How we design loudspeakers Part 4

March 30, 2019
 by Paul McGowan

39 comments on “How we design loudspeakers Part 4”

  1. Sad to hear your AMT project didn’t work out. So is that one officially abandoned now?
    As for the BG mid you’re using now, are they actually a valid choice AND still in production?
    Or do you have another ribbon/AMT up your sleeves for the AN-series??

    1. Bohlender-Grabener is indeed no longer in business, but the drivers are still being made (although I can’t recall by whom) and are available at Look under the speaker components tab, and then in the “hi-fi” sub-category. There’s a section for planar and ribbon transducers. The Neo-10’s can be found there.

      I asked Parts Express if these were NOS but they confirmed that they are still being manufactured.

        1. Good afternoon Paul,

          Please can you share the make and model of the “Vertical CD” rack in the video of Part 4, How we design loudspeakers.
          Thank you for your awesome, educational and human (funny) video series.

          Thank you

            1. Thank you so much Paul for taking time out of your busy schedule (Axpona to come, AN* (* = 4…9)) to provide an answer. I called PS Audio and TJ was gracious enough to take my call. Your organization is incredible and accomodating. What you project on your videos (appoachbale) permeates through your organization. Answering a call by your front desk is an example. Thank you again.

    2. Hi Paul – hi together,

      your “miniseries” how PS Audio does design and build the AN3 speaker is very informative and gives a good impression how a company of devoted and professional engineers, commited to best musical reproduction designs a transducer along a certain sound philosophy.

      To the actual product, the AN3 prototype, I have some questions in addition to a few remarks. You mention the difficulties and challenges to integrate your PS Audio AMT transducer. Have you ever considered to outsource these chassis and make use of the experience of producers like e.g. Mundorf, who equipped the Royal Albert Hall in London with his AMTs ?

      The description how to finalize the crossover of the speaker raised the question whether you will equip the AN3 with a different crossover targeted at the USA, resp. European/Asian market according to their different listening environments (big rooms, wood and plaster constructions vs. smaller rooms buildt with brick, stone, or concrete)? Such regional different musical receptions once were subject of an Ask Paul video.

      A design related remark: Since you choose a black baffle and most likely will not fit grilles on the AN3 to avoid negative influences on the sound, you might think of to have all visible metall parts and the cone of the mid tone chassis at the front baffle also produced in black.

      Looking forward to hear the AN3 soon also in Gernmany (High End 2019 in Munich?) I remain with very best regards, Klaus

  2. You gave us only a very small bit to listen to. But to me, the singing sounds like singing inside a small enclosed cubicle. I am not sure I am correct because the listening is so brief but at least I try to be honest to be helpful. Arnie’s speakers are not that good either as the hand over mouth effect is still there even with the twitters turned on. I know it takes lot of time to get a speaker to sound right. Do continue your efforts to produce one that sounds best like the real thing.

    1. Don’t think listening through a (badly recorded, sorry..) youtube video can tell us much about the speakers capability. With that said, I actually thought the IRS Killers sounded better than the AN3 through the short clip we were presented with.
      Only way to judge them properly is to be in that room. much is a plane ticket to Colorado..;-)

  3. To me it looks like Paul forgot to switch off the amp of the mid-bass-coupler of the left speaker. From 2:45 to 3:00 he’s switching off the amp for the subwoofer (by rc) and the amp of the right mid-bass-coupler (by switch). When he turns to the left amp he’s not switching the amp off but continues to explain about the construction.
    Could it be that the left mid-bass-coupler was still engaged when the demo track played? I can’t judge due to the poor quality of my (old) iPad. But at least I have the impression that the sound played through Arnies speakers has more “body”.

    1. Good Ear AW!

      Yes, observed and heard the same scenario! The IRSK’s have the 100-400hz playing (left speaker), whereas the AN3’s truly have just the mid-high end playing.

      I have an all-in-one monitor with “descent” built-in stereo sound. Presents a very obvious difference between them (due to the mid-bass still in the mix). However, believe the AN3’s top end sounds more natural and open, just not as big a sound as the IRSK’s…but then again, would need to hear a repeat demo, but on equal playing fields!

      Can’t say anything to Paul, as I do the same all-the-time!! 😉

      1. That could certainly have been the case. I do get scattered and forget. I have done the demo so many times properly and it is obvious to me how much better AN3 sounds to Arnie’s reference that I could have simply been focusing on the increased clarity in the midrange of AN3 that I ignored the midbass coupler left on.


        1. Suggestion…. don’t try to get more than 3 octaves out of a driver. Adjust the crossover points of the woofer and lower midrange upward. Try a 5 way design if that doesn’t work. 3 octaves is usually the limit of the usable range of dynamic drivers. They are after all inherently resonant devices.

          From what I’ve read about IRS and its successors at Genesis, the region between the bottom of the EMIMs and the top of the woofers was a problem. Have you ever considered adding a mid bass coupler as you call it to the IRS? Acoustic Research grappled with this problem for nearly 30 years and couldn’t solve it. IMO when they went to a 4 way system it was throwing in the towel on the best 3 way they could design. Turning Bose 901 into a credible full range 2 way system required jumping through hoops creating an equalization curve that required 1000 times as much electrical power to play at 30hz at the same loudness as at 1 khz. No amount of power could get it to reproduce the top octave.

  4. A lot of ideas that were held true for Paul seems to have fallen by the wayside.

    Less is more fell to adding a preamplifier making more more.
    Never use tubes for the best sound fell to the Sonic Frontiers amplifier and then the BHK amplifier and preamplifier
    Class D amplifiers will never sound great fell to Stellar
    The superiority of a line array fell to the AN3
    The perfection of the IRS V fell to the IRS killer which fell to the AN3
    Flat frequency response sounds bad fell to flat frequency response can sound excellent
    A 3 way concept pushed to its most absurd limit fell to a 4 way system not nearly as ambitious or expensive

    The ground seems to be shifting. This isn’t California where you can have an earthquake every 5 minutes. Boulder is on solid ground. they don’t call it the Rockies for nothing. This is NOT a moving target.

    Sorry to tease you like this Paul. It’s in my basic nature to tease my friends.

    So here’s something to think about. What caused you to have your prior beliefs and where were the errors in thinking that experience demonstrated were falsities? In short, why were your prior theories wrong? How do you reconcile your new thinking based on experience you didn’t expect? I don’t expect you to answer now or even ever, just to give you something to think about in your off moments. One thing is clear though, there is a gap in understanding that isn’t confined to you but is pervasive throughout this industry and to challenge the prevailing dogma is to invite a torrent of uninformed criticism. At least here there isn’t so much of that. There is more black art here than hard science. Keeping it that way has proven very profitable just the way false theories about winemaking like terroir were used to keep the myth that only Bordeaux and Burgandy could produce the greatest wines in the world becuase of their unique combinations of soil, climate, and whatever else they throw into the mystique. The French were outraged several decades ago when in a blind tasting by their own best experts the best California wines both red and white beat the best French wines and some wines made California are dead ringers for top Bordeaux.

    1. SM,

      Your “7-bullet points” All represent an individual & company who thrive on product creations through ongoing learning processes created through failures, trial & errors and ultimately, Successes! I’ve only been following PSA for the past 6 weeks or so, but realize and appreciate their open philosophy and willingness to share the journeys of engineering, development and talent of potential audio products…A Great learning experience for us Newbie’s (at 65, I still have much to learn in life)!!!

      Do I seek or like change…not usually. However, I am open minded to listen, contemplate and learn what really matters, be it music reproduction or life in general. IMHO, PSA is a refreshing and talented business endeavor to improving my in-home musical enjoyment!! 🙂


      1. I am glad you enjoy this site and I enjoy it too or I wouldn’t post here. I congratulate Paul for many reasons including building and operating a successful business. I wish I had a good head for business but I don’t. Since you are new here I should tell you I’m not an audiophile, at least not in the sense you understand it. Instead I’m an engineer and this is only a hobby for me. I analyzed, understood, and solved this problem 45 years ago. My ideas are very different and there are many people who post here who frankly are tired of me repeating them.

        My criticisms are not of Paul specifically but this industry in general. Some of the people in it are excellent electronics and even mechanical engineers. But when it comes to sound and acoustics, IMO none of them seem to know beans. I view them in this regard as no more than tinkerers. Some are quite clever tinkerers and come up with interesting ideas but still in the field of sound and acoustics which is what the ostensible goal of this industry is about I’m sorry to say they are roaming around blindly in the dark for the most part. My bullet points to Paul are just a nudge to provoke him to think about not just what he learned from tinkering but why his prior beliefs led him to the wrong conclusions. It’s certainly nothing personal. I know he and his team try very hard and are at least as good as any of their competitors. Were I in the market for the kinds of products he sells I wouldn’t hesitate for a moment to buy them from him.

        1. Good to hear from you, SM…believe we are on the same page. I too have analyzed, understood, and solved the music, sound and problem 45 years ago…as a student of keyboard, instrumental and vocal learning, as a teacher and performer of instrumental music and as an avid listener to many live non-amplified acoustical performances over these past years! My solution was simple…get out and hear the real-thing…LIVE!!

          However, quickly realized that the audio industry presented possibilities for musical enjoyment in the convenience of my own little venue called home, thus have been a want-a-be Audiophile since (financial limitations withstanding). Personal progressive stages of development have been ongoing these past 45 years, with my goal being “home reproduction of music as a live and real venue” that I intimately know, understand and appreciate from my experiences!

          To me, Paul and PSA represents the same “Live” reproduction goals that I have been able to finally achieve in my listening room, at least to my excellent hearing and understanding. And as per his on-going endeavors and achievements, the long up-hill \ down-hill journey is all part of the process, just as it has been for me and many other lovers of music.

  5. Paul, the demoes are very interesting, and I am looking forward to hearing the final product, although I think the market is already well served with speakers selling for over $10,000 the pair. The essential problem with all of them is the need for space around the speakers which then, by necessity, protrude well into the room, raising important issues around domestic tranquility. I thought one of the design goals with your speakers was great sound at closer rear wall boundaries, using dsp and so forth. If so, why are you mirroring the placement of the reference? How does that get you closer to great sound nearer the rear wall?

    1. I wish I knew of $10,000 speakers I could recommend to people but, alas, I do not. There are so few that live up to my standards and tastes that it’s one of the biggest motivators for starting a speaker line. I just am at a loss as to what to recommend to people when they ask.

      Yes, indeed, one of the goals with the AN series (and our follow-up Stellar series at lower cost) will be the ability to place them near the wall. Stay tuned!

  6. given the limits of the brief demo, I thought that the AN sounded more “relaxed” and “natural” than the reference, while still having excellent clarity and presence. to my (aging) ears it just sounded more “musical,” and of the two it would be the one that I would prefer to listen to in my home were I making the choice. I think, for what its worth, that you my be about to birth something very special. I wish you well.

  7. Based on my own experience I believe that Paul and his crew are on a good path with the design features of the AN series.
    I used my Genesis 2.5 for about 16 years, most of the time also using PSA gear (PPP, PWT, PWD I/II w/bridge, DS w/bridge II). Back in that time I also contributed as much as I could from Germany in the development of new software for the PWD, the bridge (remember that bloody gapless issue?) and finally the bridge II (I was in the beta user program).
    I always appreciated the open communication of Paul and the folks on the forum. I really learned a lot from them.
    When I was thinking about getting rid of my Gens due to the unavailability of spare parts, I read about Legacy Audio Aeris in the PSA forum. In short words: I invited my son for a week in the US, starting at the Axpona 2015, just to hear the Aeris. It follows the same basic design as the AN (in fact, it’s vice versa): build in subwoofers powered by class D amps, 10” midbass and 8” midrange in dipole configuration and two AMTs for the frequencies above 2.5 kHz. Since Bill Dudleston added the Wavelet from the Swedish genius Bernt Böhmer to the system, it’s (for me) perfect. The “time domain correction” outperforms any other DSP I know (Trinnov, Audiodata, Audyssey to name a few). Since even the best designer can’t know the rooms his speakers have to work with, addressing the phase and timing issues via DSP makes perfect sense to me.
    Since the Wavelet works as a preamp, active x-over, DAC and DSP, I had no longer use for my beloved DS w/bridge II. So finding my “last” speaker at the same time gives me some rest in the never ending search for “the next big thing” in electronics.
    But, based on my almost for years with the Aeris, I believe that the ANs, especially for 12k€, will find its place in the market. Maybe the Böhmer time domain correction, build in a DS or PSA preamp, would add the icing on the cake…

  8. This effort reminds me of a product that was produced back in the mid 1990’s by a company called Flatline Design. The Model 175 was a 3-way speaker that also used the Bohlender Graebener ribbons and ran them down to about 350Hz before a 5 inch mid bass dynamic driver took over. It was an impressive sounding speaker and IMO the first ribbon / dynamic hybrid I’ve heard that successfully blended the two driver types.

  9. I have concerns with all the electronics and processing that will occur at the AN3 speaker AFTER the original signal has already traveled through a DAC, and a pre-amp, and an amplifier. If a digital signal has already been processed by a superier dac like the DirectStream, and then amplified by superior preamp and amp like the BHK, wouldn’t we see degradation when we submit the resulting high quality signal to what would have to be significantly inferior class D amp modules for the bass and midbass as well as inferior DSP digital processing?

    1. These are good and valid concerns on the surface but to get the right answer we need to look a little deeper into the subject. First, we need to look at what the frequency range we’re internally amplifying, and then second we need to consider the specific application benefits.

      The areas the AN3 separately amplifies are quite low, 350Hz down to 20Hz. These lower frequency areas are the easiest for amplifiers to get right as there’s little in the way of harmonic overtones.

      Secondly, it’s important to remember the amplifiers and speaker drivers are engineered to work perfectly together, something an external amplifier doesn’t have the advantage of. This happens very early on in the design process and by ear. Here we can apply what’s needed without concern of other frequency areas. For example, the servo woofer section employs a great deal of feedback to get the right amount of slam—an amount that if applied to a BHK amplifier would harm the upper frequency ranges and you’d be unhappy with the results. But when applied to a specific area like 20Hz to 70Hz there’s all benefits and zero harm.

      Lastly these are hardly “crappy” class D amps. And, there’s no way we could make the AN3 sound as good as it does without using these hand tailored internal workhorses tuned specifically to the task at hand. The BHK amplifier, or whatever external power amp you choose, imprints its sonic character onto the system from top to bottom. The internal amp’s are transparent enough at their specific task to maintain that same sonic character into the areas they do their work.

      1. So I assume there is no way to disengage the AN3’s internal amplification if someone chooses to run them passively or biamped using their own amplification? It sounds like even if there is that option that the AN3’s wouldn’t sound as good driven passively if someone chose that option.

        1. That is correct, there will be no way to disconnect the AN3’s internal amplification of the midbass coupler or the servo subwoofer. The systems are hand tuned to work together and offering the ability to replace the amps would be a small disaster, especially the servo subwoofer whose gain, bandwidth and output impedance must be tailored to the system’s loop. There’s no way anyone could get all that correct with an outside amplifier.

  10. If you’re listening to hi end speakers through your computer speakers or the speaker on your phone you are not hearing the hi end speakers you’re hearing the speakers on your phone or computer amplified by a crappy amplifier. How can you possibly judge the sound that way? It must be an April Fools joke. Lol.

    1. No, it actually works and that’s something that always surprises me. We can hear differences in the big system as recorded on the little microphone in the video and as played back on crappy computer speakers. I am still pondering the meaning of all that but it does work.

      1. OK I think what’s happening here Paul is we are hearing differences between the sound of both sets of speakers and maybe differences in the sound of both rooms even though those differences are not in hi end. But can we really tell which one is better sounding of the two? The better sounding one listened over crappy laptop speakers might not sound better if listening directly inside the room seated in the sweet spot.

  11. A successful line of speakers should all have the same sonic signature From the tiniest speaker to the mammoth flagship floor standing speakers. The only notable differences should be the dynamics and the extreme low end. In fact many times the smallest most inexpensive speakers in the series while retaining the sonic signature of the series actually do certain things that are superior to the giant flagship speakers like imaging for example. It’s going to be difficult or nearly impossible to hear imaging qualities of a certain model when listening over a hand held device.

  12. Can someone tell me what the demo track is? I would like to hear it on my system.

    Interesting series by the way. It makes me wonder how you “voice” electronics.

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