[medium jim]

Kevin:

Your last paragraph would tend to support Alnico magnets as they come in different strengths, Alnico 3, Alnico 5 and Alnico 9 if I recall correctly.

Jim

[studiotech]

I wish I could find it, but somewhere on one of the forums years ago(it may have been diyaudio.com) Dan Wiggins of Adire Audio and XBL2 fame wrote a series of posts describing the various characteristics of common magnetic materials.  Their history, strengths and weaknesses were all covered very well as related to loudspeaker design.  There were also more esoteric posts about the sound of each material. 

Greg

[josh358]

$8K more for a speaker that's 30% larger with non-Neodynium magnets on each side?

How about a dual magnet version of the 3.7s for people with smaller rooms?

And what caps does Magnepan use in the 20.7s?

As I understand it, it isn't just the materials, it's that they're more difficult to make. Labor costs matter and these are going to go up because of the larger diaphragm area, doubled/larger magnet assemblies, and extra foil (both are placed by hand), as well as the extra labor of assembling the dual driver. AFAIK, there are other differences as well, e.g., I gather the baffle is thicker than  it is on the smaller ones. I'm not sure what other differences there are, or what caps they use. Presumably, they also have to amortize the R&D cost over a smaller number of speakers -- they sell a lot more 3.7's than 20.7's.

I suspect a dual magnet 3.7 would be significantly more expensive than the single-ended version.

[josh358]

Admittedly I am really curious as to the way the pole piece magnets are magnetized. I've seen some internet videos for some bar magnet machines being magnetized. No big deal. But since the edges of the Magnepan pole piece magnets are the N/S poles does that mean they have their own specific jig? Do they have to do it all at once can it be done a section at a time?

Here goes a whole wasted morning trying to Google that ...

Each model has its own jig. The precut magent strips are laid by hand, two at a time in alternate channels, then, once they're all in place, one of the magnets in each pair is flipped over into the adjacent channel. A little polarity checker is then run across the magnets to make sure that no mistakes were made in the orientation. The ends of the magnets are trimmed, adhesive is srpayed on the magnets, then the jig is closed like a steam press to glue the perforated pole piece (which has previously been sprayed with adhesive) to the magnets. At the next station, excess adhesive is removed.

[josh358]

Thanks for posting that Kevin.  It seems counter-intuitive to me to use magnetic material of differing strength when I think the ideal is to drive the diaphragm evenly over it's entire surface.  Hmmmm, I'd like to know more.  I'd still have to imagine that maybe that was the design for the original patent, but not how they are created for mass production.

I was also under the impression that both the 3.7 AND the 20.7 were push-pull magnet designs.  I guess I was incorrect and it is only the 20.7?  In either case, I think this proves that even in Magnepans minds, dual sided are the superior performer, but not for all cost points.

I think a little love is due to be shown to Eminent Technologies as well.  They ARE push-pull planar designs and are also VERY affordable.  They don't get nearly as much press or hype as the Maggies, but are consistently revised through the years to stay current.

Greg

In some models, they doubled up the wire towards the center of the diaphragm to increase Bl product there. I imagine the variable magnet strength served the same purpose. The reason this is done I think is that the diaphragm doesn't move in a plane. Rather, it dishes, with the center of the diaphragm having the most excursion. So the magnet assemblies are actually bowed out to allow for large excursion towards the center.

Doubling the magnets in the 20.7 allows them to increase field strength and therefore Bl product. It also makes the field more linear at large excursions. So it's superior in those respects, but since the dual magnet drivers are significantly costlier they apparently don't consider it the best trade for the less expensive models.

I have the Eminent Tech drivers on my Monsoons and they're very good. But they have limited SPL's and use costly neo magnets. I've read the Eminent patent -- I think it's here -- and IIRC, the Eminent driver design is intrinsically inefficient and the neo magnets are used to make up for that. They aren't like the BG Neo series, which achieves very high efficiency from the use of neodynium magnets.

[josh358]

I'm curious about the magnets as well, but I'd like to read more than speculation. Even the comments based upon the outward appearance are nothing more than speculation. Although measuring excised fragments of the strips is the only way to put a number on the field strength of the things, one could address the question of whether Magnepan actually employs differing materials on a single pole piece by measuring the interaction distance of various samples and a piece of iron (we all have rulers). If the piece of iron reacts to the presence of some strip samples at a greater distance than others, I'd say the patent I referenced describes their products accurately. If they all interact at (approximately) the same distance, then I'd accept that all of the strips are comprised of the same material – at least, for that particular model. I'd still warn against extrapolating from the sacrificed model to all others.

Only the folks at Magnepan know for sure, and I see no reason why they should give us an honest answer to such an inquiry. Aren't manufacturers entitled to protect trade secrets? Still, the simple, if inelegant, experiment outlined above would be great start. So far, I'd categorize everything I've read (and written ) on this matter as pure speculation.

---

By the way, the patent I referenced was from the 80s. I'll have to do some digging again - thought I saved copies of those things, but I've changed computers since I looked into this last time. Anyway, I recall three separate patents awarded to Magnepan which referenced the magnet strips. In the later patents, increased field strength (over the original strips) was a key point.

They've used different magnetic materials over the years -- the manual of my IVa's says that they have new and stronger woofer and midrange magnets. Also, they currently use different materials in different applications (including neodynium).

[jhm731]

I suspect a dual magnet 3.7 would be significantly more expensive than the single-ended version.

 

[SteveFord]

I hate to be a wet blanket but...
In the past I've talked to Wendell about the various suggestions made on the forums and sure, they could do it.
Everything that people suggest they've already tried.  Don't forget they invented the product and are pretty active behind the scenes.
The entire product line didn't get revamped by itself.  Except for the MG12 - slackers!!!

The question is just how expensive would you like the product to be?     

[medium jim]

I was trying to find the interview that Wendell did several years back where he indicated that they do read many of the threads about modding their speakers and have tried just about every one of them at one time or another.  Maybe Josh knows the interview that I'm referring to.

Jim

[jhm731]

I hate to be a wet blanket but...
In the past I've talked to Wendell about the various suggestions made on the forums and sure, they could do it.
Everything that people suggest they've already tried.  Don't forget they invented the product and are pretty active behind the scenes.
The entire product line didn't get revamped by itself.  Except for the MG12 - slackers!!!

The question is just how expensive would you like the product to be?   

I'd pay $8-9K for high performance dual magnet version of the 3.7s.

At $14K, I think the Sanders Model 10 with its bass amp and Electronic Crossover represents a better value.

[*Scotty*]

jhm731, what are you using for loudspeakers now and what do you think a double magnet version of a MG3.7 would offer.
By the by, your system links are not functional.
Scotty

[SteveFord]

Those Sanders look pretty nice - is that what you have?
We do like pictures...

[Davey]

I'd pay $8-9K for high performance dual magnet version of the 3.7s.

At $14K, I think the Sanders Model 10 with its bass amp and Electronic Crossover represents a better value.

You could build one of those yourself.  Buy two sets of used 3.6's and take the main panels from each and mount them back-to-back with the transducers facing each other.  Attach the tweeter from one set and you've got yourself a symmetrical 3.X speaker.  You'd also have a spare set of tweeters you could sell. 

I've done this with two pairs of MMG's and it worked fine in that application.

Cheers,

Dave.

[Davey]

I wish I could find it, but somewhere on one of the forums years ago(it may have been diyaudio.com) Dan Wiggins of Adire Audio and XBL2 fame wrote a series of posts describing the various characteristics of common magnetic materials.  Their history, strengths and weaknesses were all covered very well as related to loudspeaker design.  There were also more esoteric posts about the sound of each material. 

Greg

I remember those Dan Wiggins postings also.  Although I think referring to the "sound" posts as being "esoteric" is fairly kind.  It was flaming subjectivism with nothing but speculation forming the conclusions.

I'm still not following the magnet "sound" argument..and why some are preferable sonically and some not.  To me, that's kind of like saying you don't like the performance of a Ferrari because it's red.    It's meaningless without a larger context.

Anyways,

Dave.

[medium jim]

I remember those Dan Wiggins postings also.  Although I think referring to the "sound" posts as being "esoteric" is fairly kind.  It was flaming subjectivism with nothing but speculation forming the conclusions.

I'm still not following the magnet "sound" argument..and why some are preferable sonically and some not.  To me, that's kind of like saying you don't like the performance of a Ferrari because it's red.    It's meaningless without a larger context.

Anyways,

Dave.

Davey:

Stop and think for a second, Alnico is Aluminum, Nickel & Cobalt, completely different than Ferrite (Ceramic) and to that of Neodymium...each will have their own sonic signature.   This is self evident in Guitar Pickups which primarily use either Alnico or Ceramic magnets.   But I fear that all of this Magnet speak is going to derail the thread....

Jim

[SteveFord]

That's a thought with the 3.6s.
Somebody do that and report back, please.

[Davey]

Jim,

You're missing my point completely.  I'm not making it well I guess.
Anyways, you're right.....off-topic for this thread, and certainly not important in the larger scheme of things.  Magnepan uses what they use and that's the way it is.

Steve,

It's a simple concept, but a person needs to have sufficient DIY spirit to venture forth. 

Cheers,

Dave.

[josh358]

I was trying to find the interview that Wendell did several years back where he indicated that they do read many of the threads about modding their speakers and have tried just about every one of them at one time or another.  Maybe Josh knows the interview that I'm referring to.

Jim

I could be remembering wrong, but wasn't it in the open letter that he wrote?

As Wendell likes to point out, they're tinkerers, just like us. And when you've been designing and making planars for 40 years, you have a lot of chances to experiment. They're forever trying things, for example, Wendell tried stacking MMG's and said they sounded great. Some of these experiments we see as products or uses for products, e.g., Wendell decided one day to see what would happen if he used the DWM with the MMG for bass and was astounded to find that the bass extended lower than either speaker did alone. So the MMG/DWM system was born.

[jimdgoulding]

Not exactly how I pictured the listening room of reviewer for a major audio magazine.

Yeah, that sure doesn't look like Robert Harley's room.  REG is a math professor at UCLA.  Maybe that's his work crib and not his main residence.  Sorta looks like a 1 bdrm apartment to me.

[josh358]

I'd pay $8-9K for high performance dual magnet version of the 3.7s.

At $14K, I think the Sanders Model 10 with its bass amp and Electronic Crossover represents a better value.

I think they face some constraints that limit the number of models they can offer. For one, limited dealer floor space. They eliminated the 2.x because it was too similar to the 3.x and I know they don't sell that many 12's in the US, since it's so close to the MMG's and 1.7's. They keep it in production because it's popular in Europe.