drivers

Tested at crossover points ranging from 300hz-2000hz on various horns:

-Great Plains Audio Altec 288c -- Very pleasant, musical midrange, the long throat exit of the driver yields a somewhat nasally and compressed sound compared with other drivers which feature a shorter exit and thus a more direct crisp sound. Despite this it is superior to most drivers that I tested and has a particularly sweet tonality.

-Faital HF108 -- Very good no frills sound in a compact and robust package. Some compression drivers from Faital and other brands exhibit shrillness and honk which this one does not. Good extension in both directions for the size.

-BMS 4552 // 4540ND -- Ultra clean high frequency compression driver with no objectionable qualities. Does not have quite the midrange presence of the HF108, but is more tactful above ~2000hz.

-Beyma CP755ND -- An overlooked beast of a driver. Not quite the visual fit and finish of the HF108, and quite barebones in design and build quality. That said this driver produces treble and mid frequencies with unrivaled energy and clarity. There is a shimmer that this driver presents which is by no means harsh nor unlistenable but is characteristic of this driver alone. It reads as ultra detailed presentation which some might find objectionable. At only the highest volumes at which I don't listen, I presume this characteristic becomes distortion of the diaphragm and its interaction with waves reflecting off the driver's rear cover. The BMS4552 is a different flavor of detail which rather than shimmery, I would describe as clean and liquid -- the difference in diaphragm materials between these two contributes to this difference in sonic character. The Beyma is unique and versatile in that it can be crossed quite low, at 400hz and extend to 20khz. I believe a very powerful 2.2t magnet, well designed low mass diaphragm with a thin low modulus and thus low mechanical damping surround, proper phase plug and short driver exit, are responsible for the uniquely superior sound of this unit. I prefer this driver to all others tested.

-TAD TD4001 -- This is the most accurate compression driver that I have tested thus far. At 400hz and up it presents crystalline clarity throughout its operating range. The stuff of legend. These are enormous and feel dated physically compared with comparatively small highly performative drivers like Beyma CP755ND and the BMS 4552. It is difficult to improve upon or exceed the performance of the TADs for playback above 400hz due in part to the TADs relatively large ~4" very low mass beryllium diaphragm. The Beyma and BMS are in some respects equal or more performative. The only places I see possible room for improvement with the TADs are incorporating a low mass low modulus diaphragm surround material; a denser magnet structure consolidating magnetic currents and increasing efficiency; and possibly alternative geometries such as a shorter throat. In my testing, anything other than an annular phase plug, and certain ring radiators, sounds downright bad. The CP755ND is the only alternative that I've found to the TAD thus far. The simplicity with regards to its cost, size and availability compared with the TAD renders the TAD less desirable for a utilitarian speaker build. That said, such a large curved radiating surface producing 400hz-20khz frequencies is unique and outstanding. The TD-4001 is the only compression driver that I know of with a ~4" radiating surface capable of relatively flat extension to ~20khz at midfield distances. Other drivers with heavier beryllium diaphragms of the same diameter, constrained by the laws of physics, are rolled off.

For testing I mounted the TD-4001 to a B&C ME75 constant directivity horn and AB compared it with the CP755ND mounted to the same model horn using a 1.4" to 2" adapter. The TAD is extended to the point of providing ample treble. The CP755ND is more extended. Tonally the TAD is a hybrid of the CP755ND and the 4552. The Beyma is louder and more authoritative, though more homogenous and less separated. The BMS is more liquid with ultimate clarity throughout the treble region. The treble of the CP755ND with titanium diaphragms is somewhat hashed, which like the 4552, the TD-4001 overcomes. The BMS 4552 while very clear, lacks the presence of the TD-4001. Horn loading aside, the sonic scale of the BMS with its ~1.75" diaphragm is somewhat diminutive compared to the TADs very large ~4" diaphragm. 

Edit: Experimenting with a 400hz crossover with the TADs in a three way configuration sounds better than the 1200hz crossover that I was using previously with the Beymas. A three way implementation with a low crossover for the sub and mid driver is essential under all circumstances -- elaborated upon later. The CP755ND is the closest alternative to the TADs that I've found. The Beymas are a more modern and direct sound whereas the TADs are smooth and lush. One is seemingly more correct, the TADs, though not as engaging as the Beymas with their higher magnet strength motor. Whereas most if not all other drivers present the sensation of a sonic backdrop, the experience listening to the TADs is peering into a deep well from which sound emanates. It is most holographic and nuanced. On par with the best electrostatic transducers I've heard. Combine these qualities with the energy of the CP755ND and you have a perfect driver.

-Yamaha JA6881B -- Good midrange compression driver. Very much an evolved Altec 288. Very clean sounding, though output and sensitivity is low compared with contemporary drivers. I attribute this to the use of stiff beryllium and or some other stiff material in the otherwise interesting diaphragm surround "fingers," and a relatively weak 1.6t ferrite motor. The power handling for this driver is less than 20watts. Such a driver is desirable in its ability to be crossed over low. Though crossing this particular driver below ~350hz one risks permanently damaging the advanced yet very fragile and unattainable diaphragm. An excellent choice if crossing at 400hz or above for upper midrange coverage and perhaps the best of all the midrange CDs I tested, though too niche and subdued for my needs. I believe that even in the midrange at 400hz and above the CP755ND is superior. 

-B&C DCM50 -- Very good sounding paper cone midrange compression driver. In my tests I prefer it tonally and performance wise to the Celestion AXI2050 and Faital HMF200, along with other mediocre mid range compression drivers. For lowish output playback, with some equalization and horn loading the DCM50 can be crossed at ~100hz. Serviceability seems to be an afterthought and therefore not worthwhile in my opinion.

-Celestion AXI2050 --  Very clean midrange though lacking detail compared with an extended driver like the Beyma CP755ND. I observed a resonance which occurred at a certain frequency where it seemed the back plate of the driver was vibrating synchronously with the diaphragm and or room causing a major and very audible resonance peak. 

-Faital HMF200 -- Good sounding and very well made driver though lesser sonically compared with the B&C. All three mid drivers are close in sonic performance. The external build quality and design of the Faital is my favorite of the three.

-Atlas PD5VH -- Beautiful tonality. Lush and clear. Features a phenolic impregnated linen or cotton diaphragm. Craps out below 100hz and needs an enormous horn to play at adequate volume if crossing over at such low frequencies. A driver of this nature is clearer than any woofer if implemented properly. I am hopeful to see this formulation developed further to allow for lower extension and louder playback as it is tonally as good as anything else tested. In EV Cobreflex horns the tonality becomes a good bit more nasally and less pleasant.

-Beyma CP22MF -- Outstanding bullet tweeter.

-Beyma TPL200 -- Very good sounding AMT tweeter. Does not have the energy and authority of top compression drivers. Treble is perhaps slightly more refined.

-18Sound 1480A -- Very good. Lacks the clarity, extendedness in both directions, and absolute resolution of the CP755ND.

-Beyma CP750TI -- Okay

-Beyma CP385ND -- Okay

-Visaton TL16H -- Okay

-BMS 4591 -- Meh

-B&C DCM414 -- Meh

-Beyma CD11ND -- Meh

-Beyma CD10ND -- Meh

-Beyma CP380M -- Meh. Poor build quality.

-Beyma CD1ND -- Blah. Poor build quality.

-Altec 290 -- Blah

-B&C DE980TN -- Blah

-B&C DE500 -- Blah

-Faital HF206 -- Blah

-Faital HF1440 -- Blah. Major spiky resonances likely due in part to implementation of a radial phase plug.

-JBL 2453H -- Blah

I will update the CD list upon testing new drivers and or determining new findings.

In summation, a compression driver is most logically implemented as a tweeter. A compression driver horn combination is advantageous over direct radiating tweeters, amts, ribbons, etc. in that they project hf content into air / space with far greater velocity and controlled directivity, and thus further distances than a conventional direct radiating device. This is important as hf content rolls off faster than mids and lows. One can think of a hf waveform with its short fast wave length as more fragile and easily impeded by air, compared with more forceful mid and low frequency waves. Thus lf waves travel greater distances through air versus hf waves at equal volume. Air over distance from a sound source, as it effects sound waves, acts as a low pass filter. Increasing the velocity of hf output through the incorporation of horns and compression drivers for midfield and farfield sound reproduction reduces this effect and is thus worthwhile. 

Equally advantageous is the high sensitivity motor, low mms diaphragm, and greater radiating area than a conventional tweeter, characteristic of many compression drivers. Often, this in turn yields unmatched transient response, clarity, and presence.

Doubling the radiating surface area of the upper frequency region by adding a second compression driver and horn per channel, or using a Y adapter, increases the presence of treble frequencies. This is the same in principle to a 1.4" compression driver with a larger diameter diaphragm than a 1" driver having more treble presence. A degree of enhanced detail is also observed with said implementation due in part to the halved driver impedance and increased sensitivity. For midfield listening purposes this is likely excessive and unnecessary with relatively larger diaphragm drivers like the CP755ND though maybe less so with smaller 1" drivers like the 4552.

Like high frequency compression drivers, midbass and bass compression drivers mated to horns can be advantageous over direct radiating woofers in boxes as they authoritatively present a unidirectional audio signal with high directivity which translates to minimal resonances and room interactions. In addition these designs can yield greatly increased sensitivity benefitted by both the driver and horn. Often times however, a proper implementation of a horn at below certain frequencies becomes a substantial endeavor due to size, material expense, and cost.

Midrange / Midbass Drivers

This list is semi chronological. I tested the following full range and band limited at crossover points ranging from 60hz-1200hz:

-Beyma 15P80Fe/N -- These were the first pair of 15" woofers which I purchased to test in a two way speaker crossed at 800hz. They could be used as subwoofers as they produce quality bass. I would not use a driver with mms and damping specs such as this for a two way near/midfield speaker as they do not produce adequately quick transient response to be considered a mid woofer. The result is a slow honky midrange lacking detail, clarity, and dynamism. 

-Beyma 15G40 -- These were the second pair of 15" woofers which I purchased to test in a two way design. Again inadequate and even slower and more honkey midrange. Again these would serve well as subwoofer drivers crossed below 60hz with the right amount of amplification power.

-Beyma 15MI100 -- Discontinued lowish mass 15" woofers. The transient response and detail of this driver compared with the former two was a revelation. At this point I began to realize the benefits in sound quality achieved through low mms efficient motor woofers typically associated with lower power handling non-pa drivers. After observing this audible characteristic, I scrutinized tons of drivers and their specs to find what I was looking for.

-18Sound 15W500 -- A major upgrade in fidelity compared with even the 15MI100. The cone of this 15" driver weighs a mere 55g. Highly coveted drivers like the original Altec 416s and EV15Ms are some of the few 15" drivers ever made to feature similarly low mass cones. In a two way system the resultant sound quality is quite good. For mid to large size public spaces this is the driver I would chose to cover the bass and mid bass frequencies, for convenience sake, as 15" moves a lot of air compared with smaller drivers. That said under no circumstances will a driver like this, implemented as a midwoofer be as performative as a smaller lower mass driver / cone within a certain listening distance range. 

At this point in my testing, in addition to testing woofers within the intended crossover band, I began also testing them full range, to determine which yielded the best micro details and audible transient snap, and also the least amount of honk, mud and objectionable resonances. I found that these characteristics are most directly correlated to cone mass. Lighter cones operating within their capable amplitude range produce substantially less objectionable resonances compared to heavier ones. Keep in mind this is for up to midfield listening distances.

-Visaton B200 -- This driver delivers outstanding detail and refinement. All other drivers that I have tested thus far, 8" or larger are brutish in comparison to this one. It is not as composed as the later mentioned Celestion or Zenith at the loudest volumes in the bass region. If crossed at or above 100hz and or horn loaded it becomes very suitable for mid field applications. Adding a pair of these per channel reduces the strain on the driver at a given output and permits lower crossover points for midfield application. This is my preferred 8" midrange driver for near field listening and midfield listening if a pair per channel is used. Visaton deems the B200 a full range driver as it yields outstanding sound throughout the audible frequency spectrum at low to medium volumes. At 9g mms, in addition to losing composure in the bass at very loud volumes, it becomes a bit shouty and peaky in the upper frequencies with the volume turned up. Despite being low mass and paper this driver is a successful implementation of pistonic operation yielding very low distortion and clear sound within a certain output range. The B200 produces snappier mid frequencies than any of the 8" or larger drivers tested. The only 8" drivers superior to this in my estimation are far more expensive featuring equally low mms cones and very strong magnets. A 10" version of the B200, to be crossed at ~50hz-800hz would be perfect in my opinion.

-Fostex FF165WK -- Outstanding. Comparable to the B200 though smaller and thus less versatile for midfield applications unless multiple are used. Very composed.

-Fostex FE126NV2 -- Outstanding. This is technically a full range driver though it yields outstanding sound throughout the audible frequency range. It is suitable as a midfield driver as it is more robust than the Visaton B200 in the bass region despite being smaller. A pair or quad array of these drivers would match or exceed many if not all midrange implementations for mid field listening when crossed at 100hz or above. A very special driver and the sweet spot within the Fostex lineup in my opinion.

-Celestion G12H Heritage -- Outstanding. Contemporary reproduction of a highly coveted 12" vintage guitar speaker from the 70s. One of the best drivers tested, particularly musical, and seemingly the most solid of the lot. I prefer the naturalistic and simple choice of materials here which are of a higher degree of class compared with some of the more plastic, glitzy and toyish though often more pricey drivers on this list. Outstanding tonality with electric guitar passages as was expected. Very musical and warm though not as resolving as the Visaton B200 or similar. If bass and musical sweetness are a priority over clarity and detail I prefer the character of this driver to the slightly more resolving Zenith mentioned below. This and other high sensitivity Celestion 12" guitar speakers sound very good, are robust and are perfectly suitable for utilitarian high performance hifi applications. This driver is not suitable for a two way without subs as it does not produce adequate bass. Above ~60hz it is an outstanding choice. 75hz fs version tested.

-Zenith 49CZ852 -- Excellent. 12" 1950's console woofer features a 21g mms cone. It receives high praise on various forums and websites including glowinthedarkaudio.com. I'm a skeptic and didn't want to believe the hype. The Visaton B200 is more nimble and clear. At midfield levels the larger diameter Zenith is more composed. The deep cone of this woofer benefits its sensitivity.

-Seas FA22RCZ -- Excellent. This is technically a full range driver and it yields outstanding sound throughout the audible frequency spectrum. It is suitable as a midfield driver as it is more robust than the Visaton B200 and can easily produce 60hz bass tones at reasonable levels. The whizzer seems to be effected by low frequency content in that it produces some additional resonances which are subtle though observable. The FA22RCZ doesn't present quite the same degree of midrange refinement and clarity as the Visaton B200. 

-La Voce WSF121.70G -- Excellent 12" guitar speaker. Similar in appearance and tonality to G12H at a lower price point.

-GRS PT5010-8 -- Very good planar magnetic mid panel / tweeter. I occasionally read reviews which tout this and the B&G Neo 8 which are supposedly identical as the best mid driver around. It is a very good driver, though not superior to the aforementioned Visaton or Seas, and perhaps other low mms fullrange drivers. The GRS is advantageous in that it produces uniform front and rear response, likely produces very little distortion at normal listening levels, and is of a very solid, simple and utilitarian build quality. That said, it is not as dynamic, snappy and transparent as Visaton, and it also lacks the extension of the Seas. However, it can be crossed at 100hz no problem. A great choice in my opinion though I prefer the aforementioned 8" paper cone full range drivers to this one. Electrostatic transducers and true ribbons will be more detailed and produce the least distortion of all driver types though are limited by their output and extension capability.

-Faital 10FE200 -- Very good mid range driver. Quick enough though a bit lazy compared with certain 8" or smaller full range drivers. Lush tonality.

-Ciare -- Excellent quality in general. Hit or miss fit, finish and occasionally a very tacky logo on the dust cap of certain drivers.

-Celestion CN0617M -- Excellent. Very high magnet strength intended for horn loading applications. Clearer above 300hz than any of the other drivers tested at the expense of lack of extension below 300hz. Not suitable even at low output levels below 300hz as it is non responsive to bass input. Mediocre fit and finish.

-Supravox 285-2000 MK2 -- Very good sound. Crap build quality on the alnico 285 which snapped in half in transit.

-18Sound 12W500 -- Very good. Excellent choice for a compact 2 way. Produces adequate bass compared with Celestion G12H.

-ER Audio Mini Panels -- Very pure sound though too fragile with sonic artifacts causing clicking of the panel enclosure even when excursion is reduced with 200hz crossover. Electrostatic transducers yield the most detailed, composed and lowest distortion presentation possible though often their limitations outweigh the benefits of such drivers.

-Dayton Exciters -- Good and genuinely synthetic tonality when attached to polystyrene foam insulation panels. Unique omni directional presentation. The extensive surface area of open cell insulation foam made from molecularly very rigid polystyrene, yields a very loud substrate to transmit sound through. This happens in a continuous and three dimensional, non linear non pistonic manner, and simultaneously a linear manner. Upping this methodology, though likely quite difficult to implement, an aerogel panel or similar with a relatively low modulus, with its even greater surface area and lower mass than poly foam, would yield many orders of magnitude greater passive amplification. In addition, a translucent enclosureless sound radiation substrate such as this as a singular full range driver, suspended from a mic stand or similar, would be quite beautiful and minimalistic if designed properly. Far more interesting and elegant than a pair of M2s in my opinion. One can dream, and use clear corrugated poly carbonate in the meantime.

An alternative approach would be applying an electrical signal directly to a sound radiation substrate to excite a material which in this instance would have to be conductive and ideally an open cell lattice with a high surface area at the molecular level.

Visaton FR13 -- Preferred option for near field design above 200hz

-Beyma 12XA30ND -- Excellent

-Purifi 6.5" PTT6.5M08 -- Excellent

-Mark Audio MAOP11 -- Excellent

Visaton FR 6.5" -- Good

-Supravox 285 GMF -- Good

-Fostex FE208E Sigma -- Okay

-Seas U22REX -- Okay

-Seas M15CH002 -- Okay

-Scanspeak 10F/8424G -- Okay

-Beyma 10G40 -- Meh. Shouty.

-Beyma 10MI100 -- Meh. Shouty.

-Beyma 8MI100 -- Blah

-18Sound 10MW400 -- Blah. Seems there are different versions. Mine has a concave dust cap. Shouty.

-Eminence Legend 1028K -- Blah

-Jupiter 10LAP -- Blah

-Audax PR170Z0 -- Blah. Far less sensitive than rated.

-Oberton 10M150 -- Blah

-B&C 320K/C -- Blah

-Fostex FE206NV -- Blah

-Faital 15PR400 -- Blah

-Satori MW19TX -- Blah

My preferred crossover point for a three way speakers upon testing is between 0hz-60hz/100hz for sub bass, 60hz/100hz-1200hz for mids, and 1200hz-infinity for highs. I find that similar to the advantages of crossing a low mms compression driver around 1000hz or lower depending on the design, the same applies for mids. Crossing an 8" driver with an mms of 15g or less,  or a 10" driver with an mms of ~20g or less, at say 60hz improves the clarity and performance of a speaker substantially. The simplicity of direct radiating drivers for this approach rather than trying to do the same with a compression driver like the WE555 and a 15A horn, speaks for itself. I would guess that the performance of this approach, assuming the right drivers are used, is equal if not superior to the Western Electric system. I have not heard a WE set up and intend to explore that approach eventually, perhaps converting an Atlas/Monacor driver to a field coil and swapping the diaphragms for the nicest WE555 replicas I can find. I prefer experimenting with readily available drivers in most instances rather than unobtanium.

Back on topic, the Seas FA22RCZ for example handles 60hz sonic instances at normal to loud listening levels with no problem. At the maximum rated power handling of this driver, a continuous 60hz wave will begin to strain the driver sooner than later. Under no listening circumstances would I ever find myself presenting such a task to any 8" driver. Also adding a sub rather than running the driver full range to 20hz enables it to perform better. This principle applies to all drivers. The Visaton B200 is not as robust as the Seas and is better suited for an 100hz crossover at a minimum. Keep in mind the octave range in hz is double, and thus the 40hz difference between 60hz and 100hz is very substantial. The excursion range of a 60hz wave form is much higher than 100hz at equal volume. A pair or quad of Fostex FE126NV2s for example, which handle low frequencies quite well, with a specified crossover range of 100hz-1200hz, would be an excellent choice for the mid section of a three way near or midfield speaker. Additionally advantageous is the low impedance achieved from wiring all four drivers in parallel assuming ones' amp can handle a very low impedance load. This configuration would outperform most ultra high end drivers at a fraction of the cost.

I prefer paper cone full range drivers of high efficiency, low damping, low mms, and often low power handling. This formulation contributes to a nuanced presentation and dynamic range that no other cone variety can match. Furthermore a smaller well designed magnet that saturates little with a magnetized pole piece yields greater sound quality than drivers with over built magnet structures in my observation. There is presumably a sweet spot of magnet volume / mass as it correlates to the characteristics of a given driver. And this size tends to be far smaller than one may intuit and or what the industry seems to have adopted as the standard.

All drivers have a sweet spot of volume range and frequency range within which they will perform their best. This is directly correlated to the various parameters associated with a driver. The easiest place to start is power handling, sensitivity, mms, total damping, and motor strength. Lower power handling drivers are better suited for home audio and or midfield speakers as their optimal volume range as it relates to performance is achieved with relatively low wattage. 

In addition, drivers also have a sweet spot correlated to the operational lifespan of a driver. This sweet spot varies from driver to driver, and also varies depending on the average intended playback level throughout that lifespan. For example a drivers like the 10MI100, 10M150 and 10MW400, all of which are intended for high output playback, are presumably excellent drivers when operating at their intended output level -- high output in a midsize live venue. At home listening levels it will take ages for these drivers to open up -- the spider and suspension loosening up. B&C drivers and also especially the Oberton 10M150 seemed very non-pliable out of the box -- thick stiff spider and surround. After 1000 hours of playback at loud levels, this characteristic is presumably diminished.

Voice coil / cone excursion and cone diameter are correlated and determine the frequency range within which a driver can operate without distorting excessively or failing. For example, no commercially available 4" driver can play to 50hz. The ones that are remotely extended in the bass region feature massive excursion. Most drivers of this diameter won't operate effectively below 200hz at midfield levels. I find that excessive excursion especially with subwoofers can yield a vacuum effect with regards to energizing / pressurization of the listening room in a way that is distracting, unnatural and thus undesirable. To avoid this and other issues associated with using small drivers for loud playback of low mids, bass and sub bass, opt for larger drivers when possible.

The notion that a midrange specific driver is necessary for home audio level playback is futile. In every instance a low mms full range driver of the same size will out perform a "dedicated" midrange driver. Also a midfield or nearer dedicated 6"-10" mid woofer, meaning it does not operate from ~100hz-10khz or greater, intended for low power handling applications, is a poorly designed driver. In my experience thus far, fast transient response correlated to low mms and high motor sensitivity supersedes inconsequentially low distortion specs as it relates to perception of sound quality, though the latter spec is worthwhile to optimize for. This can also be observed in a frequency response chart; the more extended a driver is in both directions, typically the more performative it is from a sonic quality standpoint. Poorly implemented examples of such extended drivers will have a ragged frequency response, inherent when damping is reduced, a more fragile cone is implemented, and thus faster transients are achieved, all of which can induce higher distortion compared with a higher power handling driver of an otherwise similar nature.

Reiterating, every driver has an optimal operating range that relates to its intended application. While the Beyma 15G40 may not sound good at 100w in a living room compared with the lower power handling 18Sound 15W500, the inverse is likely true when higher output levels are called for assuming the proper crossover point is used.

For my purposes, the best sounding drivers under a certain diameter, within an intended listening distance / power handling range, are full range drivers. A pair of FE103A's at 3' in small bass reflex cabinets sound better than Meyersound Blue Horns.

Subwoofers

I tested the following at 60hz and 100hz 24db/oct crossover points and stress tested all drivers with 20hz, 30hz, 40hz, and 50hz continuous waveforms at various amplitudes:

-18Sound 15W500 -- Favorite.

-Visaton W250 -- Outstanding within its power handling range with extended low frequency response. The chalky tonality of this paper cone, much like the 18Sound, is much to my liking. This driver produces bass notes with an ease that I have not heard from any other driver. It seems to be extra sensitive. Exceed the amplitude capability of this driver, which occurs only at fairly loud listening levels, and it will crap out on you. I don't listen at levels where this is an issue though I prefer the 15W500 enjoying the knowledge that it can handle whatever I throw at it. To my surprise the W250 presents lower bass with greater ease than the 15W500 as it is a dedicated subwoofer, though the 15W500 used for bass, as it coheres with the rest of the system, partially due to the larger size woofer and necessarily lower excursion, is more to my liking.

-Visaton W300 -- Very good. Utilitarian, no frills, robust and performative. A single or pair of these are a utilitarian choice for a mid field bass/midbass woofer in a two way or three way design. Close in performance and capability to the 15W500. I prefer the 15W500 as it seems a bit quicker though maybe not as extended. I prefer the tonality of the W250 to all though it is not suitable for midfield purposes as it cannot handle loud passages as well as a 12" or 15" by any means. Within a near field listening range the W250 remains my preference.

-Visaton W200 -- Very good performance though not excursive enough and thus limited in low frequency bandwidth. Cannot play loud below 50hz. Great choice for midfield desk monitors.

-Seas H1878-08 -- Best premium near field woofer I've tested. Most composed despite its smaller 7" size. Unsuitable for midfield purposes.

-Visaton AL200 -- Very good premium near field woofer. Digs a little deeper than the W250 though I prefer the sonic presentation of the W250. Despite being a smaller diameter, the Seas H1878-08 is a more robust performer. Reiterating, a pair of 8" drivers per channel functioning as subs is not suitable for a midfield speaker. A pair of 10"'s per channel is the minimum one should consider when building midfield speakers, though a single 12" or 15" per channel is a far better choice.

-Visaton WS20E -- Best inexpensive near field woofer I tested by far. More composed and robust than the W200, though at the expense of tonality and speed. Suitable as a midrange driver in a midfield application though not as performative as the B200 covering 100hz-1200hz. 

-Celestion TF1525 -- Very good. Simple though mediocre commercial build quality.

-Visaton WS25E -- Not performative enough to be deemed a midfield version of the WS20E. The WS20E seems more refined. 

-Seas L26RO4Y -- 4 layer voice coil version. This is a different beast compared with the other drivers tested in this section as it features a 30mm p-p xmax and a 56mm excursion limit. Excellent below 50hz. Horn load this in a vertical open baffle to avoid rear wall reflections. It features a relatively heavy cone and voice coil and thus is inefficient and also not suitable above 50hz. Ideal when implemented as an ancillary active subwoofer for achieving ultra low bass down to 20hz or lower with relatively minimal distortion. This woofer produces bass tones that the other drivers in this test cannot even approximate. Given this, if one is going all out, a pair of these would be the icing on the cake in a perhaps unconventional four way speaker build.

-Beyma 15MI100 -- Good

-Seas CA22RNY -- Okay

-Beyma 15G40 -- Meh. Excessive for home audio purposes. Likely ideal for louder output sound reinforcement.

-Kartesian Sub250 -- Meh. Sounds good under a certain volume then craps out.

-Visaton W200S -- Meh

-Visaton W250S -- Blah

-Seas A26RE4 -- Blah

-Audax PR240M0 -- Blah

All the smaller diameter woofers in this test, with the exception of the Seas L26RO4Y, are suitable as midrange drivers in midfield applications above 60hz-100hz depending on the driver and application. Lower mass often paper cone drivers like the Visaton W250 are better suited for this sort of application than aluminum or plastic. That said, the drivers in this test, primarily dedicated sub bass drivers, are not as performative in the midrange as certain full range drivers mentioned in the midrange test. If designing a compact two way midfield speaker without an ancillary subwoofer, a pair of Visaton W250s per channel, or possibly the W300 would be my driver of choice.

I prefer using a 60hz crossover rather than 100hz or somewhere in between as the former yields less room interaction and associated resonances. Also resultant of the low mass cones of the Zenith drivers or similar handling frequencies down to 60hz, bass is more defined than when higher crossover points are used.

Additional drivers tested:

Visaton Tweeter