Dream setup

August 26, 2016
 by Paul McGowan

If I were to roll my sleeves up and design a speaker system from scratch, I'd not use passive crossover components connected directly to the drivers—as is done in 99% of all speakers. Being an electronics guy, I'd self amplify this mythical speaker, assigning the right amp for the right driver: class D mega amp for the woofer, tube input/MOSFET outputs for the midrange, simple all-tube design for the tweeter.

To separate the frequencies, I'd use simple passive networks of very high quality, small size, working into the high impedances of the three power amplifiers.

Of course, this is only a dream, one I wouldn't even think to begin.

Even under the best of circumstances, it'd probably wind up sounding like crap. Why? Because I am not a speaker designer. I would be like the proverbial wanna be culinary expert that knows just enough to be dangerous: best ingredients + great recipe + no skill = good, not great.

But the concept I laid out is still a better alternative—in the right hands—than what most speaker designers have to work with. The smaller components in the passive networks can be of exquisite construction, the amps tuned perfectly to the driver.

Yet rarely do you see such a setup because it is simply not commercially viable. Audiophiles like to pick their own amplifiers and, besides, even if that weren't true, designers would be taking a huge gamble that their tastes matched yours. Once you go down the rabbit hole of an inclusive design, you're limiting people's choices.

That can be a good thing, but also, a scary thing.

There are those brave designs like the Devores, Meridian, Genelux, Emerald Physics, and I am certain I have missed a few.

Tomorrow I'll tell you about a speaker project I built years ago and what made it unique - and yes, it was active.

Subscribe to Paul's Posts

35 comments on “Dream setup”

  1. In the end even Bascom King voiced his amp design based on a specific loudspeaker-room setup. Or did he voice it via headphones? The question remains: what makes him confident that having voiced the amp with speaker A and Music Room 1 will work for speakers B to Z and music rooms 2 to 20000000 as well?

    1. ..which is why I say:
      1. there is no "neutral"
      2. people need adjustment options somewhere. As many as possible with no impact to sound quality (which is possible...with active design)

    2. Which is a great question and yet there's a reasonable answer.

      Bascom's speakers happen to be one of Arnie's older designs. But that's not the important point. The point is, the speakers are extremely revealing. And that is the point. As long as designers use speakers with revealing qualities that hide nothing and emphasize nothing, in other words, relatively neutral in tonal qualities but revealing - then that is a great speaker for designing and voicing electronics.

      I have owned many speakers and not all were great choices. Years ago I had a pair of Revels that I loved, but unfortunately, they sweetened the sound of all that was played through them. And a pair of Avalons did the same thing, though to a lesser extent. It's taken time to learn what it means to be revealing and neutral, but it is possible and that's what you want.

      1. Isn't for example a Maggie a revealing but not necessarily suited speaker for product voicings, as it's very special in many regards compared to the majority of speakers?

        1. Maggies are revealing and I have used them with great success for years to voice equipment.

          Electrostats are revealing too, but limited in other areas (like dynamics) too much to be useful.

          It's a tough blend of revealing, uncolored sound, without too many restrictions one way or the other. Not many speakers qualify.

          1. ok, didn't know that...I just had two smaller ones long time ago (SMGa and 1.4) without ribbon. I guess you had the real thing at the time.

            So mine were best value for small money at the time for me. I just noticed too rich upper bass to compensate the missing low end and also missing top end extension (probably due to the missing ribbon) compared to what's possible with electrostatics and conventional speakers. Generally they sounded quite forgiving to me in terms of bright/dry/thin recordings and for sure they were placement critical and didn't like overdamped rooms.
            They can work very well, but I wouldn't call them easy and typical.
            I'm sure you had an appropriate setup with the bigger ones.

            1. ..and I can't say mine were dynamic sounding...but they had "impact".

              I loved them but a product I would have voiced with them would have sounded quite bright/thin with too much lowest end I guess and all that very much depending on room and room damping I used.
              But I'm sure with larger models and tight, fast amps this is different.

      2. I understand that designers of all components need a reference system that they know well. But,I have asked this question before, on the LANRover, and never got an answer.
        Do you, and other designers, at some point, test your new designs with a variety of different equipment? With speakers, that would involve maybe three or four other amps, that are quite different, for example solid state, tube, class d, and whatever is a lower priced, very popular amp. Maybe one of the midrange Parasounds a 21 or 23?
        I understand that almost every speakers needs certain minimums of performance, but if you look at the specs, there is often a recommended range of power needed, say 30-200 watts recommended.
        And with components, they need to be able to function in a variety of setups.
        I'd like to see, warnings, with components, that would say, will not work well, with certain types of equipment. When Atkinson tests equipment, he will say that a component or speaker will not be compatible with say a component with low impedance, or low current, etc.
        I realize it might lose a few sales, but it would save a lot of aggravation.

        So, the question is, do you test your components with a variety of associated gear, to be sure that reasonable results can be had with a wide range of gear?

        Having a component that will be great in all but a few systems, seems like it would be part of the design initiative. There are tweaks that are designed for a specific component. An example of that would be an upgrade to say a Linn tt. You wouldn't want someone with a Rega, trying to jam a Linn specific part into it. So, wouldn't some general compatibility info, be good for everyone? How often have you heard someone say they will never buy that brand again, based on one bad experience, an experience that may have been their own fault. The people on this newsletter list and in the forums are a lot more aware than the average buyer.

        1. It's a great question and I apologize for not having responded to it sooner.

          We do and we don't.

          Officially, we have one reference system as you're probably aware. The system in Music Room One is state of the art in terms of its ability to easily show differences and gives us an unwavering reference we can count on day in and day out.

          But no reference system is perfect.

          We rely on others as well. The system of Ted Smith, Arnie Nudell, Bascom King, and a handful of other trusted listeners. Units are shared, opinions proffered, evaluations considered, and we make decisions.

          I wish the company were big enough to have multiple listening environments. There are systems at our homes and these too are valuable.

          Some day, I'd love to move to new quarters where we will make a point of having several completely different listening environments.

          For now, we do what we can with the resources we have.

          After all that, we then rely heavily on our customers with a beta testing program, typically 50 to 60 strong. That's a lot systems to test amongst and the results are always instructive.

      3. 'Revealing' doesn't seem to be a technical term or method of measuring as the theme you opened when addressing frequency or step response. Thus the approach seems to be similar to the way traditional healers act - not at all science based.

  2. It was also a big step for me to switch to a fully active design.
    Previously owning several heavy and more or less hot and powerful amps from Threshold, Gryphon, Pass, Vitus Audio and Lamm with passive speakers, it was hard to imagine, that much less sophisticated, large and powerful amps built into a speaker without passive crossover could give a benefit over those babies. But except slight restrictions at really over the top SPL's, they did so with their advantages in control, fast response behaviour, dynamics, homogeneity and adjustability.

  3. One of my first foray into the realm of "hi-end" was the Paradigm Active 40s (version 2, to be exact). Not the last word in resolution, low level details, soundstage, etc. But as jazznut alluded to, the level of control, dynamics, and balance made it a great choice for me in my late 20s.

    Wes Phillips reviewed these speakers back in the days. I think he summarized it perfectly:

    "No matter how good they are, no matter how convenient, logical and sensible they are, I know that audiophiles won't buy 'em, because audiophiles don't buy active loudspeakers. And non-audiophiles won't buy 'em because non-audiophiles don't buy $2000 loudspeakers."

    As a result, Paradigm ceased to produce these speakers.

    Fast forward 15 years later. Although I have since upgraded my main system, I have been very reluctant to part with my Active 40s.

    Just my 2 cents.


    1. Maybe most music lovers have no dedicated listening room and have to respect the WAF. Thus design or better styling aspects come into the game. The speakers and the electronic components have to be stylish primarily. Look at the huge efforts made to cut a cabinet for a speaker or a 'simple' SSD based music server from a big block of aluminum. As a music lover I don't care about styling. I even don't require to have the system placed in my listening room - with the exception of the (active) speakers. The turntable greatly profits from this remote placement! 🙂

  4. As I've written a couple of weeks ago the most impressive demo of a loudspeaker in the last twenty years was for me the demo of the new B&O Beolab 90.
    These fully active, fully digitally controlled loudspeakers are compared to allmost any other setup of speakers and amplifiers in that price-range a bargain.
    These are the most covincing speakers to me.

    1. Being neither a line source design nor a point source design the pse speakers are simply a sound machine as judged from a puristic point of view. However the concept hits the requirement of 'customizing' the speaker's output to the listener's preferences and the listening room's acoustic!

  5. Paul, what would you say:

    if you have a power amp selling for 10k and a passive speaker selling for 15k in your product line.
    What would you think would cost that quality speaker with that amp quality put and fit in as active design?
    And what with not that, but a sufficient amp quality necessary for an active design put in? (in case there are lower or other demands for active amp design than for passive)

    1. If I understand your question correctly you want to know how taking a $25K setup and building it in would affect overall cost?

      It's a good question and I haven't given that any thought, to be honest. Of course you'd save the cost of the metal chassis of the amp - and that's never a small amount of the amp's cost.

      However, I don't suppose it would make sense to do all that just to save on the cost of the chassis. If I were to internally amplify a speaker, I'd use multiple amps each tuned to a specific task. It's be an entire redo so I haven't the foggiest notion of what that might cost.

      1. Yes I mean the transformation of two BHK monos and such a speaker into an active design with 2 x 3-way matched amps inside the same speaker as active design and the question how such active design amps differ from what's needed for passive speakers. I guess you wouldn't need the power handling of a BHK inside an active speaker.

        I was not sure if you might have experience with such thoughts.
        The amps seen in active designs are usually less class A biased (for sure at least than the extreme amps) and have considerably smaller transformers and caps.

        It seems for active designs a very specified, but (regarding material) much less costly amp design is necessary than for passive designs.

        I initially left the housing out of my consideration.

        It looks to me as i.e. two mono blocks with a sales price of together 30-60k could be replaced by active amp design of let's say 10-20k (looking just from the material side, assuming construction effort is similar)

        At least that's what designers of active speakers say, that much less power is needed if there's no passive crossover.

      2. Paul, you say the chassis is never a small part of the cost.
        I guess your chassis design, which is used in its general design for all components is one reason why you can make more cost effective products than some others, correct?

        I like the puristic design, especially the size of the DS chassis, but although it's very good quality, it looks less difficult to manufacture than others on the market (especially in that quality range of products).

        I generally also like understatement, but to be honest, the design of the PSA products was the reason that even thoug I knew the high end market quite well, it took a long time until I noticed PSA first time (which was at availability of the DS).

        The design for me somehow doesn't reflect the quality and innovation status of your products, althought I like it and have no reason to complain about it's quality etc.

        I just tell this in case it's interesting for you. Maybe it's only a single opinion. It's strange that I also prefer that it is like it is because I like it and wouldn't really want to pay 1k more for a complicated design chassis. Anyway I see how it influenced my perception long time.

  6. Paul, no idea how much detail you want to get into with designing active speakers. But I'm sure it must be complex, more so than simply picking an appropriately power rated amp for each driver.

    I know from discussing it with him that John Dunlavy worked on a large, active, full-range design over the last few years he ran DAL. He believed it had greater potential in capturing the "live experience" than any of his passive designs. However I'm quite sure it never got beyond the prototype stage, and he was a smarter designer than many others.

  7. In my experience, there are many levels of clarity in amplifiers but they are swamped by the inequities of speakers and rooms. The analytical superiority of measuring voltage response and being able to safely ignore the complexities of impedance is enough to justify an amplifier per driver.

    If you are concerned with peak output, as in acoustic music, rather than the sustained continuous levels of compressed pop/rock/hip-hop/electronica, then multiple small amps are more dynamic than one big one. From a system standpoint it is also superior to have high efficiency drivers. I use, for example, LM3886 tri-amped to achieve uncompressed 120dB peaks with 98dB/Watt drivers. This approach using uncompressed music needs a tweeter that will handle a 60 Watt peak like an AMT1. This chip is Class A for the first couple of Watts and is quite refined given a good power supply and layout.

  8. I also have a design for a speaker system that meets my criteria as a 2 channel system or as the front channels of a more ambitious system.

    The system has the following design criteria;

    The first arriving sound must have flat FR
    The reflected sounds from the wall behind the speaker and the side walls must have flat FR
    The ratio of direct to reflected sound from different directions must be adjustable for each direction of arriving reflections
    The speaker must be adjustable to perform the same way in most rooms typically found in homes
    The speaker must reproduce the entire audible range without any BS.
    The speaker must sound the same everywhere in the room to the degree that such a thing is possible.

    The system is a 4 way design having electronic crossover, independent equalizers and amplifiers for each directional segment and the subwoofers.
    For each channel there are six 3 way segments (lower midrange, upper midrange array, tweeter array., One segment firing forward, one firing diagonally backwards to the left, one straight back, one diagonally back to the right, and one straight up at the ceiling. It is a direct indirect design.
    The woofers are an array of acoustic suspension drivers that extend across the floor the width the front speakers.
    The front firing 3 way segment must have very wide dispersion at all frequencies and will include dimensional arrays of midrange drivers and tweeters to achieve this goal.

    The input signal must have independent equalization to adjust for variables of input signal spectral variables to work optimally with any source device and recording.

    The success of the design does not depend on using the most expensive drivers obtainable. Good quality drivers that meet these criteria and good quality amplifiers which audiophiles deride as "mid fi" are sufficient.

    On my first job I was advised not to build what I could buy. The tweaking is performed in the process of adjusting the installation, not of endlessly and hopefully fiddling with the design in a one size fits all fixed speaker system that can't be adjusted and must always be redesigned to make any improvements. The design also recognizes the limitations of room treatments and exploits the room's acoustics rather than fights it, a hopeless battle every high end speaker designer loses.

    1. Even this rig will be compromised. A few years ago at a CES I asked the gentleman running Kimber Cable's 4 channel demo to turn off the back two channels. The sound went from magnificently real sounding to just normal hifi.

      I get great satisfaction from two channel at home having done everything in my power to optomize my music delivery within my skills and finances. But since it is obvious by now that we will NEVER have much in the way of high end multichannel, aren't we forced to just compromise and admit that we are striving for Nth place at best?

      1. The setup I described above is the most practical solution to the "they are here" problem for a two channel system that is adaptable to a wide variety of recordings. The acoustics you hear will be largely the acoustics of your listening room. Acuvox's solution is technically superior but it is not practical for home use by people who have a collection of commercially made recordings.

        My interest in acoustics started with my failed experiments with four channel sound in 1974. I concluded it could not be made to work. I'm not going through my own ideas here again but I figured out my own solution to the problem and it is very different. The only one here besides me who heard a scaled down and adapted experimental prototype is Paul. It was originally conceived of as an instrument for scientific research. But after 40 years I've lost interest in it. Too bad, I've got access to an enormous anechoic chamber that hasn't been used in a very long time. I haven't seen it yet but it's probably very large. They tell me it's now the second quietest room in the world (used to be the first quietest.) If the technology were to be developed, that would be the place. Ho hum.

  9. The necessary ingredients for a really good anything in audio including speakers are-1. A lot of patience 2. a lot of time 3. A lot of elbow grease 4. A pair of good ears. Designing experience cuts down on time by avoiding mistakes the newcomers make. Add to all this a never say die attitude. The sky becomes the limit keeping in mind the limitations posed by the law of physics. And lest I forget the most important is bearing allegiance to no particular technology. The latter helps thinking outside the box easier. And remember universal appeal is wishful thinking. Regards.

    1. Audiophile sound systems are assembled like a tossed salad. An amplifier here, a speaker there, this cable, that CD player. Each one is designed as a stand alone unit, supposedly the perfection of its function. But there is no system engineering, no real system performance goals. And there are no adjustments to adapt the system to even the most common variables encountered such as room acoustics and different recordings. It's like buying a race car that can go from 0 to 60 in one second but can't turn because it has no steering, can't go in reverse. What happens when you reach a corner and there is no road ahead. Sooner or later anyone who buys one of these high end sound systems becomes frustrated because there are recordings they like that sound bad on them and there is nothing they can do about it. All they can do is shop and swap or look for a promised magic bullet that never hits the target. And it always costs more money. It's not just the equipment that is badly engineered, not just the systems, not just the concepts behind them, but the lack of thinking and skill and the insistence that one side fits all so that no adjustments are needed because it is already perfect.

  10. Does anyone remember Henry Kloss's powered Advent speaker. I sold these at Opus2 many years ago. Very progressive for the time, reliablabily with the amps was a problem. Also the Accoustat X was a brilliant solution to driving electrostatic speaker and the amps out of these are selling for thousands of dollars, if you can find they.
    I have sold and owned Maggie's for 40 years and currently have the 3.7i and enjoy them very much driven with ARC vintage tube gear, SP11mk ll, D250 MK ll. And while at it let me stick my neck out, after selling audio of 40 years I have gotten so tired of hearing about the WAF. Maybe most audiophiles are PW by their wives or maybe audiophiles should be as careful at picking a wife as they are their equipment

  11. The Linkwitz speaker design approach sounds very close to what you explained, with the exception of passive crossovers. DSP processing controlling multiple amps driving multiple drivers.

    The great sound of his rooms at the shows continue to amaze, given the eccentric design.

  12. Funny you should say that, since that's pretty much the kind of system we at ICOS have been doing for over fourty years. It is commercially viable but indeed most of our crowd are not the typical audiophiles you described. Of course 40 years ago, we didn't have class D for bass and would keep them mono on a very big woofer, separate from the medium and treble speakers ; and we don't use tubes or Mosfets. But still, we agree that it's the right idea and sounds awesome. Dealers and amp manufacturers are indeed not crazy about it.

  13. "Maybe most audiophiles are PW by their wives or maybe audiophiles should be as careful at picking a wife as they are their equipment"

    The most audiophile remark I've ever read in these posts. And for me skip the word "maybe".
    Every (would be) audiphile should know the 2 basics if you like to have a decent sounding system.
    1. DON'T pick the wrong spouse. Better don't pck a spouse/girlfriend at all.
    2. DON'T be an audiophile (Soundmind's law)
    Okay, just kidding. Forget number 1.

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