The Absolute Sound posts a video about a new product development from Magnepan, but...


 

...fails to give any mention of the impetus for the development, for which Tom Martin has coined the term "Open Architecture". Here’s the full story:

 

Danny Richie of GR Research offers this service to his customers: Send him a loudspeaker you want him to evaluate, and he will put it through all his tests at no charge. He takes measurements of frequency response (on and off axis), cross-over characteristics (and the quality of the cross-over parts), individual driver responses, a spectrogram/waterfall plot, evidence of comb-filtering, impedance, sensitivity, etc. He evaluates any failings he finds, to see if he can develop measures to remedy those failings. Danny is a well-known expert at cross-over design, and if he feels the loudspeaker has the required potential he puts together one to "fix" the failings he finds in the speakers sent to him, selling the x/o in DIY kit form.

Over a year ago he received a Magnepan MG3.7i for evaluation, and ran it though his full battery of tests. In the video he posted on YouTube (see below), he describes his findings on that model Magnepan. He came up with a major redesign of the cross-over, to eliminate what he considers the MG3.7i’s failings. Finding fault in the measured frequency response of Magnepans is not new, but before dismissing what I just wrote, consider watching the video.

What he found was that the Magnepan cross-over slopes results in the three drivers (bass, midrange, tweeter) over-lapping each other, reproducing the same frequencies at the top (bass driver), top and bottom (midrange driver), and bottom (tweeter) of their ranges. That can cause comb-filtering, which is exactly what Richie found in his measurements of the MG3.7i. A lot of it. That filtering wreaked havoc on the response of the speaker, with lots of phase cancellation occurring due to the same frequency reaching the listening position from different drivers at different times (the definition of comb filtering).

Danny also found the cross-over to be comprised of absolute junk parts---push-on connectors, steel nuts and fuse assembly, electrolytic capacitor, iron-core inductor, etc. But THAT was already well known about Maggie cross-overs, with many after-market products offered to replace the stock parts.

So Danny created a new cross-over, which you will hear about in his video. What I want to highlight here is that he made a new cross-over plate to install in place of the stock one, but that plate merely holding three sets of connectors for the three drivers. Those inputs are fed from a new, separate x/o box, with all new x/o filters designed to---amongst other things---eliminate the comb-filtering, allowing the three drivers to create a beautiful frequency response. Danny suggests anyone considering the purchase of a pair of the MG3.7i to ask Magnepan if they are willing to make a pair without a x/o, in it’s place three jacks connected directly to the three drivers.

In his video, Danny also mentions how his GR Research/Rythmik Audio Open Baffle/Dipole Sub makes a great partner for use with the MG3.7i, or any other dipole planar for that matter. I’ve been touting that combo for years here on Audiogon.

 

So, I see the heading of the TAS video (posted below, if all goes well), and start watching it. One of the first pics I see is a pair of MG1.7i, with three sets of jacks where there is normally those crappy Magnepan speaker cable binding posts! I guess Magnepan has also watched Danny’s video 😉. On top of that, standing next to the MG1.7i’s are open baffle/dipole woofers!

Magnepan has been talking about offering an OB/Dipole sub for use with their planars for several years now, but there is already a dipole planar-magnetic loudspeaker with integral dipole (though not open baffle) woofers---the Eminent Technology LFT-8c. One can also add a pair of the GR Research OB/Dipole subs to the ET LFT-8b, using the OB sub in place of the LFT-8b’s monopole woofer. Just leave the 8b’s connecting strap off the woofers binding post.

 

 

 

 

While Danny’s x/o keeps the single-amp design of the 3.7i intact, Magnepan’s design requires three separate power amps, one for each driver.

128x128bdp24
Post removed 

 

That Maggies sound as good as they do in spite of their crappy x/o parts is a testament to the quality of Jim Winey’s basic magnetic-planar driver (and ribbon tweeter) design. Hard-core Maggie enthusiasts have been upgrading the cross-overs in their speakers for decades.

I bought my first pair of Maggies---their original model, the Tympani T-I---in 1973, and now have a pair of Tympani T-IVa. I don’t run them with the stock cross-over (sorry, it’s junk. Unless you like the sound of iron core inductors, steel parts in the signal path, etc.), using instead a First Watt B4 x/o, designed and built by Nelson Pass. It’s an all-discrete design, using high quality resistors to provide an amazing number of filter x/o frequencies (from 25Hz to 6275Hz, in 25Hz increments) and slopes (1st, 2nd, 3rd, and 4th order---6, 12, 18, 24db/octave)..

 

Speaking of improving: It was Winey’s magnetic-planar driver that inspired Bruce Thigpen of Eminent Technology to develop his LFT driver. He saw that the Maggie drivers were single ended---magnets on only one side of the Mylar diaphragm, and thought "Why not make a planar-magnetic driver that is a balanced, push-pull (magnets on both sides of the Mylar) design?" So he did just that, and named it the Linear Field Transducer, or LFT.

 

More details on the upcoming "super" Maggies: The cross-over will be active, just as is the x/o provided with the Linkwitz LX541.4. It is inserted in the chain in between the pre-amp/sources and the three required power amps (again, just as with the LX521). The x/o Danny Richie offers is of course passive, receiving it’s signal from a single power amp. Two very different architectures.

 

 

Ric---Thanks for the heads up on the Rich Hollis info. I know Rich has had interaction with Danny Richie for years. I used to read the AudioCircle Forums, don’t know why I stopped.

 

The OB subs pictured in the TAS video are four of the GR Research 8" woofer models per channel. They probably put out about as much bass as do two of the 12" woofers. Are 8" woofers "faster" than 12" woofers? Danny says (as does Brian Ding of Rythmik Audio, with whom he collaborated on the OB Sub) in this case no. It has to do not only with the moving mass of the driver cone, but also the the motor (magnet) strength, the "settling time" of the cone, etc. The servo-feedback design of the Rythmik woofers/plate amp combo is another factor in how fast the OB Sub sounds. "Stops on a dime". 😉

Brian Ding offers the 8" woofer for instalations where a higher x/o frequency is desired. The 8" can be used up to a higher frequency than the 12", though the 300Hz limit on the 12" is already much higher than is the case with "normal" subs.

 

I have heard many of Danny's supposed tweaks - IMHO nothing really there.

Just me though.

It is never just one thing that is important....Everything makes a difference.  For instance, you can see on the Audiocircle site the latest thang from Hollis labs....He has a 1.7 maggy that he is tri-amping.  However he put junk binding posts on the speaker and still has the fuses in line.  These things wreck the sound.  WBT Nextgen posts and no fuses would sound way better.  Will Magnapan do the same with weekend on their tri-amped speakers at Axpona?  And whose class D amps are they using (stock Hypex?).  When LInkwitz showed his speaker in LA a few years ago they did an AB of the stock Hypex Ncore amps he normally sells and some Pass amps......most everyone preferred the Pass amps.  Bruce at Eminent showed a prototype super triamped speaker at the Florida show a year ago and evreyone thought is was way cool.  I bet he was just using stock Hypex amps in that thing.  Your system is only as good as the weakest link.  

I suggested to Bruce at Eminent Tech that he sell his new more effecient mid and highs planars on the baffle he showed at that show seperately (no woofers).  Just wires hanging down.  He would sell a ton at $3k.   This way you can get his latest 90 db drivers on a baffle and make your own woofs and actively or passively crossover yourself.....for the best sound possible.  Basically this is what Magnepan is going to do.....sell you the speaker with no xover parts and just binding posts on the back.