Editor’s Note: Readers will find this articles to demonstrate that, if anything, Bill Dudleston is as competent a communicator as he is an engineer for a probing Doug Schroeder. I hope you will find this article as exciting to read as I did.
Doug Schroeder (Q): Briefly tell about the recovery from the Allen Organ ownership of Legacy.
Bill Dudleston (A): Initially my duties were quite spread. I was asked to review the design of their organ speakers, co-develop the architecture and applications for an 8 x 8 DSP product, and design a number of speaker systems for commercial PA. This was all quite interesting; especially the time spent in the recital halls, Basilicas, Steinway venues, etc. However as I stepped away from the day to day customer contact, we had several executives step in to distribute the line over a six year period. Each chose their own path, and the focus wandered a bit. The last year of product development before I bought the company back, I was asked to design a number of smaller products using 5” woofers. Not exactly where I wanted to take things personally.
During the eight years with Allen, I still had my facilities in Springfield for product development. I put in thousands of hours on driver optimization, diaphragm material studies, acoustic studies and DSP applications. In this respect, this was very good for Legacy.
Q: You have named the pinnacle of your line the Helix. Please share the inspiration for that name and the thought process behind it.
A: The name Helix references that which is “fundamental to life”, the genetic DNA ladder. No matter how small the deficiencies of a loudspeaker may be, they become quite detectable over time. What is new and exciting at first, can fade once the brain is onto these shortcomings. The goal of Helix was to reduce these telling detriments by an order of magnitude. This would allow us to relax our brains enough to trust the authenticity of the source which is truly “fundamental to life”.
Q: You are located in Springfield, Illinois yet half of your business is abroad, and India, for instance, recently shown interest in your products. How would you size up the global market currently for high-end speakers?
A: India has been supplying precision hard drives, electronics components and computer technical support for years, and they are catching onto high-end audio now. Globally, business is very good if the speaker system’s performance can differentiate itself from the crowd. The new audiophiles are a slightly different breed. Exotic alone does not cut it anymore.
Audiophiles the world over desire a demonstrable difference, otherwise it becomes a stagnant swap club of novelties. By focusing on performance and value we have always experienced increases in sales whenever the economy gets tight. People stay at home more and invest in what matters most to them.
Q: Touching on the aesthetics of the Helix, one of your cabinet makers, Joe Green, did an exquisite job on the pair of Helix I have been using. They truly have “to die for” luxury and beauty. Some would suggest this level of finish is uncalled for; what do you say to that?
A: I’ve always defined value in audio as performance/cost. Our premium finish is the icing on the cake. Our head cabinet maker is allowed an artist’s freedom in veneer selection and matching. It is a feat to behold as he scurries around the flitch with chalk in hand, selecting this section for the base, that section for the front. It is a lot like watching a tailor cut out a fine suit. He loves his job. Our staff is quite critical of materials chosen. Every individual of our staff has the ability to pull a cabinet off the line if it doesn’t represent their vision of a Legacy product…and they do.
Q: Please explain the major steps involved in the process of building a Helix cabinet, the thickness of MDF, isolation of the midrange and tweeter drivers, etc.
A: We CNC mill of each of the 34 MDF cabinet components:
The base is assembled from laminating 3 layers of 1.5” thick MDF with the edges trimmed in hardwood.
Front baffle overlay: 1.25” MDF with hardwood framing, this section primarily acts a lens opening for the front drivers.
The central face hosting the midrange and treble drivers begins as a 2” thick slab of MDF and then is machined to its oval shape. The left and right sides are milled to slope away from the raised central region which is pocketed to host the tweeter lens.
The tweeter lens is a specially machined device which adds acoustic impedance while keeping the dual diaphragm’s interference pattern constructive.
The main enclosure consists of various thicknesses of MDF, each cut to the proper dimension and angle, then interlocked and braced to divide the enclosure into 4 separate chambers. Sidewalls are two composite layers of ¾” thickness for an overall 1.5” wall thickness. The main enclosure is glued and clamped together, then coated with a rubberized asphalt-like material which resembles truck bedliner. The other components of the cabinet are sanded, stained, and sealed before attaching to the main enclosure.
Q: How old is the Helix design, and how many iterations of it were there? I know that the current generation has transitioned from the metallic coned subwoofer to Rohacell. Any other major changes?
A: We shipped the first unit in 2001, however the acoustic design work for the driver arrangement began in 1995 with a skeleton frame. Side panels could be added or removed. We computer modeled a number of driver arrays, and built three configurations for evaluation. In one design, there were five 5” mids in a circular array. We could make the pattern look good at any one frequency or listener position, but the lobe collapsed miserably with any movement. A design that fared better was a triangular array of midranges with a vertex directed towards the floor. When this design was compared with our existing Whisper system, a listener panel readily favored the Whisper’s “two above and two below” arrangement. Further modeling and experimentation led to the gently angled front baffle with splayed mids.
The 15” spun aluminum cone was recently replaced by the Rohacell backed, silver/graphite woven diaphragm, and the 15” midwoofers are the latest version of the Neodymium series from B&C (from Florence, Italy). The other drivers have yet to be improved upon.
Q: The Helix’s in my listening room are set on casters; do your installations typically utilize spiked footers?
A: Actually no, the weight of the cabinet is such that it has little difficulty anchoring itself and coupling to the floor. The cabinet is sufficiently decoupled from vibration just on casters that I personally feel the advantage of being able to tweak the speaker position and toe-in ‘outweighs’ the advantage of spiking anyway. We of course, can supply spikes if desired.
Q: There is a bass radiator located discreetly underneath the hulk of the cabinet, directly below, firing downward. I have rather thick carpeting, and the wooden “apron” at the bottom of the cabinet nestles against the carpeting. The radiator essentially seems snuffed by the carpeting. Would this largely nullify the influence of the bottom radiator?
A: If you had a wooden floor lacking carpet, I would actually recommend placing a ½ foam pad under the radiator. The radiator is as close to the floor as we can get it so that it squeezes the air in and out the rear slot. The wavelengths radiating there are about 40 feet long, so they will easily escape from the base.
Q: In a wonderful convergence, the 750W module used internally to power the 15” rear firing subwoofer of the Helix is IcePower, the same brand used by Jeff Rowland in his six-channel MC-606, which is also being used along with the Helix in a “Super-Review.” You spoke admiringly of Jeff; care to comment on the IcePower module and Jeff’s designs?
A: Jeff’s designs are always well executed. His willingness to use a digital design in a high-end amplifier demonstrates he is out in front of the pack. For Jeff to utilize another designer’s circuitry also speaks well for him. The ICEPower module is the most reliable design in amplification. It won’t self destruct if shorted, and doesn’t clip hard. It is by nature a balanced design and features balanced inputs as standard.
Q: You have given some thought to making the Helix an active speaker utilizing IcePower internally. Precisely how would this design look? Is the consideration of an active version of the Helix an accommodation to the marketplace? Would there be a sonic concession involved?
A: We have plenty of room to add additional modules internally. XLR jacks simply replace the binding posts for any channel. We have some larger studios, and even a University that have requested internal amplification. Recently, we made the decision to only add one more channel as an option: powering the mid-bass section internally. This leaves the end-user to supply a pair of stereo amps. Flexibility is preserved for the audiophile, and performance is not compromised.
Q: You have made a pair of Helix with passive crossovers, as well as a pair of Whisper using an outboard crossover. Are these options still available to the public?
A: Yes, they are available.
▼Helix with only sub section amplified. Jumpered rear panel of passive crossover version.
Q: What would you consider your three most crucial design decisions associated with the development of the Helix?
A: The decisions were based on the goals of the design:
totally realistic dynamic range with ultra low distortion
With designs like Focus and Whisper under our belt, we knew how much air displacement was required. We took advantage of the advances in magnetic materials and utilized neodymium in the motors to kick up the acceleration. We needed an efficiency of at least 100dB to remove possibility of driver compression or overload.
bass attack with a natural decay
This is achieved by avoiding the typical 5-millisecond bass lag behind the treble. This late start and unnatural decay is the primary reason we are so sensitive to resonance in this band region. DSP is critical for the impulse alignment and the required parametric correction. To maintain a controlled radiation lobe, the 15” mid-bass are mounted in an “over and under” arrangement, with the center of the array being coincident with the tweeter. Most unusually, the upper 15” is dipolar (open air) while the lower 15” is sealed. The summation of the figure-of-eight and the spherical omni patterns result in a directional cardioid pattern well down to the room pressurization frequencies, where the sub takes over.
vivid soundstage without smear
This is accomplished by virtually eliminating early reflections. The front baffle of HELIX provides a symmetrical layout. The high frequency driver is located at the center of four, splayed, midrange drivers. The 6” midrange drivers are spaced and angled to prevent sidewall and floor reflections, while summing maximally on axis and gradually attenuating as you move off axis.
Q: The top woofer is not hard wired, but instead uses spring clips and binding posts. What was the reasoning for this? Some might suggest this is a sign of compromise in the construction of the speaker. How would you respond?
A: I am a big believer in solid mechanical connections. The tension clips will not vibrate loose, nor do they require floating of the input wires in molten solder, which typically has 1/7 the conductivity of a tensioned contact. The only other thing comparable is a silver solder connection, which we use on the binding posts.
Q: Which companies are making your drivers?
A: The 15” subwoofer is our own creation, and is the 4th generation design. The diaphragm and spider are custom built for us by one specialist, while the motor is built by another. An adhesive supplier north of us oversees that the assembly is to specification. The driver features a second coil for dynamic braking.
The 15” midbass drivers are made for us by B&C of Florence Italy. These drivers exhibit an Xmax of 14mm, while maintaining a sensitivity of 99.5 dB due to a flux density of 1.25 Tesla. Unlike other neodymium motors, the structure can handle 1000 watts without demagnetizing.
The midrange drivers are custom built for Legacy by B&C. They are the result of us starting with one of their stock drivers, and then refining it for our application. B&C has been terrific about meeting and upholding our ultra-strict tolerances.
The tweeter is based on an Audax neodymium design. It is a dual silk diaphragm unit with ferro-fluid in the gap.
Q: The bass and mid drivers are made of what you called “natural pulp”.
The phrase reminds me of Hemp drivers. What are the advantages of this material in speaker cones?
A: Long fibers of random lengths provide great internal damping. Even the shaft diameter of the fibers vary individually. In pulp form, they are vacuum filtered and pressed to a diaphragm of uniform thickness. Imagine trying to create such a thing from manufactured fibers. You would need hundreds of dye sizes to extrude the fibers, and they wouldn’t nest together as well.
Q: The bass drivers look like the ones used in the Escalante Fremont. Escalante says that the larger dust cap acts like a midrange. Is that a reason you chose these drivers?
A: I would say the smoothness of the driver at higher frequencies is quite unusual for a large driver. It is indeed a good midrange.
Q: You seem very excited by the development of the Rohacell material for driver cones. Please explain what Rohacell is and its advantages.
A: Rohacell is a lightweight polymethacrylimide rigid foam. It is used at Fermi National Laboratory as a structural material, also as a core material in present aircraft construction. It is available in six different densities. Rohacell 71 is selected as it provides the highest strength/mass ratio.
Q: The Helix is the most high efficiency speaker I have used, despite its enormous size. The website lists the efficiency at over 100 dB. What is the actual measured efficiency?
A: 101.5 dB @ 1watt/1m
Q: During installation you commented on the extreme high extension of the mid drivers. You alluded to the fact that in some professional applications they can be used as tweeters! How does that top-end extension help in determining the mid/tweeter crossover point for the Helix?
A: Upward extension of the frequency response correlates with superior impulse response. More simply, it indicates the transient behavior of the driver. We want a high acceleration factor as it allows us to overlap the piston area in a beneficial way. If both the mid and the tweeter can operate at 4kHz, you want to combine the two in the way that voices most naturally. Since we can control the phase relationship between the two with DSP, we are allowed more degrees of freedom in the crossover with an extended driver response.
Q: During set up you worked to create gentle slopes for the crossover points so that the drivers “support each other.” How does this principle correlate to your choice of drivers and their use in arrays?
A: See above.
Q: It also looks like the tweeters use two types of wave guides, internal and external. Also, there is a perforated ABS screen, does this add diffraction?
A: Each tweeter is loaded into a short flare which increases the acoustic radiation impedance (lowering distortion by raising output in the forward plane). A vertical dividing blade further increases this radiation impedance. The diaphragms are splayed horizontally, then partitioned by this blade, which prevents destructive comb filtering off-axis at frequencies where wavelengths are less than 1” long. The cloth covered mesh passes pressure normal (orthoganol) to the diaphragm while slightly reducing diffraction at the baffle surface.
Q: In a rather unorthodox manner all the midrange drivers are pointed slightly away from the sweet spot. Is the idea to join their inner edges of their waves, similar to a “midrange phantom image” per channel?
A: A single piston will have too narrow of dispersion at high frequencies and too broad dispersion at low frequencies. A rule of thumb is that the driver will become beamy at wavelengths shorter than the width of the driver. We can use this to our advantage if we plan carefully:
1. Assume a pair of drivers in a line.
2. Introduce a splay angle such that they together provide the output of a single driver at higher frequencies on axis, and do not destructively interfere until 30 degrees off axis.
3. Now note that the width of the two adjacent drivers is effectively doubled at lower frequencies. This reduces the radiation angle desirably.
We have now widened the high frequency radiation pattern that was too beamy, and narrowed it where it was too broad. You might ask why we didn’t do the same with the 15” drivers? Well, we did (see the Double Helix Professional system in the Technology section of our website). However, we engineered an alternate solution for the Helix using the rear cancellation driver to prevent the need for a 30” plus baffle width.
▼The DOUBLE HELIX
Q: Not only do the front grills come off, but also two individual rear grills, one for the top bass driver and the other a specialty driver to absorb unwanted bass reflections. Do you recommend removing these rear grills?
A: I prefer that the rear grilles be left on the speaker, for prevention of dust accumulation in the magnetic gap.
Q: While programming the outboard crossover you inverted polarity on the drivers temporarily in the process of eliminating room modes. Please explain why?
A: I always like to see what the room is contributing overall. When I flip the polarity on one of the inputs while feeding mono pink noise to each side, I can look at the spectrum analyzer and see what the room has added. The speakers null out, leaving a plot of the room contribution.
Before shipping the Helix we always run a test with alternating drivers out-of -phase. The deeper the null at the crossover frequency, the greater the quality of summation is when in-phase.
Q: There is a hidden 15” rear driver underneath the angled grill you called the “Pac Man” driver. Its purpose is to, as you said, “…gobble reflected low frequency waves which do a round trip off the rear wall, then the front and toward the listener.” If this is not a passive radiator, but an active driver, what is the source for its signal? How is that signal treated for that driver?
A: The rear energy terminating driver is connected out of phase with the front drivers. Its transfer function is created by a passive band-pass filter and the small cubic volume of its angled enclosure. This allows us to reduce the energy buildup due to a rear wall in the 50 ~ 120 Hz range. Its operating level is set proportional to the front radiation. If you put your ear to it, it sounds like it is radiating garbage. It is actually reducing garbage that would otherwise reach the listener.
Q: Both the open baffle bass driver and the “Pac Man” bass radiator are located to the top of the backside of the Helix. Is it the purpose of the location of the radiator to correct anomalies caused by the upper bass driver?
A: Actually the composite output of both of the B&C 15” is being factored in.
Q: The grills of the Helix stand out quite a bit from the drivers. What is the purpose of the extra forward extension?
A: The farther the grill is away from the driver, the less it impedes the driver. In this case we want low impedance to the composite signal off the baffle.
Q: The Legacy binding posts are superb. You have spaced them wider than necessary to accommodate dual bananas. Am I correct in recalling them to be platinum covered copper?
A: We space them wider to provide more finger room. This also facilitates CE approval. The posts are made from a proprietary alloy and then plated with silver and nickel.
Q: In my setting the Helix is approximately one foot from the side walls. The speaker also sits just one foot below the ceiling. This is a very tight fit. Yet, you were of the opinion that the Helix would be equally at home in a moderately sized room. Please comment on that.
A: Helix has a carefully designed radiation pattern that avoids side wall and ceiling reflections. It also has enough piston area to pump the room pressure directly up and down at low frequencies. It does not ‘wait’ for the room to sympathetically load up, then wait for it to dissipate. Again, Helix bass starts and stops 5ms before conventional designs. The impulse response demonstrates this fact.
Q: There is a faint “Ahhhhhhh” sound emanating from the Helix drivers when there is no signal. Explain what I am hearing.
A: First, you are employing balanced inputs, which is +6dB over unbalanced on amplifiers with serious gain. Additionally, you have introduced a loudspeaker which will magnify the output intensity (and dynamic capability) by more than a factor of ten over that of a typical high-end speaker. The average high-end speaker has a sensitivity of 88 dB. Helix has a sensitivity more than 13 dB greater than that. This means you will hear what is upstream more readily. It does not mean you have compromised your noise floor relative to dynamic range. It is quite the opposite.
With a given amplifier, Helix can supply much greater dynamic range above the noise floor. Helix is also configured as a 4-ohm load, to take advantage of the additional current capability that most amplifiers can deliver below 8 ohms. If one merely listens at very soft levels, they can request we series-wire drivers at the factory instead of parallel wiring them, as the Qts of the system will not change. This will drop the apparent sensitivity by 6 dB. This has not been an issue with any customer thus far, as the noise level is similar to a phono stage when no LP is playing, but with the benefit of about 30 dB of additional performance headroom above the noise floor. Also keep in mind, Douglas, you have designed a wonderfully quiet listening room, quite isolated from the outside world, and 8 to 15 dB quieter than most.
Q: Many speakers incorporate either a subwoofer or passive radiator, but not both. How do these two drivers work in conjunction with each other?
A: The passive replaces what would be a very long tuning port… so long with adequate diameter it wouldn’t fit in the enclosure. Above 24 Hz, the woofer does the work, but as frequency drops the passive begins to contribute in-phase signal (i.e. they are both moving outward at the same time). At system resonance, the woofer virtually stands still and the passive strokes a distance up to twice that of the woofers longest throw. Frequencies near DC do not pass to the subwoofer, therefore the woofer does not unload.
Q: For all the drivers of prodigious sizes in this speaker, the cabinet is remarkably vibration-free. Do you have a general rule of thumb in regards to cabinet mass for your speakers? How do you judge the inertness of your cabinets? Currently it is popular to use high-cost, extremely high-density materials such as aluminum for speaker cabinets. The suggestion is that these materials will allow the drivers to have less coloration. What is your thought about that?
A: Some think that inert means stiff. It actually means nonreactive, or slow to move. A jug of sand is inert. Stiffness is a way to prevent low frequency transfer, but as the material stiffness increases, efficiency of high frequency transfer increases. When one thinks of damping, we think of absorbing energy. The key to enclosure damping is to keep vibration inaudible. Just as one optimizes each driver over its frequency range, we work to optimize the mounting frame, gasket, the individual acoustic chamber and the overall enclosure. Some parts are thick and heavy, some parts are light and stiff, and all parts are well reinforced.
Q: I see that you spec each driver to +/- 2dB. Why did you choose to list your specifications this way as opposed to the standard +/- 3dB?
A: We rely on a voiced target function that must fit +/- 2dB for the overall structure. Our driver assemblers must hit +/- 1.5 dB, before we match parts or tune the crossover. Our speakers are optimized for a realistic listener position, not an arbitrary 1-meter position.
Q: It is difficult to tell whether all the front-facing drivers are on the same plane. If they are not, is that due to horizontal time alignment?
A: They are approximately aligned. Within the crossover is where things are locked in precisely.
Q: You mentioned that the equalization curve of the Legacy Wavelaunch Processor is a composite of four rooms. Please expound on that.
A: The standard algorithm that ships with the unit has a fully established crossover, which is both time and phase-aligned. Timing is independently set for each driver cluster and the phase is aligned via the marriage of complementary slopes. The optimization is verified via subtraction. We also have influenced the algorithm by combining common characteristics in the final EQ settings based on multiple listening rooms. One is in Chicago, one is in Elko Nevada, one is our own listening room in Springfield, another is from a room in Pennsylvania. We did not average the curves, but merely included common EQ elements found in all the rooms.
Q: If an individual is technologically limited, describe how much work would be involved in them learning to use the Helix system.
A: I think the most critical element is simply plugging in the wire harness correctly and running the balanced interconnects to the matching channel. It is very straightforward. We can supply pre-labeled cables for those feeling sheepish and without dealer assistance.
It is best to listen for a bit before making any adjustment to the settings. Otherwise they will find themselves second guessing every other mastering engineer. I also suggest making the EQ adjustments to the input channels only, to preserve the transitional relationship between the drivers.
I would estimate that less than half of users involve their computer in the system after the installation. If one wants to trim a driver level, it’s pretty simple to select the channel from the front panel and make an individual adjustment. Then there are the more adventurous souls…
Only twice has someone asked us to fix a data file. In both instances, we had the solutions back to them by e-mail within 48 hours.
Q: How is production currently running? Is there a waiting list?
A: The waiting period is typically four to six weeks. We try to keep them in stock in each finish, but they seem to evaporate on us. Interestingly, we have sold more in the past year than in the previous three years.
Q: Let’s say someone with the means to obtain the Helix is so distant from a shop or Legacy’s factory that they won’t be able to hear them prior to purchase. How would you describe their attributes and sound?
A: Authentic. Very faithful to the human qualities imparted to the music. Weighty. Transparent to the degree it is fun to compare electronics. A very useful tool to evaluate the recordings that I make at our studio.
Q: For this review pair of Helix, David Salz of Wireworld supplied some of his new Mini Eclipse internal wiring. Which drivers was this cable used on, and how did you find its performance?
A: Douglas, we ran a continuous run of the new cable from the binding posts to the tweeters. My staff found the Wireworld Mini Eclipse a pleasure to work with, and the sound was at least comparable to the silver cable we have in house.
Q: Bill, I hope that you have found this interview exhilarating! Thank you very much for sharing your thoughts with Dagogo and its readers!
A: It was my pleasure Douglas. Thanks again for your thoroughness.
- (Page 1 of 1)