Speaker ID

[mmoeller]

I've been looking to clone the Revel Performa F52. I have a pic of the drivers. I was wondering if anyone could identify.

I see the tweeter has a custom waveguide attachment. Not too worried about that, just what tweeter it may be.

Some or all of them may be exclusive OEM stuff for Revel, but there may be something similar.

Pic attached.

[kirknelson]

Chances are very high that they are all proprietary drivers. Revel is owned by the Harman Group so they have plenty of production available.

[mmoeller]

I just read an interview with one of the Revel Designers. Was bragging about how great it was to be able to have the resources to design his own drivers, and not have to use OEM stuff. I guess that awnsers my question as well.

So. Switching gears. Are there any 6.5" woofers out there that can extend into the 30's in a triple set up?

A matching 5.25" would also be nice.

What tweeter would be suitable for a custom waveguide attachment? I can get SLA's made for 1 off devices fairly cheap.

Looking at the Revel Center Channel uses the same mid 5.25, but then extends to 2 8" drivers? Nice to have a matched set there as well.

I guess its time for the Zaph Audio speaker performance search.

Thanks again.

[Jonasz]

I think you should look at the Seas 27TBFCG for tweeters and the L18 for woofers. I'm pretty sure there is a 5" Seas to complement the L18's. This could turn out to be a very nice speaker.

You could also go with Dayton RS7, RS5 and RS28 wich would in my eyes at least make a nicer looking speaker (and probably as nice sounding or better than the Seas alternative).

[cjd]

RS180 can be ported in the 30's with the right cabinet size. I port my RS150's at ~38Hz. And the RS150 is, IMHO, an excellent midrange option.

The RS180 has better bottom end IIRC (compared to the Seas options). Seas has better midrange and top end.

Tweeter I would grab a Seas 27TDFC or RS28A personally.

C

[mmoeller]

Thanks for the input guys. Would investigating the Peerless Nomex standard 5.25 and 6.5 drivers be worth while?

If I chose to add a waveguide would a particular tweeter be more suitable?

[technimac]

Try this....

http://www.zaphaudio.com/Waveguidetmm.html

[mmoeller]

I wanted to look in to a Elliptical Oblate Spheroidal waveguide. I saw Geddes has a book with a section on this type of structure. There is not a whole lot of free information about this available. JBL has some sort of IP regarding this as well. The trick here is being able to plot out a model for something complicated like this. I have yet to order the book, but it might be a good idea.

[Dennis H]

The Oblate Spheroid equation is posted here somewhere if you do a search. In practice, if you were going to carve one by hand, you'd take a block of material and draw your circle or ellipse or whatever on the front side. Carve straight from that line to a point (zero diameter) on the back side so you have a round or elliptical cone. Drill a hole through the back the diameter of the throat and then smoothly blend where the hole meets the cone. Finally, round the front of the cone to reduce diffraction (not included in the equation.)

[mmoeller]

I did search for that. There are some subtle differences between an Oblate Sphereoid and an Elliptical Oblate Sphereiod.

r = sqrt( a^2 + (x * tan(theta))^2)

The Oblate Sphereoid will have a cross-section of a circle at the throat and the "mouth", where the Elliptical Oblate will have a circle at the throat and an ellipse at the "mouth". I suppose your technique would work for both. I was hoping to be able to create a 3D model.

[Dennis H]

r = sqrt( a^2 + (x * tan(theta))^2)

Mmoeller, I've been meaning to get back to you on that. While not trivial, you can create a 3D model of an oblate spheroid horn with an elliptical mouth. The challenge is the value of tan(theta) changes as you go around the ellipse. So we need to define the ellipse in terms of angle of rotation, t, around the center rather than the more common way of using two foci. Wikipedia to the rescue. I'll stick with their conventions for symbols even though some of them are the same as the equation above. Just be aware that a, r and x mean different things for the two.

http://en.wikipedia.org/wiki/Ellipse

x = a*cos(t)
y = b*sin(t)
0 <= t <= 2pi (or 360 if your software uses degrees)

Once you have x and y, you can calculate r, the distance from the center of the ellipse to a point on the ellipse.

r = sqrt(x^2 + y^2)

Now, if we know the depth of the horn, d, we can calculate tan(theta).

tan(theta) = r/d

I typed this in a bit of a hurry so double check my math but I think it's right.

[mmoeller]

Very nice. Your math looks good.

I've been doing a little diggin of my own. I read Geddes patent relating to waveguides. He had a very interesting point of view. Creating the structure you have shown the math for here will give crossections that are all elliptical. He suggested using a Prolate Spheroid which will give a circular crossection at the throat. He then suggests ploting Elliptical Cylindrical coordinates to map the mouth section. The trick remain in blending the two structures to a smooth finish. There are some remarks conserning maximum angle to attemt dispersion.

I've been able to creat the two structures in a first run attempt to see the structure. I have attached a small pic. I've about got it. The smoothing however will require some fudge factor.

Wikipedia was a great source for the coordinate systems, (prolate spherodial, elliptic cylindrical). If you want to see the calcs PM your email and Ill send along something.

Thanks again for the assistance. ;x(

[Dennis H]

Creating the structure you have shown the math for here will give crossections that are all elliptical.

Nope, it will be round at the throat and elliptical at the mouth when you plug the various tan(theta) values into the oblate spheroid equation -- tan(theta) changes but throat diameter remains the same for each slice. Each slice is a radial one from the center out (like cutting up a pie), not a planar one from front to back.

[Dennis H]

Hmmm, I tried it in excel and I got some goofy plots -- okay near the throat (round) but some kind of silly figure-8 shape near the mouth. Gonna have to rest up and think about it some more. I know I'm doing something simple wrong but my poor tired brain can't find it right now.

[mmoeller]

Hehe... I spent a bunch of time trying to plot it out as well. There's something going on, but I'm not sure what it is. I don't know how many times I've tried to put it together. I know there are some discontiuities in the tan function, but I'm not convinced that is causing the problem. It could be that I'm using Mathcad and my arrays are screwed up.

[Dennis H]

Okay, I finally got it to work. Sweeping the OS equation around a circle doesn't work for some reason. But this does.

Specify the vertical and horizontal angles
Each slice is a plane parallel to the baffle
Use the OS equation to calculate the vertical r at the slice depth
Repeat to get the horizontal r at that depth (different theta)
Plug the r values into the ellipse equations -- horizontal r is "a" and vertical r is "b"
Build an ellipse at that depth -- it starts as a circle (a=b) and morphs into an ellipse

Excel doesn't do 3D graphics very well but the pic is a 100 degree by 50 degree waveguide with a 1" throat and a mouth approx. 10" x 4".

[mmoeller]

Good work. I tried some today to get it to work. I noticed that when calculating theta based on the x,y ellipse coordinates was some what flawed. The distance from the axis of rotation was the x,y ellipse coordinate, which is the same as the r from the OS equation. However, the OS equation takes into account that the value for "a (circle radius)" separate from the triangle used to find the opposite leg distance. Once I saw that I was able to subtract "a" from the ellipse coordinate magnitude, calulate theta and then the subsitution looked ok. My plots on the other hand didn't look nice, but the data seemed to mirror what your showing me here. I'd be interested in seeing your Excel sheet. I'll PM my email.

[Dennis H]

XLS sent.

[mmoeller]

Thanks, I received.
Mabey I can convert some of your work to a tool that can plot some nice 3D images.