[captainjack115]

What do you think it would sound like if i cut out a section and suspend it with rubber bands?

I bought some of those 55 cent onkyo tweeters and they dont sound that bad.



Wich do you think would sound the best?

[/b]

I'm not sure if the intention is to use a direct radiating tweeter along with an exciter.

Personally, I'm a bit of a purist and want to keep this project as NXT technology.
The Teac system that Zygadr remarked on appears to be totally NXT technology.  I think what they're doing is what I'd like to try.

I saw the "YouTube" video with the 12" panels. There may be some extra highs on the back side of that configuration, but he also loads down the panels by mounting everything by the back of the panels. That's a lot of mass to handle. No hope for bass there, he used a separate low frequency source. He also used a cheap driver with no crossover along with his panels. That set up can't possibly be anything close to high end performance.

I'm not trying to be an audiophile snob here, I'm just trying to get the very best results from a very promising technology.

I don't mind using multiple exciters, but I do think they should all run full range. In fact, I thinks it's imperative to get higher spl's and deeper bass by using multiple exciters. I just don't want to use extra direct radiating drivers and crossovers. The Teac units appear to be a way of extending highs and maintaining true bending wave operation.

This will be my next move in my empirical approach to a successful system.

Jack 

[BowerR64]

I just want a decent midrange sound in as small of a profile as possible.

THe sub can be hidden so thats taken care of, the transducers i have mounted on the panels im using just seem to lack a little clarity thats why i wanted the tweeter.

I dont want to mount a bookshelf speaker on my wall or i would of done that but i dont mind somthing thin like a panel. Ive tried many cubes and they dont sound good to me either.

[BowerR64]

I think the speakers they used for these transducers just didnt have a good sound from the start.

I think there is more to it then just the cone material they use for higher end speakers. If the drver isnt that good the best cone in the world isnt going to make it sound good.

[Donka]

The number of exciters seems to make a difference in the HF response. With one exciter my 2'x4' gatorfoam panel goes up to 15Khz with 2 exciters the HF rolls off at 12Khz and with 3 or more it is lower still around 11 Khz. I presume the multiple exciters must be canceling out the high frequencies. I wonder what role spacing between exciters plays.

(I tried different spacing and it did not seem to make any difference)

After reading some of the earlier discussion about dipole vs. bipole, I thought I'd try something very different - exciters on both faces of the panel. Complete Fail! The sound was just awful. The exciters must be canceling out.

Well it looks like people decided to breeze over your testing and findings.  I am happy to read some logical testing, and your results are exactly what ELAC states; which is the dB coupling when using multiple exciters is less in the HF range than the LF.  Basically each additional exciter is adding more LF than HF, which ultimately equates to increasingly lower HF response.


Sedge,

I don't see how you could have a panel material that is too light and rigid.  Based on ELAC's optimal panel material (paper and resin honeycomb), the SG is half of Gatorfoam.  The question is whether it really is as rigid at equal thickness and panel size.  However, if it is, you are in a good position as you can apply dampening (as recommended by NXT) to the panel to bring it into compliance, where one cannot remove the over dampening of the Gatorfoam core.



Regarding cutting holes for the exciters,

I was thinking about something last week where I would drill a hole through the board and glue a thin wall plastic/alu ring inside the hole, coupling the front and back sheets, with the exciter glued to the ring.  However I am afraid if the exposed inside back of the exciter acts as a DR tweeter, that it would add beaming and directionality.  I would think minimally, if one was to cut a small hole smaller than the foot, that you should edge harden the core with some resin and even some thin paper edging strips to stop the board from delam tendencies and also to assist in transfer to the front sheet. 


Last, did anyone have a chance to try white vs. black Gatorboard yet??

[BowerR64]

How are you running multiple exciters on one panel? series or paralell.

If you are running them in series i can see why its rolling off the high freq because your adding another coil. The coils act like a choke i think and the tend to block the HF the larger the coil is.

Im only running one exciter but im running 2 of the cheaper ones because they are 4ohms but i dont think they even sound as good as the dayton ones.

[usp1]

To keep things the same I am connected two in series and just attached one or two to the board. When I wanted to try three or 4 I connected two each in series and the pair in parallel.

[captainjack115]

How are you running multiple exciters on one panel? series or paralell.

If you are running them in series i can see why its rolling off the high freq because your adding another coil. The coils act like a choke i think and the tend to block the HF the larger the coil is.

Im only running one exciter but im running 2 of the cheaper ones because they are 4ohms but i dont think they even sound as good as the dayton ones.

[/b]

Hmmm!
 I'm not sure how exciters will behave.

A 4 ohm direct radiator will be around .04 mh near 16000 hertz.

Jack

[Donka]

BoweR,

The reduction in HF gain over LF by adding additional exciters, as described by ELAC, is due to negative cancellation effects.  Simply stated, yet very involved in terms understanding DML and material reactions.

Since many seem to be ignoring your questions, probably since your goal is different from others, I will try to give a few tips.   If your goal is to make a small thin panel this is what I know from reading up on DML et al.  1st Scrap the tweeter, the use of sheetmetal will transfer the HF and is no longer needed.  Now you have no need for multiple configurations, so cut the panel material you decide to use to whatever size you want.  Next glue the exciter off-center, hang it, or rest it on foam.  Add a HP filter to correspond with your subwoofer XO point.  Enjoy.   

Buying thicker foam is going to achieve squat, read the principle's of DML.  You'll either need a heavy frame that is isolated from the panel, or need to use some sort of foam/insulation between the panel/frame and the wall, otherwise it will vibrate and rattle.

[usp1]

zygadr,

As I mentioned in my very early posts, before I got my white gatorfoam, I was using corrugated cardboard panels. I was using the tri-fold science displays and the side panels give it rigidity without overconstraining the panel. The sound is clear and loud (louder than my gatorfoam panels). My wife actually preferred the sound of the cardboard panels!

Maybe this weekend I will try to recreate and measure the cardboard panels.

[fivestring]

I must admit I have been following this thread with great interest for a while now. I got a chance to get a few exciters from a local dealer (Hi Bostjan!) here in Slovenia and was able to make some measurements with the materials that I had on hand. One material that I have tried is called ultrapas here in my country, I think it`s some sort of artificial veneer of some 1.5mm thickness and is usually used to cover cheap kitchen elements or furniture - now you tell what the English word for it is. I also tried a 4mm thick real veneer of some 80x50cm, but it performed poor, sonically of course.
Contrary to all expectations, MDF panels of 3mm thicknes proved to be quite interesting in both, mesurement and listening tests, but I don`t claim it`s the "right" material.
In the past I have tried countless dynamic drivers, planar magnetic, ribbon, horns and even a plasma tweeter, various box or boxless working principles, different crosover techniques, etc.
In short, even with suboptimal panel material like MDF there is a great potential in this approach at least to my ears. There is a certain "rightness" about the sound that is hard to describe, but is quite obvious when compared to ordinary boxed cones and tweeters which I`m not very fond of. By listening experience I came to appreciate the "panel sound": big planar magnetic, ESL and ribbons are my favourite loudspeaker types and NXT/DML seems to fit in here as well.
I have reported the results of my experiments at our slovenian diy forum, but here are some of my findings for all you english blabbing folks 


The following panels were used (MDF), all 3mm thickness:

- 60x17cm
- 70x28cm
- 60x40cm
- 60x40 with two Dayton exciters

Some quick conclusions:

small panel dimensions are limited not only in their low frequency extension, but in the upper regions too. All the magic of bigger panels
is gone here. One exciter on a 60x60cm or 70x70cm panel already gives almost full range sound, but it is limited in terms of maximum SPL, of course. A 60x40cm panel could be a nice mid high range transducer if properly highpassed at cca 300-400Hz.
Two exciters mounted on the same panel with the recommended 8" spacing are quite destructively interferring with each other, at least at short listening distances. They do have a great polar characteristic in the upper octaves.
Observe the impedance peaks in the bass, large panels bring lower resonances and better low frequency extension but at the expense of lower sensitivity in the bass region.
Except where noted all measurements were done with the panels suspended in free air, I hold them with one arm

http://www.audiodiy.org/viewtopic.php?f=9&t=251&start=20


MDF 58x60x3mm 5cm mic distance

MDF 58x60x3mm 50cm mic distance

MDF 70x70cm on ground 10cm mic.PNG

MDF 70x70cm on foam 10cm mic.PNG

MDF 70x70cm suspended in air 10cm mic

ultrapas 40x60cm 50cm mic distance

stiropor 30x30cm 50cm mic distance

MDF 60x17cm

MDF 70x28cm

MDF 60x40cm

MDF 60x40cm anechoic

MDF 60x40cm f 50cm mic distance an 45dg

MDF 60x40cm f 10cm mic distance an 2x Dayton

MDF 60x40cm f 50cm mic distance an 2x Dayton

MDF 60x40cm f 100cm mic distance an 2x Dayton.PNG

[BrunoB]

I must admit I have been following this thread with great interest for a while now. I got a chance to get a few exciters from a local dealer ...

Nice measurements Fivestring.  I wonder if you could measure a waterfall plot as well? It is easy to do with SoundEasy. I did such a measurement with the mic placed  a few cm away from a piece of glass having one Dayton exciter glued to it.  The waterfall plot was awful: the glass was ringing forever. May be the results would be better with a 3 mm thick MDF?


Bruno

[scorpion]

Hi Miro,

Interesting results ! I think Ultrapas would translate to Melamine, at least some googling suggest so.
I noticed on your Slovenian froum that some may have tried KAPA-board, are there any results regarding these panels ?

/Erling

[el`Ol]

The mystery of NXT is not in measurements..........it's in practice with actual REAL materials and experimentation............lots of it as most here already know.

Mounting the exciter and listening is by far the best method of judging whether a material is suited for DML or not. But if it qualifies, how to proceed then?
A suggestion:
The Arta package has a realtime impedance measurement working with a noise generator (I don't know what others have it as well). One could press the exciter to the board without glue and move it to see the modes on the impedance curve.

Concerning my own experiments,
balsa plywood with hardwax oil sounds overdamped, maybe I'll coat carbon fiber next.

[usp1]

I tried a 1/4" 3-ply baltic birch panel 18x24 and it sounded terrible. More highs but very harsh and overall sound quality was pretty bad. Probably not a very good material.

[fivestring]

Thanks zygadir, for your warm welcome, I can assure you MDF is probably not even in the middle of the all time worst panel materials for the NXT principle , let alone the worst. There are tons of other candidates I am sure.
That reminds me of the foolrange mania, the lowther fans, which are constantly after that "certain something" that those drivers supposedly should poses, especially in the midrange, with absolutely no attention given to the fact that their entire presence and high frequency region is completely unlistenable and there is no bass to speak of...but the mids...they are to die for 
I mean, how can a loudspeaker that transforms a classical guitar into steel stringed one, be considered natural sounding and lively?
My point is, measurements by all means cannot replace our hearing, but not knowing how to measure or at least how to approach measuring this specific type of transducers (NXT/DML) is a bad excuse for not using them. Panels that measure horribly but sound incredibly sweet and tonally correct...are either nonexistent or probably useful only for some narrow spectrum.
Personally I don`t want to spend countless hours listening to panels that have "magic" midrange but completely lack overtones or harmonics which give the instruments their characteristic tonal color and overall sound signature, it is one thing to expect a completely ruler flat frequency reponse or... a reasonably correct one. Of course, there are too many ruler flat loudspeakers that sound totally lifeless and uninvolving, but the opposite is also true, there are too many mystery driven fullrange cone loudspeakers that sound acceptable only with small number of instruments, if at all.
Personally I don`t care for the ultralow bass or at least for the bass below 80Hz or so and likewise, I don`t care for the ultra high spectrum or at least for the spectrum above 14-15 kHz, but I do aim for a reasonably correct spectrum in between those two extremes.
To achieve a natural tone color along with a healthy dose of the original liveliness and dynamics... I hope I`m not asking too much.

I intend to try as many different panel materials as possible and put them all under listening as well as measuring tests, just to try to find some correlation between ear and the mic, if there is any but I`m sure there is.

Using ears and microphones in a balanced manner should only shorten the path and bring benefits, but I do admit that if I would have to choose between the two I would pick listening only without hesitation but fortunately, I can have both.

BrunoB,
I`ll try to supply waterfalls tommorow.

Hi Erling,
someone at our forum has tried Kapa board, our forum has a meeting this Sunday, I will report when I hear them.

Best regards,
miro

[j gale]

Zygadr,   I'm following your speculation re: cardboard as a panel. I wonder  since violins etc are lacquered if cardboard could be with the same good result? At any rate I guess it would need a coating that would stiffen it without adding too much weight. There should be cardboard in sheets that are somewhat ridgid vs the stuff that comes off a roll. Just wondering what might be available.

[captainjack115]

So, I suggest that the next time you play around with useless materials such as MDF and hold them with one hand while measuring their ''horrible'' frequency response curves, and posting them on this thread, keep in mind these variables(and there are probably even more that we are not aware of) and then try and convince us all that they are accurate and reveal a midrange only panel that ''lacks overtones or ''harmonics'' or '' sound acceptable only with a small number of instruments, if at all'

[/b]

One other thing that Fivestring has done is to make his measurements in near field, "very" near field at that.

Bending wave transducers are the most "room friendly" devices I've ever come across. As I've said before: "You can't listen to a spec sheet"!
I use a spectrum analyzer and a pink noise generator, but only for general reference.

BUT!!!!! Here's the really cool part that I cannot duplicate with my direct radiator speakers. When I run pink noise and take my spectrum analyzer microphone back to my listening location, I can move all it around in "any" direction, including backwards!
I can hold the microphone up down, side to side, walk to the left or the right.
The result? Nothing changes to any serious degree, I see no combing effects and the configurations of the various frequencies and their levels remain quite constant.

Now for me, that's  visual proof of the magic! I can see what my ears have been telling me all along.

When Paul Klipsch made his first horn system system, he admitted that it sounded bad. By tweaking he finally achieved what he was looking for and he did it with his ears. Yes he employed mathematical calculations and the latest test equipment of the day. Loudspeakers are made to satisfy human ears, not a rack of test gear. Mr. Klipsch used his ears as the final factor for success.

A quick and paraphrased story about Thomas Edison and his light bulb:

One of Edison's workers came to him and remarked that their endeavors were unsuccessful. Edison picked up a journal and showed the man page after page of materials they tried for a long burning filament.
The man didn't understand why Edison was pointing things that didn't work.
The point of it all was that they had a journal full of things they knew wouldn't work and there was no need to try it again.

Well folks, that's where we are with NXT technology. We know what doesn't work, but we also know what works better than other things.

We know we need a material that's as light as possible and at the same time as rigid as possible. We also know that we need to couple the back surface to the front with as little loss as possible. "Enter Nomex"!!! Yes, expensive stuff and probably not worth the money for something that may not be better than what we've tried thus far.

I stand with Zygadr on empirical design, we need to find the best techniques and and materials by listening with our ears.

Jack

[fivestring]

zygadir, actually there are more points on this technology that we agree on than you`re willing to admit, just re-read my posts.
I never said MDF is a perfect material, I just said even with suboptimal panel material like MDF one could hear the potential of this technology - what is wrong with that? Instead of  `suboptimal` I should say `worst`, I know.
The fact MDF is suprisingly linear and reaches very high, doesn`t mean much of course and I do realize there must be better materials, no doubt about it.
I`m sorry for making you feel I would like to turn this discussion into technical prattle because I`m not.
However, this thread might be "yours", but you certainly don`t own this forum as well as you don`t own the NXT technology.
I regret if mentioning the word "measurements" makes you nervous to the point you completely overlook that I don`t neglect listening tests (quite contrary) and I`m also trying to find a panel material that will sound true to the original as much as possible, but forgive me if I don`t get excited only by your claims, they say one picture tells more than a thousand words and here I`m not alone, some other people from this forum would also like to trace down all the causes that lead to a great subjective performance.
I wasn`t the first that brought up the subject of measurements, but if you feel they pollute your thread, bring more confusion than benefit and should be forbidden, then I apologize. You started this thread, you`re the boss, no need to feel threatened, I just thought we`re allowed to express our opinions.
So many loudspeaker technologies were claimed perfect, yet they`re all far from that and some are downright silly (widerange cones for example).
Personally, I feel large planar/ESL/ribbon loudspeakers are still the closest to the live unamplified sound and NXT/DML might closely follow them, but still has a long way to prove itself as an equal or better alternative. There is no doubt NXT/DML is however far more potential than cones`n`domes for example.
I started experimenting with exciters because I have heard Podium Sound speakers two years ago at the Munich audio fair, heard the potential and because people constantly reported on their capabilities at various forums all over the world.
I will continue to do so, with listening and also with measurements because I believe in the middle path...now you can expel me from your church...

Bruno, here are the waterfalls (God forbid!), panels look bad, but they do sound better than dome tweeter...for whatever reason

[BrunoB]

Bruno, here are the waterfalls (God forbid!), panels look bad, but they do sound better than dome tweeter...for whatever reason

Fivestring, thanks for the WF plots. They look much better then the WF plots I measured on glass. Glass was not sounding good at all: the music sounded like ... glass. I would predict that the waterfall plot of a good sounding material such as GatorFoam would be better.

Bruno

[captainjack115]

Well said Jack!

I'm glad you demonstrated the ''other side of measurement techniques'' ........just goes to show what we're dealing with here

Rob.

[/b]

Hey Rob:

One little amusing tid bit in my last post about moving the microphone all around.

Are you ready for this?

The readings on the spectrum analyzer were "not" flat, in fact no where near flat. But, my ears were loving it!  aa

Jack