Tip #18: The Hammer Dynamics Super 12 Loudspeaker Kit
by Dick Olsher (5/29/00)
It gives me great pleasure to introduce the Hammer Dynamics Super 12 loudspeaker to a wide audience, and not just because the Super 12 represents a milestone in the evolution of the full-range dynamic driver. At its asking price ($575/pr in kit form without cabinets), it is also a veritable godsend to music lovers looking for great sound and high efficiency. In my global search for the best and most cost-effective audio gear, it seems that I almost missed what was practically parked in my own backyard.
Magdalena, New Mexico, home of Hammer Dynamics, is a mere 3.5-hour drive from my listening room. Magdalena, a small town even by New Mexico standards, earned its name around 1540 when a Spanish soldier saw the face of Mary Magdalene on the north side of the mountain that overlooks the present town. Another sort of vision has kept designer John Wyckoff occupied for the past several years: the pursuit of full-range driver technology. I discovered Hammer Dynamics on James Melhuish's delightful web site titled "Single Driver Speakers" (http://www.melhuish.org/audio/hammerdynamics.htm). I was sufficiently fascinated to arrange for a visit. What I saw and heard in Magdalena convinced me to purchase a pair of the Super 12 loudspeakers. In order to save time, I commissioned Wyckoff to build me a pair, complete with cabinets, to his specifications. Hammer Dynamics now also offers assembled loudspeakers at a cost of up to $1,600/pr – depending on desired finish and grade of crossover components,
Let me turn the podium over momentarily to Wyckoff for a few words of explanation: "The mission of Hammer Dynamics is to create great sounding, well balanced, high-sensitivity loudspeakers, to owners of low-powered tube amps, for reasonable prices. Five years ago I was struck by the fact that there were no inexpensive high-efficiency loudspeakers on the market. Once I settled on a "full-range" approach I began testing
most of the full-range drivers on the market. Not one of them gave me the sound I wanted no matter the price. The few, which had reasonable mids and highs, had no bass. The few, which had good bass, had lumpy mids and no highs. My solution to these problems was to design my own driver. My final design represents the best compromise I have been able to come to. The driver is capable of fast, accurate bass, silken mids and exceptional response to 10,000 Hz. I believe it is better to add a fast super-tweeter to a system than a slow sub-woofer. During my research I found that the usual 8-inch full-range drivers were never going to offer the kind of dynamic range I wanted so I finally settled on an ultra-light 12-inch. The entire moving mass of this driver is less than 29 grams. With a surface area of around three times that of an eight-inch driver the dynamic range is increased by a similar factor. The S-12 represents my efforts to create a loudspeaker that sounds like Quad ESLs with much greater dynamic range, sensitivity and another octave on the bass-end."
Thank you, John. My own research and recent listening experience with several full-range drivers has confirmed the validity and potential of this technology. To be blunt: full-range drivers are capable of coaxing more magic and more drama out of reproduced music than a horde of expensive multi-way speakers. My passion for these drivers has little to do with nostalgia; rather it is driven by the technology's ability to kindle the soul.
The Quest for that Old-time Religion
There was a time, two generations ago, when the full-range cone driver reigned supreme. In an age when the radio console together with the shellac 78-rpm record defined audio quality, a frequency range of 60 Hz to 9 kHz was about as wide a window as was needed or desired for the enjoyable reproduction of available source material. If anyone dared to open the window any wider – especially in the treble - they ran the risk of exposing gremlins such as needle scratch and other high-frequency hash and noise. With the advent of the high-fidelity phenomenon in the 50s, the audio industry moved inexorably toward multi-way loudspeakers, such as two and three-way designs, as a means of expanding the bandwidth at the frequency extremes. The advertising campaigns/hype at the time were so effective, that for many people Hi Fi became synonymous with bandwidth. Many consumers expected to pay a premium for a high-fidelity loudspeaker basically because of its increased bandwidth. Even today, many audiophiles and audio engineers seem convinced that the road to hi-fi heaven lies at the frequency extremes. Of course, music lovers know that this is patently false, and that the emotional content and drama of live music have little to do with the frequency extremes.
H. A. Hartley - a contemporary of Voigt and Lowther - devoted much of his life to perfecting the single-driver, full-range, loudspeaker. He lived through the transition from full-range to multi-way, and clearly realized that the widest frequency response was obtainable from a multi-driver design. Hartley noted that if mere width of frequency response was the ultimate goal then such a system would be ideal. Sure "audiophiles want it," he said, "but I have no real evidence that music lovers want it." His original bandwidth desideratum for music lovers, laid down in the 30s, was perfect reproduction between the limits of 32 and 9,000 Hz. With 30 years worth of hindsight, he expanded the limits from 20 to 12,000 Hz. The crucial point is that while there's nothing wrong per se with the pursuit of bass or treble extension to the threshold of audibility, the heart and soul of the music remain within the range of 60 Hz to 9 kHz. And that, my friends, is precisely the frequency domain of the full-range driver. Its magic comes from having a single coherent voice span the critical midrange and lower treble octaves.
The road to "audio hell" is paved with well-intentioned multi-way speakers. In its most benign implementation, as a two-way, the music is split apart between a woofer/midrange and a tweeter, typically in the octave from 2 to 4 kHz. The fundamentals are reproduced by the woofer/midrange, while most of the harmonics are assigned to the tweeter. This then is the sonic equivalent of that kitchen wonder - the Veg-O-Matic. You slice and dice harmonic textures with the hope that somehow the drivers can then properly synthesize the music's waveform. You don't have to be a rocket scientist to realize that when these drivers are mounted on a loudspeaker baffle, often with their acoustic centers 6-inches apart, that sonic integration is problematic. The differences in time delay between each driver and the listening seat typically result in severe interference effects in the frequency range overlapped by the drivers. In an effort to better integrate the transition region, many designers have pushed the tweeter's crossover frequency lower, sometimes precipitously close to its free-air resonance. This approach is based on the firm acoustic ground; in that driver spacing becomes a smaller fraction of the wavelength at a lower frequency. Hence, a lower crossover frequency minimizes interference effects and produces a smoother transition between the drivers over a larger sweet spot at the listening seat. While such a solution results in a smoother frequency response, often the sound quality is far worse because of increased distortion. That poor tweeter is made to work much harder than it really wants to. In the end, it is reduced to painful screaming. This problem is endemic to many modern designs, and it doesn't matter whether that 1-inch dome tweeter is made of fabric, plastic, paper, metal, Kevlar, or even cheddar cheese. As a musical peak hits the tweeter, its distortion spectrum rises sharply, and in addition, it compresses the music's dynamics. To make matters worse, all this happens in the upper midrange (2 – 4 kHz), where the ear is most sensitive. I don't care how old you are, or how much high-frequency hearing loss you may have suffered, our auditory system is optimized for acuity just in this range. By definition, distortion in this range is very audible and objectionable. I have literally cringed when auditioning $10K and even more expensive speakers while trying to reproduce something as simple as female voice.
The Super 12: An Inside Look
As you might have guessed by now, my definition of a full-range cone driver is a design that spans the frequency range from at least 60 Hz to 9 kHz. Rather then use a full-range driver as a stand-alone loudspeaker, the modern trend is to supplement its response at the frequency extremes with the aid of a subwoofer and/or super tweeter. The S-12 uses a modified Audax metal-dome super tweeter to augment the speaker's high frequency response. The tweeter is mounted on a bracket, which positions it over the central region of the wooer to approximately align the acoustic centers of the drivers (at least in the vertical plane).
The woofer is a proprietary twin-cone design, featuring a uniquely damped paper pulp composite. The cones contain a mixture of four different wood pulps for an optimized combination of broadband response, lightness, strength, and self-damping. In addition there is a lightweight polymer mixed with the cellulose to minimize moisture content and further dampen the cones. The edge damping on the free (whizzer) cone is precipitated chalk in a polymer base to reduce edge interaction.
The woofer is to be mounted in a bass-reflex vented 4.5 cubic foot enclosure. My cabinets, as built by Hammer Dynamics, are simple unbraced boxes, constructed of ¾-inch MDF, glued and screwed together. Of course, if you build your own, you have complete control over materials and construction techniques. The most unusual aspects of the enclosure are the materials used for internal damping. The inner walls are lined with bubble wrap. In addition, a square piece of felt is mounted on the rear wall behind the woofer. Finally, the cabinet volume, up to the lower lip of the vent, is filled with Styrofoam packing peanuts. Unorthodox to be sure, but Wyckoff assures me that these materials work extremely well in controlling internal standing waves, while being very cost effective. If you overfill the cabinet with peanuts, you're in for quite a treat. It's a cheap thrill, but highly amusing to watch bass transients grasp hold of, and then blow peanuts right through the vent.
I have to confess that I did have a role in tweaking the crossover network. The final schematic shown in Figure 1 below represents the results of a collaborative effort between John Wyckoff and I. My first serious listening impressions and initial measurements showed that the tweeter was not properly integrated with the woofer, being significantly shelved down in level relative to the woofer's upper range. Since I owned the review pair, I felt comfortable in tweaking the network to my liking. I shared my findings and suggestions regarding the tweeter's high-pass and the woofer's low-pass networks with Hammer Dynamics, and after much listening and several iterations, a consensus emerged that pleased everyone.
Two features of the S-12's crossover schematic deserve a fuller explanation. The major problem facing all full-range drivers, and in particular paper cones, is cone breakup. The extreme accelerations experienced by a typical paper cone above about 2 kHz, cause it to flex in a complex pattern. The cone no longer acts as a piston but rather as a collection of vibrating elements. The resultant frequency response is cratered like the surface of the moon with giant dips and peaks, commonly referred to as breakup modes. The magnitude of these modes can be controlled to some extent by using a cone material with high internal damping. Full-range drivers with thin paper cones (e.g., Lowther designs) are particularly susceptible to uncontrolled breakup modes. However, damping alone is often insufficient to fully smooth out a driver's response, and the designer must resort to electrical equalization techniques to get the job done.
The S-12 uses a parallel resonant circuit in the woofer network to control the woofer's breakup modes around 3 kHz. John assured me that I wouldn't be happy without it. And he was absolutely right. According to Wyckoff, who has spent an enormous amount of time tweaking this circuit to perfection, both the cone resonance in the coincident region, and free (whizzer) cone resonance all lie within a narrow frequency band so that a single contour filter can be used. The coincident region is defined here as the band where both the main and free cones are active, and is usually the stage of the worst spikes in full-range paper drivers.
In addition, the woofer is gently rolled off in the treble with the help of a first-order low-pass network (i.e., a 0.24 mH coil), so as to tame an 8 kHz breakup mode. Note that the drivers are connected in phase. If money is no object, I suggest that you go the extra mile and bypass both caps in the tweeter network with a 0.01 mF Teflon film and foil type capacitor.
The crossover board, on my pair, is mounted on the back baffle, except that the woofer's "Zobel" network represented by the 12 mF capacitor in series with a 7-ohm resistor was soldered directly across the woofer terminals. No input connectors are used, as the speaker cable is soldered directly onto the board. Take my word for it: no input connector is sonically superior to the most expensive connector money can buy. In this case, less is clearly better. I ordered my pair with the Hammer Dynamics' Litz speaker cable, which is also an exceptional value. The amplifier end of the cable was tinned and connected bare (no spade lug or banana plug) to the amp output terminals.
The S-12 represents an easy amplifier load, with an impedance minimum of about 8 ohms, as can be gleaned form Figure 2. The twin impedance peaks below 100 Hz are typical of a bass reflex tuning. The box tuning appears to be at a frequency of 45 Hz.
This is a high-efficiency design that is suitable for use with low-power tube amplification. A 10-wpc single-ended triode amplifier provides plenty of headroom even in a medium-sized room.
It is imperative that full-range paper drivers be broken in for a considerable length of time prior to any critical listening. A minimum time period of 50 hours should give the cone and suspension an adequate work out. Be sure to also toe-in the speakers directly at the listening seat, since the treble sweet spot (ala the QUAD-57) is focused within a fairly narrow beam.
Let me start with tonal balance considerations. The in-room frequency response is shown in Figure 3. The lower midrange (centered around 500 Hz) is clearly emphasized relative to the rest of the range. This serves to add weight to an orchestra's foundation as well as confer a touch more chest to male voice. This is no lean, low-calorie, midrange. If your notion of tonal beauty runs toward that of a 100-pound skin-and-bones model, stop, and do not pass Go! This clearly isn't your cup of tea. Actually, I happen to enjoy a midrange on steroids, with a bit of added muscle to liven up the harmonic palette. True, it's a bit bigger than life, but more often than not, I find such a balance to be consonant with musical expressionism. In an imperfect world, oh Lord, your humble servant beseeches thee: please send more speakers like this my way.
The upper midrange is quite smooth and well controlled. The lower and upper treble sound surprisingly quick and detailed. The tweeter adds much needed sparkle, air, and finesse to what would otherwise be yet another closed-in, dull sounding, paper full-range.
The bumps in the frequency response above 7 kHz are due to manner in which the tweeter is mounted on the front baffle. Treble energy diffracts around the narrow tweeter baffle and reflects off the woofer cone, which looks and acts like a parabolic reflector. The reflections give rise to an interference pattern, much like that of a comb filter. The impact on tonal color is slight but noticeable.
The response roughness below 200 Hz merely shows the effect of room modes and should not be taken as intrinsic of the S-12's performance. The deep bass is extended to about 40 Hz, while the midbass sounds tight and detailed. This represents a quantum leap in performance over my anemic Lowther PM2A rear-loaded horn. However, the Super 12 is not a long-throw woofer. Its limited excursion means that its maximum SPL is fairly limited. This isn't much of an issue for domestic listening, but don't plan to use it as a PA/dance party speaker.
In my experience, the greatness of a speaker has rarely been reflected in its nominal frequency response. It seems that every speaker that has won entry into my personal pantheon, fails the flat-is-perfect frequency response criterion. My first audio love, the speaker that redefined reproduced music for me, was the original QUAD electrostatic (QUAD-57) loudspeaker. Its major tonal trait was a significant midrange recession. It sounded laid back, but its astounding level of midrange clarity and harmonic cohesiveness was so captivating that it nudged me closer to the soul of the music than ever before. Having said all that let me simply state that the Super 12 merits a direct comparison with the QUAD-57. The main reasons that the QUAD-57 failed to obsolete the competition were its limited macrodynamics and bass extension. The Super 12 manages much of the cohesiveness and clarity of the QUAD-57 while bettering it in terms of dynamics and bass punch.
With the S-12, the harmonic envelope rockets into space, revealing pitch and volume modulations with the greatest of ease. This is the one aspect of full-range drivers that puts multi-way speakers to shame - even those very expensive audiophile darlings. It would not be unusual even for a $5K high-end speaker to use a $30 midrange, crossed over at 3 kHz to a $20 1-inch plastic dome. For the moment, let's ignore this speaker's fancy woofers and cabinets, and focus on reproduction of female voice. When a soprano lets loose, and transients build up from soft to loud, the distortion spectrum of the multi-way changes rapidly as that $20 dome tweeter is made to work harder and harder. The overall sound becomes rougher, edgier, dissonant (pick your poison) at the upper end of the dynamic range, as a swelling of distortion products pollutes timbre. In contrast, the S-12's dynamic palette is Teflon coated. Transients' ebb and flow across the lower mids, upper mids, presence region, and lower treble originate from a single voice with uniform personality. The feeling of effortlessness and purity in elucidating the music's tapestry is precisely why I so thoroughly enjoy this speaker.
Another strong suit is the delineation of spatial detail, both in terms of focus and stage depth. With the help of a good front end, image outlines may assume a 3-D palpability. I recently installed a Kuzma Stabi Reference turntable in listening room #2. This is an absolutely wonderful table that has had me addicted to nothing but analog for the past few weeks. The difference between analog and even an excellent CD player as far as the rendition of the original acoustic and the perspective of instrumental outlines is enormous. It made me realize yet again how much of a convenience medium the standard CD actually is. It was remarkable to observe the performance of the Super 12 in the context of a high-end front end, which included the Graham 1.5T tonearm, Symphonic Line RG-8 Gold moving coil cartridge, Audio Note AN-S3 step-up transformer, and the Air Tight ATE-2 phono preamp. No problem at all. As the front end improved, so did the Super 12's ability to recreate an illusion of flesh and blood musicians performing in a real acoustic space.
The Super 12 loudspeaker joins a select group of speakers that I could enjoy until the end of time. I would gladly choose it over a host of $5K or even $10K multi-way speakers. No, it doesn't have the killer bass or ultimate treble finesse of these other speakers, and it certainly isn't a status symbol. But it is intensely musical and absolutely affordable – a winning combination in my book.
Magdalena, NM 87825
Tel.: (505) 854-2370
Warrantee: Lifetime warrantee; 30-day trial with full refund except shipping
Price: $575/pr in kit form without cabinets
Frequency Range: 45 Hz – 20 kHz
Nominal Impedance: 8 ohms
Efficiency: 97 dB 1W/1m
Power handling: 125 W rms