
Home  Forums  Rules  Articles  The diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 
MultiWay Conventional loudspeakers with crossovers 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
4th December 2013, 07:18 PM  #6691 
diyAudio Member
Join Date: Dec 2004
Location: Novi, Michigan

First, you need to understand that refraction does not take place within the medium, but at the interface between two different mediums. One medium, in this case, is air and the other the plates. Since the plates require a longer path, the wave speed is slower where they are present. When the slower waves meet up with the airs normal velocity refraction occurs. It is not ideal, which is why it works so bad, but that's the idea.

4th December 2013, 07:42 PM  #6692 
diyAudio Member

Got it Earl. At the same time wouldn't the speed in the aluminum plates themselves be faster than the air? It would seem due to the extremely thin aluminum the actual emissivity at the end of the plates would be rather small compared to the air itself except perpendicular to the plate surface area. Refraction would appear to be very minimal while diffraction would seem to be the dominate factor here.

4th December 2013, 08:27 PM  #6693 
diyAudio Member
Join Date: Apr 2011

I doubt that many here were confused at all about it, but what fun is that? (:
__________________
My Signature is..... I don't even tolerate my own BS. 
4th December 2013, 09:19 PM  #6694  
diyAudio Member
Join Date: Dec 2004
Location: Novi, Michigan

Quote:


2nd May 2014, 09:33 AM  #6695 
diyAudio Member

Wow, long time no chat here!
I am using mabat's equation to calculate an OS profile, but now I would like to also calculate the area expansion (with the correct wavefront shape). Any clue?
__________________
No loudspeaker system even approaches real life so there is plenty of room for interpretation  Greg Timbers 
2nd May 2014, 01:58 PM  #6696 
diyAudio Member
Join Date: May 2004
Location: Toronto, ON

My best guess is to go to where the math came from. Oblate spheroidal coordinate transforms that allow nice solutions to PDEs. The transform should relate the waveform shape and expansion.

2nd May 2014, 02:05 PM  #6697 
diyAudio Member
Join Date: Dec 2004
Location: Novi, Michigan

Josh is correct, you would have to do this in the OS coordinate system by integration, just like you would do for spherical coordinates but substituting the definitions from OS. Not an easy task, but not impossible either.
My question would be what do you need the area expansion rate for? 
3rd May 2014, 06:58 PM  #6698 
diyAudio Member
Join Date: May 2004
Location: Pensacola, Florida

OS Note
Earl, this is for other readers.
To the first degree of approximation, a spherical surface assumption should be sufficiently accurate for calculating the area expansion of horns of circular section, as an OS horn of this variety is asymptotically conical. However, at the horn mouth, wave front geometry gets far more interesting and so does the horn geometry required to mitigate refraction and reflectance of the exiting wave front. Here the horn profile and the wave front as well, necessarily depart radically from the OS régime. Regards, WHG 
3rd May 2014, 09:11 PM  #6699 
diyAudio Member
Join Date: Dec 2004
Location: Novi, Michigan

The mouth is near conical, but the throat is far from that. If one wants the wave front area from the throat to the mouth then it could get complicated. It would be true that this wave front is always a spherical section, but near the throat the radius and angle subtended would vary continuously, starting at an infinite radius and zero subtended angle. Unlike at the throat, near the mouth the radius would vary directly as the distance along the device.

3rd May 2014, 09:57 PM  #6700  
diyAudio Member
Join Date: May 2004
Location: Pensacola, Florida

More...
Quote:
The ist. derivative of the horn profile curve (a hyperbola with apex at x=0, y=rt) gives you a tangent line. The intercept of that line with the x axis gives you dx for a given dy. Thus the radius (r) of the approximating spherical cap may be calculated thus: r = ((dx^2)+(dy^2))^(1/2). Of course as r > oo area of a flat disk is a sufficient approximation. The area difference between oblate spherical cap having an elliptical profile as opposed to spherical cap having an approximating circular profile is negligible for any circular section OS horn of typical size. Regards, Bill N.B. In my previous post the word refraction should read diffraction. At that time I was thinking about an acoustic lens. Last edited by whgeiger; 3rd May 2014 at 10:05 PM. 

Thread Tools  Search this Thread 


New To Site?  Need Help? 