HP's Workshop: Maggie 3.7 Speaker System (TAS 212)
April 7th, 2011 -- by Harry Pearson
Source: The Absolute Sound
Source: The Absolute Sound
Two Magneplanar panels, the spanking new and sexy-looking 3.7s, arrived just after noon this day. And just as the FedEx guy was pulling away, who should arrive but Wendell Diller of Magnepan in the flesh—fresh, like the speakers, from the audio doings in Las Vegas.
Within 30 minutes, Diller had not only unpacked the speakers, but had them set up in Music Room Two, where, three decades or so ago, he had first installed the earliest Mageplanars, the Tympani I-Us, then being distributed by Audio Research Corporation (which arrived along with Bill Johnson, a complete set of ARC tubed electronics, and a full playback system, down to a Decca cartridge). The two Magneplanar panels of the day were divided into three sections six feet high per side, extending in width almost the entire room, itself a shoebox-shaped affair, ideal, as it turned out, for Maggies.
The original Magneplanars were horrifically inefficient and had to be played loud, loud, to achieve a real sense of life, which, in the lower midrange to midbass, they did in a fashion still unduplicated to this very day.
They were also rolled-off in the top octave (not necessarily a disadvantage given the quite “bright” sounds back then), lacking airiness, dynamics, and harmonic extension into the atmosphere(s). They also had to be bi-amplified, and required an external crossover (from ARC, of course) since crossover design was not one of the strengths of the Magnepan products for many a year after.
In a day and age when almost no speaker, perhaps outside of the KLH Model Nine panels and original Quad electrostatic, could reproduce the critical harmonics of midbass fundamentals accurately, the Tympani’s were a revelation because, unlike the electrostats, the I-Us could reproduce the orchestral fundamentals, but also do so with a great deal of power, moving air much in the way air is moved in the concert hall. Thus, they were getting right what no one else could—the basis of music itself. To some, including this reviewer, they were a revelation even if, as revelations usually are, flawed.
Soon enough, all of Jim Winey’s planar designs were under the Magnepan aegis. He set to work on refining the system, usually by lowering the mass of the moving parts of the system, a mass that had been quite high in the original Tympani, thus the lack of efficiency and slowness of response higher up in frequency. He devised a true ribbon for the upper frequencies, one widely admired (and illegally copied) with a “sparkle” and “airiness” new to his speakers, one whose inherent colorations (that “sparkle”) he in time tamed. [It is a little known fact that Winey’s first design, the one shown to ARC’s Johnson, was entirely a quasi-ribbon design, one that originally intrigued Johnson, but didn’t quite work out as a viable product.]
Heard in light of the succeeding models of Magnepans (or Maggies), the $5500 3.7 is a hybrid fertilization of Winey’s true-ribbon design and the company’s more recent ventures into quasi-ribbon technology (as in the Model 1.7s), and it sounds unlike any of its ancestors. It is the culmination of Winey’s art. The technological ins and outs, and their workings, are things the company is trying to keep as secret as Iran its atomic research programs. (See the sidebar.) Maybe they fear being reverse-engineered. No surprise there.
In saying it doesn’t sound like its ancestors, I mean to suggest, before going into detail that the 3.7s do not sound at all discontinuous as they have in the past, but rather as if cut from a whole cloth. Before this (and perhaps the same with the 1.7, which I haven’t heard), the perceptive listener could hear the differences between the ribbon tweeter, the midrange, and the separate bass planar elements, and these differences were audible not only as difference in rise time, but also as a kind of characteristic texture. As Winey’s speaker designs evolved, there was greater continuity within the system, but still, one could pretty much guess where the crossover points were. No longer. With the 3.7 continuousness is so flawless that the speaker sounds as if there are no crossover points. And so, the first thing we heard this day was a unified field of sound.
And so for a moment, a bit of geography. The latest version of the ribbon is in its own panel, placed either on the outer (or inner) edge of the speaker, depending on how you choose to orient the speaker, while the separately mounted midrange and low-frequency quasi-ribbon strips comprising the guts of the system. Or, in Diller’s wording: “The midrange is a narrow, vertical, quasi-ribbon line-source, next to the true ribbon. The quasi-ribbon bass driver is adjacent to the midrange and runs the full length of the speaker; it is so wide that ‘strip’ doesn’t seem like the best word.”
All content, design, and layout are Copyright © 1999 - 2011 NextScreen. All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
Comments
Great preview, HP. Thanks.
Amandela77
Finally -- finally, somebody has commented on the 3.7's efficiency.
OK, maybe I didn't read every review. Guilty. But of those I read, this simple -- but crucial -- bit of information was never addressed in a real-world sense. And that sense I now have is: Don't expect these to need any less amplifier than their predecessors.
Rods: What amp did you use on the 3.6, etc?
I've heard the 1.7 on a typically low powered mid-range all-Naim stack, and it sounded like the PRAT (pace, rhythm, and timing) were there. Used to be super amps were at most 300W per channel at 8 ohms back in the day, and now double that or 1000W+ at 8 ohms with decent sound is available quite readily, in less than super amp categories, starting at 14.000 per pair, or less, for less power.
Would some amp like the Bryston 4BSST2 at a similar price point to the Maggie 3.7 not be amp enough, in your book? (250W conservative), or are you hinting that something like top-end Boulder (or more) is needed to drive a Maggie properly?
Do you need at least a good 1000 watts+ to get any sound out of these panels?
Curious,
Staxguy
Staxguy:
Perhaps I was too leading in my comments because I never owned a pair of Maggies -- although they have always lurked in the back of what's left of my mind as the only great panel speaker my means could deliver. But what has always stopped me from even starting a "Maggies For Me" bank account is that I would also have to turn my general amplification implementation on its head.
All things being equal, I prefer both tubes and the Class A amplification your screen name helped develop in a stunning way -- although I assume I'm not telling you anything you don't already know in that regard. That said, I did once hear a DA-300 on a pair of 'stats (could've been Janszens; could've been Dayton-Wrights ... 40 years in this hobby is not always a benefit). And from what I can recall, that might have been the best five minutes of reproduction I ever heard (I kid -- had to be 15 minutes easy before air conditioners three blocks away started turning themselves on).
Silly jokes aside, though, if I wanted to keep my favorite type of amplification (I'm using Class-A Monarchys right now, in fact), but beef it up to a degree satisfactory for the Maggie 3.6s that was a deal-killer. So I was hoping the 3.7s would change the financial nexus, but alas, not to be.
-- RODS
Yes, “the original Magneplanars were horrifically inefficient.” I bough a par of Tympani 1D’s from Lyric Hi Fi, NY, back in 1979. I went there to buy a pair of Sequerra Metronomes (the large pyramid like with a ribbon super tweeter sitting on the truncated top) to be driven by a Bryston 4B. After about 20 minutes of my wife and I listening to the Sequerras/Bryston combo (preamp perhaps was a Levinson), Mike Kay quitely switched to the Tympani 1 D’s, driven by an Audio Research D-350B (solid State) and the now legendary SP-3 tube preamp. Mike performed one of those unique demos he was famous for. It was a no contest, I just bought the Tympani/Audio Research system on the spot. Mike was sincerely happy never mind that the Tympanis were maybe 1/3 the price of the Sequerras. I was sent to Mike by his brother who was running the Lyric store in Athens, Greece. Soon after, I started to explore Bi-amping with tubes at the top. The D-350B provided what appeared to be plenty of power (500WRMS per channel at 4 ohms). It was actually two bridged D-110s and had the top end harshness associated with bridged configurations until not so long ago. At my dealer’s recommendation I used the Audio Research D350B for the bas/mid panels, and, for the tweeter panels, an Audio Research D-90 (tubes, 90 wrms per side) which appeared to be powerful for a tube amp and with a more delicate top end than Audio Research’s higher power offerings. Problem with this and any other attempt to biamp the Tympani 1D’s is that being crossed over at about 1100 hz, the tweeters handled a large portion of the frequency spectrum and they happened to be as inefficient as the low/mid frequency panels. Tubes at the top made the system sound more musical than single amping with the solid state D-350B, as well as less shrill when overdriven because of power tubes clipping characteristics, but compression and clipping harshness were always still there,thus, making dynamic symphonic passages rather unpleasant, quickly leading to sound fatigue. Accordingly, the regular prescription from dealers of using a large amp for the lows and a smaller (tube) amp for the highs did never really work; for certain, not for listening to symphonic music. At my dealer’s recommendation, I kept changing different parts of my system (many types of hi-end amps, preamps, but, particularly DACs and phone cartridges, thinking problem was perhaps digital harshness or cartridge mistracking) to no avail. I reached a level of comfort by using one of the late Richard Brown’s BEL1001 Mark V for the bottom and a Grant Fidelity tube amp for the top (65 wrms per side, dual mono, class A amp, triode/ultralinear switchable, using 4 KT88’s). With time, I came to suspect (fairly recently, I must confess) that the problem was that 60-100 watts per side tweeter amps were never powerful enough for the actually highly inefficient Tympanis’ tweeters. But, where to find a powerful amp with the required hi frequency characteristic (dear to me) of a quality high end tube amp without the super-premium price, that I actually could try, at home, in my system? Not very easy these days. Fortunately, I stumbled into a couple of reviews of the products from the Digital Amplifier Company. The designer, Thomas O’Brien, does not modify B&O ice modules or anybody else’s. He uses his own and holds several patents in digital circuits (all of this in his website). Prices are extremely reasonable for the kind of power and quality, and there is room for customization (including over 1000 watt monoblocks). More important, he offers a rather flexible 30 day full satisfaction guarantee. I selected two 4800A amps delivering over 640 watts rms for each of the two frequency segments, that is, over 1200 watts rms per speaker, in a vertical biamp configuration. I selected that particular model because a well respected industry player has expressed preference to use that model with his premium tube line stages--I use a Cary Audio all tube line stage.
This configuration, used with a full line level passive x-over (resistors and capacitors, no coils, 6 dB per octave), has completely eliminated the problems that I have been suffering for over 30 years. Now, I have no compression or shrillness, regardless of musical program. Just fully coherent and extended sound, with limitless, distortionless dynamic range and the same mid and high frequency smoothness of my favorite tube amp I was previously using. This in addition to the exceptionally extended and controlled bass that appears to be a trade mark of this brand of amps-- in this case being delivered through the Timpani ID’s bass/mid panels, themselves still considered a benchmark for bass and low mid range reproduction, as Harry Pearson states it in this review (and in Part II, in the current issue of The Absolute Sound). All of this with a phenomenal sound stage of over 19 feet wide, close to 7 feet tall ( I have 12 foot ceiling), and depth reaching far beyond the wall behind the speakers. Plus all the audiophile attributes: musicality, resolution, detail, excellent transient attack and decay, pin-point imaging, three-dimensionality, palpability, etc. I had those elements before to a large extent, just even more of it now. Most impressive is the density of the sound field, to the point that it feels as if one is being transported to the performance itself and drawn into it…actually, it is like one is being immersed into the sound field and engulfed by the body of musical energy filling the performance’s venue. I am sure most of this is related to the immense power response that these amps make possible for my system to deliver in my largish room, in combination with the sheer size and nature of the Timpani 1D’s themselves (fully refurbished by Magneplanar 7-8 years ago, a significant improvement in itself due to improved materials and work techniques, after 30 years), but I have no doubt that all the observed benefits are a result of a synergy or, better said, the ability of these amps to deliver all the driving power with the quality that Magneplanar's very fine speakers evidently require.
dr.larkos
HP'S WORKSHOP TEN MOST SIGNIFICANT AMPS (TAS 213) Very enjoyable article. (Since the article isn't online yet, I am putting my comment here at the previous HP's Workshop article.) A minor correction: The Citation II was not a mono amp. It was a 60 watt/channel stereo amp. I owned one in the 60s and drove stereo stacked pairs of AR-3s and Bozak 313s. Using one of each of these speakers on each side gave me something like a poor grad student's version of the Bozak Concert Grands. The amp had plenty of power for all 4 speakers. If you look at a picture of the amp, you will see 3 transformers, 1 for power and 2 for stereo output.
BMK47
dr.Larkos describes a familiar journey. I would like to know what crossover network he is using in his system. I have found the Belles 350A Reference a suitable amplifier to run the high end. This solid state amplifier has a tube like quality and is 350 watts per side into 4 ohms
Stuart, what type of Magneplanar speaker you have? The network I am using was specifically desgned for the Tympani 1D's, to replcate the cut-off points (-3dB) of the first order, two way, parallel built-in speaker line X-over. Accordingly,this passive line level X-over is extremely simple--a first order, two way, parallel one. The X-over was designed taking into consideration the output impedance of the preamp and the input impedance of the power amps. In its initial configration (when I was using the BEL1001 MkV for the lows and the Grant Fidelity for the highs) each channel had one 4.9K resistor in series and one 47nFcapacitor in parallel for the low pass filter, and one 4.7nFcapacitor in series for the high pass filter (only three element per channel). Once I changed to the vertical biamping configuration using the two Digital Amplifirer Company DAC4800-A amps, I added a resistor in parallel at each amp input so that the resulting input impedance matched the value of the original amps (it was simpler than recalculating the network values with no sonic degradation given the high quality and relative small resistor values--just a small increase (1-2 dB) in gain loss, wich was minimal to begin with). The network was designed by a third party audiophile using one well known electronics modeling program (I never got the full specifics). He used an iteration that selected the R/C values which provided the desire filter shapes while minimizing the gain losses inherent to line level passive x-overs. The advantage of this type of X-over is that by definition (no power supply, absolute minimum amount of parts, and minimal signal handling or manipulation), it is much more quite with relatively much less distortion than an active line level x-over (of, course using quality R/C components for the pasive, but no need to go crazy here). However, it obviously lacks the flexibility of an active when you change preamp or switch to amps with different input impedances and/or gain characteristics.
Back then, to facilitate passive biamping, Magnepan used to market a hybrid passive line level X-over, the XO-1, which, I believe, was designed and initially marketed by Audio Research (who also marketed the Timpani ID’s and earlier models as HP mentions in his review). The XO was, later on, marketed by Magnepan for the MG-IIIA’s, MG3.5’s and Timpani-IVA’s. The XO-1 was not a pure PLLXO, it allowed line level passive filtering of mid and high frequencies but the low pass filter still used the speaker level coil network (“miles & miles of wire” still in the circuit). So, the XO-1 allowed only for a hybrid passive crossover as both line and speaker level components had to be used. Transparency, detail, noise floor, resolution, transient response, and frequency extension at both ends all improved with the full passive line level x-over I am currently using. More notably, the Tympanis now have life at low volume levels. As volume is lowered, the effect is like when one moves back in the concert hall, to the back of the balcony, away from the performers, but the three-dimensionality, frequency balance and lower volume detail is still there (within the normal low volume response pattern of the human ear). The height of the sound stage is reduced a little bit but the depth keeps increasing as the performers move back. It could also be a combined result of the vertical biamping configuration and the power response at lower volumes of the DAC4800A amps I am using..
dr.larkos
Dr.larkos I have lived with and enjoyed the Tympani IIIA for many years, crossed over at 35 Hz to a pair of KEF B139 woofers housed in a large transmission line enclosure (6 ft x2 ft x 1 ft). I eventually switched to the 3.7. Upgrade might not be the best term. The trade off of the ribbon tweeter never made up for the loss of the IIIA’s midrange and upper bass quality. The only way I eventually found to achieve some approximation of the IIIA Midrange/upper bass and better blend the panels was to biamp the 3.7. It may be counterintuitive, but I found that a correctly chosen powerful tubed amplifier for the bass/midrange and the Belles 350A Reference on the tweeter comes close. I have tried an active crossover, totally bypassing the Magnapan crossover, and the XO-1, and settled on the XO-1 using a high quality MIT capacitor. The KEF woofers remain using a modified Dahlquist crossover still set at 35Hz.
Stuart
Stuart, you definitely have gone the same path, with a caveat, your Tympani IIIA sytem has or had eight planar-magnetic panels 6'(h) x16"(w) x1"(d) each, plus two transmition line bass-towers 6'(h) x2'(w) x1'(d). That is, you have or had four bass panels (two per channel) and two pairs of tweeter/middle panels. My T1D system does not have the extra two bass panels (it has six panels per side identical to yours). I do use two Sunfire True Subwoofers (latest model) very succesfully crossed over at 35HZ (like yours). The fact that I have 12' ceilings (thus, avoiding the bass anomalies tipically associated with 8' ceilings--I don't need to use at all the sunfire's build in room correction system, which I prefer to avoid, anyway) helped a lot in the integration of the suwoofers (as you know, using two subs also help in cancelling out other room bass response problems). In any event, with all those bass/mid panels, no wonder why you are spoiled for life with the Tympanis' sound, and any new Magneplanar product iteration feels just a "switch," a compromise, not a unqualified upgrade!!
Incidentally, I thought you could not bi-amp the 3.7's; W. Diller told me so a few weeks back. I guess he meant that there are not connection features at the X-over plate to facilitate that, and you are actually getting inside the speakers, am I correct?
dr.larkos
dr. Larkos. I must have the 3.7 on my mind. I have been using the 3.6R which is not is not very difficult to biamp. The high frequency output from the XO-1 connects with the tweeter amplifier. The speaker output of the tweeter amplifier is connected directly to the MR/TWT connection on the 3.6 (bypassing the Magnapan crossover). The bass output from the XO-1 is connected to the midrange/bass amplifier which is then connected to the “Low” input of the Magnapan crossover (the jumpers are removed). The Bass output from the Magnapan crossover is then connected to the Bass terminal on the 3.6. I have not seen the 3.7, but would assume (perhaps incorrectly) the 3.7 has the same setup. You are quite correct regarding the importance of room dimensions on low frequency problems. My system resides in a room with a 15 ft high cathedral ceiling. My Tympani IIIA’s did not include the bass panel, only the midrange and tweeter panels, and the KEF 309’s to cover the bass.
Stuart
The 3.7 does not have an external cross over box as does the 3.6. There are inputs for the speakers directly, and two jumpers for the tweeter and midrange resistors if needed, as well as the fusing.