Building a new speaker Part 1

March 28, 2019
 by Paul McGowan

Starting today, and going through the next three days, I’ll be detailing the making of the new AN3 speaker. If you’re not interested in learning about the development process I’d skip the next few days of posts.

One of the reasons I wanted to spend time with this subject is its fleeting nature. Once the process is completed: the pain, sweat, angst, triumph, and failures along the way are lost—and to my way of thinking they are all part of a product’s DNA as much as a young person’s upbringing shapes their lives.

As you read this, know that the AN3 prototypes are now bouncing along I-80 in the back of some truck on their way to Chicago for the Axpona show.

The battlefield where engineers Darren Myers, Bob Stadther, and I (ably assisted by Jordan Kamper) devoted our days nights and weekends in service of making great music has been relegated to the inner canister of a Dyson vacuum cleaner.

A look inside Music Room One would offer no clue of what occurred there: the building of a new kind of loudspeaker—a 4-way hybrid of ribbons and cones and analog and amplifiers and servos and DSP capturing the magic of Arnie Nudell’s work and the bounty of music properly reproduced.

I have filmed four videos to go along with each of the four parts of this saga. The first can be viewed here.

In this opening video, I spend the entire 11 minutes explaining Arnie Nudell’s reference system: what it is, why it works, what he tried to achieve and how. If you’re familiar with that system, the one following his triumph of the IRSV and all that came afterward from Genesis Technologies while he was at the helm and me at his side, then wait until tomorrow to watch Part 2.

Subscribe to Paul's Posts

19 comments on “Building a new speaker Part 1”

  1. Paul until now I understood the AN 2 will replicate the Nudell IRS killer (as line source) and the AN 3 and 1 will be a smaller and bigger unit.
    By your video I understood that the AN 2 will already be bigger than the IRS killer which so far looked like a very compact nice line source.

    Is there no equivalent (also in size) planned anymore for the IRS killer or is it still the AN 2?

    1. AN3 is the middle speaker in a line of speakers we will produce.

      Yes, it is the baby in the PerfectWave speaker line followed by taller pairs in the AN2 and AN1.

      We will also move downward in size and pricing too, maintaining many of the same characteristics.

      For example, the next one below AN3 will look very similar but it will be thinner and sport multiple servo driven side firing 10” woofers.

      The master plan call for a total of 6 models of loudspeakers (3 in the AN Series, and 3 in the next series down) all with powered servo bass, AMT tweeters, separate midrange, and performance mirroring the AN3 as closely as possible.

      Think of the AN3 as the reference standard, the pivot point in a line of speakers we hope will cover the needs of anyone in any living situation.

      Our goal is to get to $3K the pair if we can manage.

  2. In contrast to the most convincing design concepts of the PPP, P10, DMP and DS DAC I never understood the technical design concept of the IRSV. And having watched the video I even do not understand why you have stacked the boxes of the electronic gear without using a decent rack or your vibration damping platforms. Concerning the speaker design I only learned that the design goal was to get the Nudell sound reproduced. I hope there will be more and convincing concept ideas shown in the next videos primarily concerning the relevant data for minimum listening distance, space required behind the speakers, minimum distance to the side walls and the listening height where the sound waves from the different drivers meet in perfect phase coherency. The latter I never saw for the IRSV and maybe that’s why Arnie Nudell always wanted to design a pair of speakers without those subwoofer towers which rarely will deliver a phase coherent sound wave at the listeners ear.

    1. Most of us that follow Paul’s audio engineering crusades in search of audio perfection in his daily posts and videos realize they are tailored for different audiences.

      The daily posts afford the time to delve into engineering details.

      The videos must deal with the ever decreasing attention span of the ever increasing population by limiting content depth.

      I personally enjoy both format for what they are.

      Paul, if I could recommend a possible third format to document the AN3 development. Have you considered writing a technical white paper?

      Wish I could make it to Chicago to hear the AN3s. Enjoy some good deep dish with mushrooms.

        1. Paul,

          Having written many technical papers, I have just one word for you, “Coauthors”

          Having coauthors not only shares the work, it is also an opportunity to get young engineers to hone their writing skills.

          Since engineers are so well known for their superb communication skills.

    2. I’m not big on stacking components either. I worry about dents, scratches or components being too close causing some kind of interference in the sound. Also never stack on top of an amplifier which would limit the amps ventilation. I’m sure Paul has taken precautions to prevent those problems. Vibration free platforms are also a plus but recently I heard Paul say he’s not into tweaking the way he once was. As far as damping I feel the best damping is done with real heavy speaker stands which couple small speakers to the floor which improve bass and limit smearing of the sound stage on of the advantages heavy floor standing speakers have over bookshelf speakers.

      If I had the IRS V I would place the subwoofer towers right next to the main speaker array on the inner sides of each main speaker to prevent any coherency problems. I’m sure some might feel that cause some kind of diffraction to the mids and highs. I don’t think it will as long as the room is wide enough. I prefer speakers being places on the wide side of the room if possible.

  3. So in those speakers there is a crossover in the lower midrange at about 350Hz between the planar panel and the mid-bass driver? Where is the crossover with the bass driver?
    There was a comment early on about system dependency. I assume the reference to a complete chain was to Linn Exakt, where the source, amplification and active speakers are an integrated system. At the other end you have Harbeth, who are completely system-agnostic and can used with anything of sufficient power, and in the middle you have someone like PMC who recommend Bryston (used in all their active speakers for the last 30 years and PMC are Bryston’s European distributor), but other amplifiers are fine. Were those prototypes system dependent or agnostic? How were they meant to be driven?

    1. … As per PaulSquirrel, I would like to know about the speaker design and concept. I checked the other day, Infinity was not distributed in the UK and Genesis only briefly (some smaller speakers for a few years), both before my time. USA designed planar speakers are popular over here, the 2-way ones with added bass, and we have electrostatic panels as well. However, so far as I am aware, this 4-way configuration is completely unheard of over here and it wold be interesting to understand the rationale.

      1. Good questions, Steven. I am certain you understand the idea behind a 3-way so let’s just focus on the 4th way which many are not familiar with because it’s a new concept first introduced by Arnie.

        The 4th is the midbass coupler. In our design, this covers the range of 100Hz to 350Hz. In Arnie’s original work it was 100Hz to 500Hz. This region is critical to music, especially classical because it’s where the majority of energy is located. If you look at a curve of amplitude for an orchestra you’ll see a huge bump in this very region.

        Adding a fourth driver to cover this area has a number of advantages. And making it powered and variable by the user has even more.

        The obvious advantage we first see is separating this area from the midrange. The kind of midrange we need to get reference standard performance out of the speaker is a ribbon. Unfortunately, ribbons struggle to get really loud at low frequencies unless they are huge and even then… So, by adding the extra midbass coupler we remove the requirement for thse all critical frequencies from the ribbon madrange.

        Secondly is set up and personal preference. We cannot know how far apart the speaker pair will be placed within a home. What we do know is that however we voiced the speaker it was done so for a specific distance between left and right speakers. Once the user changes that distance to match his room our personal taste requirements the midbass is sure to be incorrect. This happens because of the coupling between left and right enclosures at those all critical midbass frequencies: too close and they are exagerated, too far apart and they lose body and become thin.

        By making the midbass coupler adjustable users can place their speakers where they work best then dial in the midbass to where it’s perfect.

        Lastly, amplifying the midbass internally gives us the opportunity to do it right so the extraordinary dynamic peaks of an orchestra can be attained without tazing the connected power amp. Thus, in the AN3, even a small tube power amp should be just fine to power these beauties to realistic levels.

        1. Thanks, Paul, now I understand.
          I thought it was a planar rather than a ribbon midrange driver.
          I suspect it is push to get a 4-way system to a retail price around $12k. There is a vaguely comparable product in Piega, without an active coupling driver, that keeps the price down by combining the lower midrange and bass in a single passive driver, and the midrange and treble is produced by a single coaxial driver combining a treble ribbon tweeter surrounded by a midrange ribbon driver. So it may lack the punch and ability to dial in, although they do an IRSV 4-tower aluminium thing if you have $200k spare change.
          So maybe the Stellar version will forego the coupler, as the ribbon midrange should really be what makes these things sing, and adopt a sensible mid-bass driver down to about 40dB.

          1. You’re very perceptive. Indeed, ours is a planar ribbon driver and capable of going lower than the big ribbon on Arnie’s.

            The Stellar series will have the same AMT tweeter we use on the AN3 (which is a custom built one that is spectacular) and a 5.5″ cone driver that we specially built to cover the range down to 100Hz.

            We cannot afford to add a 4th driver in the Stellar series so they will all be 3-way with powered servo subs.

            1. So Stellar will be more conventional, but with ribbon tweeter and active bass. Ribbon tweeters have had very good results in recent products from ProAc and Quad. I was hugely impressed when I heard the ProAc K series a few years ago.
              I’d forgotten you’re a bass junkie and a lower mid/bass driver alone would not be sufficient for PSA rockers!

        2. NHT’s 2.9 and 3.3 also have a midbass coupler they call lower midrange in that exact frequency region. I have the 2.9. They are great sounding speakers. Here’s the crossover specs on the 3.3. Same in the 2.9 with 10″ subwoofer. Ken Kantor is a great speaker designer. Great speakers for the money or regardless of cost. He designed some very nice bookshelf speakers too like super ones and zero.

          12″ polymer cone in 70-liter acoustic suspension enclosure
          Ultra-long throw 2-layer voice coil
          59 oz. magnet in die-cast frame
          19Hz free-air resonance
          Fc = 35Hz, Q[C = 0.65
          12dB/octave low-pass at 100Hz

          Lower Midrange:
          6.5″ polymer cone in 10-liter anti-modal enclosure
          Long-throw 2-layer voice coil
          42hfz free-air resonance
          Fc = 67Hz, OK; = 0.66
          12dB/octave high-pass at 100Hz; 12dB/octave low-pass at 320Hz

          Upper Midrange:
          4″ polymer cone in 3-liter anti-modal enclosure
          Fiber-composite molded frame
          110Hz free-air resonance
          12dB/octave high-pass at 320Hz; 18dB/octave low-pass at 3.5KHz
          Impedance compensation

          1″ fluid-cooled aluminum dome 1.4KHz resonance 18dB/octave high-pass at 3.5KHz Impedance compensation

          1. I’m very interested in PS Audio’s new speakers. I think they are going to be winners and priced right too. Love the finish on the cabinets. It’s always a plus when speakers look like a fine piece of furniture.

  4. Congratulations in finishing a product and about to present it to the world in a HiFi show. I am looking forward to hearing some of the music from it on YouTube. If it is as good as the Infinity, oh boy, it will surely be a huge bang to the world. I hope it will.

  5. Here are my observations. To duplicate the sound of acoustic instruments which was Arnie Nudell’s goal which areas of the audible spectrum are most important to get accurately? In my experience all of them. Which areas in the non audible spectrum are important? All of them because they can screw up what you can hear by creating distortions in the audible spectrum and eat up power uselessly. They can also damage your equipment. Which are the hardest to produce accurately? All of them. Which ones do recording engineers get differently from one label to another, one engineer to another, sometimes even one track to another? All of them. Which ones are affected in the way you hear them by room acoustics? All of them. For example in a concert hall the first sound you hear from an instrument has one spectral balance. As the reflections die out the spectral balance changes as different reflections are the result of reflective surfaces absorbing sound to a different degree at different frequencies and even the humidity in the air can affect the change and rate of change too. If a sound dies out in 2.5 seconds it has traveled about half a mile before its last reflections have died out to inaudibility. Which of these variables have high end audio systems understood, addressed, and compensated for successfully given the less is more philosophy and the tendency to try to isolate the sound system from the listening room and fix the system to perform one way only. None of them. With that attitude how could there not be fights all the time? No matter which design you pick using this approach it won’t ever meet the criteria Arnie set out. It it a hard problem to solve? It’s hard enough just to understand so of course it’s hard to solve. Arguing and going brickbats until the one who shouts the loudest or controls the money wins is a lot easier. It doesn’t get you the answer any of them is looking for but it is much easier. All you need to win is endurance and the largest club. The people who battle it out are extremely passionate, enthusiastic, get very excited, and I’m surprised they haven’t come to blows or pistols at 20 paces at dawn. Sorry, just my observation. I have a new definition of audiophile. It’s someone who will fight to the death with anyone who argues that his ideal sound system ins’t better than someone else’s.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram