Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
64.252.118.104
In Reply to: RE: Should we send a "representative" to Magnepan? posted by Satie on October 14, 2011 at 15:00:38
Didn't some OEM mfrs modify the BG line sources to do that? I'd expect IM to be a concern, but we aren't dealing with bass frequencies so its geometrically less of a problem than it would be if you tried doing that with a woofer, and, really, we already know that a single driver can satisfactorily cover a broad range -- e.g., the BG's -- so we'd only be adding in a single octave.
The Wisdom approach is another possibility, albeit a pricey one and one that sacrifices the advantages (and loses the disadvantages) of a ribbon and could have interference/power response problems around the crossover point. So is the Quad delay approach using multiple amplifiers, but that opens a new can of worms.
Follow Ups:
Actually, they got PDR drivers. Those are damped physically to prevent the outer portion of the diaphragm from emmitting the higher frequencies (they are produced but are absorbed). That reduces sensitivity and thermal dissipation and thus power handling but keeps the broad driver from canceling the high frequencies. Not really the same execution but gets results.
I was thinking of some companies that use modified RD drivers. I gather that they use the center trace(s) in much the same way that Magnepan uses the outermost traces on the 1.7 tweeter. It's presumably a more expensive solution than the PDR since it requires crossover components and extra wiring, but AFAIK it will produce better results, without the tradeoffs inherent in the PDR damping/tapering approach.
It would be great to have that work, but I don't think you can overcome the IMD that a shared diaphragm would produce. However, you can obtain a timing benefit since the origin or wavelaunch are identical in both ranges. That always seems to be a problem in manufacturing that a concentric line source sharing the diaphragm will solve. So I don't really know how far the tradeoff of timing and IMD would put you ahead.
Actually, I am sure that magnepan tried this out at some point. I am guessing that it was put away in favor of keeping the tuning buttons in their symmetrical positions and putting the HF traces at the edge.
This is like the ribbon Tympani, where putting the mid/tweeter panels in the middle ends up losing baffle gain in the midbass panel.
I know that they put the tweeters on the edge at least in part because the amplitude of the vibrations is lower there, it's in one of the patents. So less IM. But I'm not thinking in terms of bass, which causes IM problems even with the tweeters on the side, but rather combining mids and tweeters.
If IM were a problem, it could be minimized by using an overlapping frequency range approach. The center could even be clamped off to further reduce IM, as the midrange is in the better Maggies. Though not of course in a kapton-backed ribbon.
I guess you would lose some effective baffle size with the mid/tweet panel in the middle, wouldn't you. Never thought about that. I've assumed it wasn't done because of lobing and modulation of the ribbon tweeter.
Are you sure there would be a lobing problem from the center position? I see it with a concave setup but not if it is flat or slightly convex.
I'm thinking that the woofer panels would be too far apart, we're getting to about a wavelength separation a bit above the crossover point, at about 500 Hz. So you're going to get interference between the woofers which will cause lobing error. This is a known issue with D'Apollito-type configurations:http://www.birotechnology.com/articles/VSTWLA.html
Also, the midrange driver will be located asymmetrically wrt to the woofers. I assume you'd have everything equidistant from the listener, but I imagine that will make the lobing somewhat asymmetrical.
The question is, how audible would this be? It would really depend I think on the first and second backwave reflections more than anything else, and that would depend on placement and room geometry as well as speaker geometry and crossover slopes, plus the characteristics of the drivers themselves. So I'm not sure what it would do to the sound, I think you'd have to model it or try it to know. You might be able to tame the effect if necessary with diffusion at the reflection points.
Edits: 10/16/11
I have the XO at <300 hz so I calculated it as a marginal effect, besides, despite the lobing the d'appolito design works very well.
There should not be that big a deal on the mid's offset - and if it is, you can widen the placement to return it to symmetry.
There were a few flat and centered mods to Tympani posted here in the early 2000's and I got the impression that it worked fine without any problems that were not common to the stock speaker as well.
Well, the standard D'Apollito implementation does space the drivers as tightly as possible. They also usually use the 4th order L-R, which has a pretty broad central lobe. And the axial lobes are symmetrical and vertical.
Of course, you have lobing with the standard Tympani configuration, too, and it's going to have axis tilt. But didn't you say that you had trouble crossing the ribbon high up because of the separation between the ribbon and mid? With a crossover of 300 you're going to have significant output at frequencies at which separation is greater than wavelength.
Didn't know people had tried a centered configuration. My main question was always about low frequency acoustical modulation of the ribbon tweeter and whether it wouldn't cause Doppler distortion or even failure. I don't think you could just take the mid and tweeter, put them next to one another, and remove all the space, although it would be great if you could. But I'm just speculating here, there has to be a reason why they separate the ribbon tweeters as they do -- from the perspective of blending and crossover design, you'd want them to be as close together as possible.
Yes, the high freq is a problem because of the small wavelength at 9-10khz and the narrow beam from the mids at that area. the combo provides comb filtering as you move your head - true head in a vice. Only works at LR4 - but that kills transients and adds gain stages to the XO.
Essentially, the difference in ear to driver distances should not excede the XO wavelength so that a normal head movement of +-5" should result in a change of less than 1/2 wavelength and preferably less than 1/4 in the difference of ear to driver distance from one extreme to the other - in the tympani that would be at a 10 ft listening distance roughly 10" * 8"/120" = 0.7" which is 13500/(0.7*2) about 10khz for the 1/2 wavelength criterion - ended up too close for me, and 13500/(0.7*4) = 5khz for the 1/4 wave criterion - where you can not go from positive to negative reinforcement by moving your head in a 10" range - that did work in my setup.
Taking Fc on mid LP to 4-5 khz while keeping Fc for ribbon HP at 9+ khz - XO about 6khz solved the problem because head movements have much less of an effect with the much broader mid beam in that freq range to 6khz.
I think the holes in the Tympani around the tweeter are supposed to prevent both bass induced damage and reduce doppler effects since there should be cancellation at that point so less (if not insignificant) of a doppler effect from the bass wave.
Josh, get's my vote to rep for the forum. I would definitely give more weight to his opinion than professional reviewers. Who have various conflicts of interest.
I would like him to report on the 3.7's, the Tri-Center concept, the Mini Maggies, CCR and center channel woofer, and anything else Magnepan can come up with.
Josh, I am curious does Magnepan have any conventional speakers on site to compare against. If so what brands and models?
That's an interesting question.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: