Yes its a easy test, and I was explaining why its harder to hear the difference on a DML panel because high frequencies are spread evenly throughout the panel unlike a conventional cone driver where the HF are pronounced in the center due to the diaphragms cone shape. I was just explaining the reasons for your findings.
I don't think thats what Kef intended. My theory/opinion on that particular Kef model design is that they were not trying to reproduce DML bending wave panels. Instead they are reproducing a form of BMR technology which is very similar to the Sony APM line of speakers.
I agree that Bertagni's design compared to NXT's design are totally different. I will go out on a limb and even say that Bertagni's design is more advanced then NXT, which further proves that measuring tools don't always mean shitz. LOL Like I said before all the measuring tools in the world wont tell one how to make there DML panels sound better which is the reason why threads with a lot of measurements usually lead nowhere. What will make ones panels sound better is the use of certain techniques that other famous designers utilized in there own designs that have proven to WORK.
Out of all the designs out there NXT is my least favorite. First the reason for NXT demise is they didn't market there products to the average audiophile consumer instead they only marketed to Pro, live pa audio. I doubt there main concern was hi fidelity. It was more voice intelligibility and the ability to play loud as they were used for live bands and huge concert halls/auditoriums etc. Bertagni speakers were marketed along side other famous brands of loudspeakers to the average audiophile consumer.
Yes I've stated many times that Bertagni's design is similar to BMR design. The only difference with BMR tech in the mechanical aspect is that BMR's do usually have higher pistonic excursion then most exciters. BMR's voice coil design is thinner and lighter then most exciters plastic voice coil because it was not designed to support its panels weight and therefore can use less power to reach maximum pistonic excursions. Certain exciters like the DAEX30HESF-4 have very similar pistonic exursions but will require more power to reach those exursions due to a heavier plastic voice coil.
Also the reason why DIY DML panels are limited is due to the limited types of exciters that are available. Yes different types of exciters do make a difference. Most exciters of today were design to be novelty experiments to be used as cheap gimmicks to hide behind walls for surround sound or used in places where no other speaker could reach. Bertagni custom made his own exciters. If I could custom make my own exciters I would make each exciter frequency specific meaning there will be a exciter that is meant to be used for the HF a specific exciter for the mids and a specific exciter for the bass duty all individual custom made to be ideal for each frequency spectrum.
I am not here just for the sake of arguing. If I am wrong I will admit I am wrong as I am not afraid to do so. I just want to set people on the right path to better ones DML panels. I wont give out all my secret techniques but I will give out some tips and hints
It didn't take long for this great question to get buried! The short answer is that F0 varies inversely with panel area. That is, if you double panel area, F0 is cut in half. The long answer follows:
Yes, there is an equation: The fundamental frequency of a panel can be estimated from:
F0=(pi/A)*(D/u)^1/2
here, A is the panel area, D reflects panel stiffness and u reflects panel weight roughly as follows:
D=(E*h^3)/12 where E is the panel's elastic modulus and h is the panel thickness and
u=r*h where r is panel density and h is (again) panel thickness.
The definitions of D and u are more complicated if the panel is a composite, but the first equation applies even for composites.
I can help the calculations, if you want.
Be aware that F0 depends on how the panel is supported, but the equation ignores that effect.
That said, I've seen it stated (Azima I think) that the bottom end of a DML panel is at about 2.5*F0.
Eric
Hey folks, let's keep things civil and personal attacks out of these discussions. It's not fun to clean up after this stuff, and the mods would like to enjoy the forum too.
Don't worry I will help you out by not posting anymore.
P.S. anyone want my advice PM me.
Last edited:
Thank You so much Eric! Thats exactly what I was looking for...too much time wasted using google
Its interesting the Azima comment about the bottom end being about 2.5 * F0...an example I've been studying showed a small commercial dml panel with an F0 of 110 Hz. The manufacturer claimed a FR of 200-18K for the speaker but of course didn't specify the -3db points. So assuming that 200 Hz bottom end was probably a bit optimistic it would agree with the theory to some extent. And that helps me a great deal as well in understanding the size dimensions I need to be looking at for a specific low end.
Now your other statement about f0 depending on panel suspension. I assume you mean dampening of the panel (at least around the periphery.) i.e., going from an infinite plane structure to a finite one. Can you explain to me does such damping increase or decrease the F0 and are their equations that may describe the effect on F0 by dampening.
One last question, when you mentioned the varable D reflecting panel stiffness, I assume you ment the Bending/flexural strength (aka modulus of rupture)?
And thank you for the offer to help with the math, I'm comfortable with that. The hardest part is just trying to equalize all the values to be either metric or english and m/cm/mm/in/ft squared and cubed etc...thankfully google can be a great resource for the conversions to at least double check my hand and calculator answers
Its interesting the Azima comment about the bottom end being about 2.5 * F0...an example I've been studying showed a small commercial dml panel with an F0 of 110 Hz. The manufacturer claimed a FR of 200-18K for the speaker but of course didn't specify the -3db points. So assuming that 200 Hz bottom end was probably a bit optimistic it would agree with the theory to some extent. And that helps me a great deal as well in understanding the size dimensions I need to be looking at for a specific low end.
Now your other statement about f0 depending on panel suspension. I assume you mean dampening of the panel (at least around the periphery.) i.e., going from an infinite plane structure to a finite one. Can you explain to me does such damping increase or decrease the F0 and are their equations that may describe the effect on F0 by dampening.
One last question, when you mentioned the varable D reflecting panel stiffness, I assume you ment the Bending/flexural strength (aka modulus of rupture)?
And thank you for the offer to help with the math, I'm comfortable with that. The hardest part is just trying to equalize all the values to be either metric or english and m/cm/mm/in/ft squared and cubed etc...thankfully google can be a great resource for the conversions to at least double check my hand and calculator answers
Burnt - this is a great idea and hopefully you have some sort of document like that already. I had a look at the options on the diyAudio site, but they are not well suited. The Wiki section might be an option, but I get the feeling very few members actually go there or reference that. If you have such a document the simplest would be to post (with FAQ as heading maybe) in this thread with your document (standard MS Word format) attached. Xrk971 can them simply create a reference link to your post in the very first post of this DML thread. He is very good at maintaining his threads like that. I am sure such a document will be very helpful to newcomers and for future reference. Just a thought.
Great idea twocents and thank you XRK971 with the kind offer to link it to the front page. I will complete the doc as a pdf and post it to the thread. Should be done by tomorrow latest.
We can add in any further builds as they arrive.
Thanks for the help folks
Burnt
Thank You so much Eric! Thats exactly what I was looking for...too much time wasted using google
You're welcome.
Concerning the 2.5*F0 statement, I can't be sure of the original source. But you can see it mentioned here by Graham Bank at the bottom of page 198.
Loudspeaker and Headphone Handbook - Google Books
Oh, almost forgot: My source for the equation for estimation of F0 is this paper, page 2756:
(PDF) Development of panel loudspeaker system: Design, evaluation and enhancement
I'll answer you other questions in separate posts.
Eric.
Last edited:
Close, but not quite. D is most closely related to the Bending/flexural Modulus, also called MOE for Modulus of Elasticity. It is often measured in the same test as flexural strength or Modulus of Rupture (MOR), but they are very different things. So do not try to use the MOR number in the determination of D.
The number you want to start with is the MOE, which may also be referred to as Elastic Modulus, Young's Modulus, tensile modulus, or flexural modulus, and is almost always denoted by the letter E, This parameter reflects the resistance of a material to deformation at loads well below the load required to break it. Do not confuse this parameter with the MOR or flexural or tensile strength. Those numbers reflect the load required to break a material, are usually denoted by the Greek letter sigma, and have nothing to do with DML performance. It can be confusing because both MOE (what you want) and MOR (not what you want) have the same units (psi or Pa or N/m2), but are very much not interchangeable.
Okay, back to the bending stiffness of the panel "D". First, be aware that "D" may also often appear as "B" just to confuse us. For example, the chapter by G. Bank I referenced earlier uses the "B" notation, while the Bai article uses "D".
So the elastic modulus E is a material property, not a panel property. That is, steel has an elastic modulus and it's modulus is always the same no matter what shape or size or thickness the panel or whatever the steel is made into. But the stiffness (D) of a panel of steel varies as a function of it's thickness, so you calculate D from E as:
D=(E*h^3)/12 where h is the panel thickness.
Note also that this is actually an approximation. To be more accurate you can put the factor (1-u^2) in the denominator, where u is the "poisson's ratio" for the material. But in practice, the number (1-u^2) is virtually always between 0.9 and 1.0, so it makes little difference whether or not you include it.
Eric
Again, Thank You very much for the references, I already had those docs buried among a few hundred others I've read over the years...read too much but don't retain enough sigh.
And definitely a big TY for the engineering explanations...yes I've noticed the D and E variables being commonly used in research papers, though often with the different names associated to them. For us non-engineers, the terminology can be just so confusing when different words are used to mean the same thing and the normal inexact and sometimes incorrect terminology us non-engineers use to try to describe the same things.
So the only other question I have is does dampening of a panel raise or lower its F0?
BYW, have you used many of these properties and formulas in helping you to design your panels? Or do you by chance calculate them after the fact to see how well your panels may compare to theory?
And definitely a big TY for the engineering explanations...yes I've noticed the D and E variables being commonly used in research papers, though often with the different names associated to them. For us non-engineers, the terminology can be just so confusing when different words are used to mean the same thing and the normal inexact and sometimes incorrect terminology us non-engineers use to try to describe the same things.
So the only other question I have is does dampening of a panel raise or lower its F0?
BYW, have you used many of these properties and formulas in helping you to design your panels? Or do you by chance calculate them after the fact to see how well your panels may compare to theory?
