Custom made horns and waveguides

[jzagaja]

Guys, you can also consider ellipsoidal OS waveguide. It has lower axial dip (mouth reflection). Recent customers telling me above presented horn is better than Stereo-Lab Tractrix (stereophony) however not CD.


List of throat (full) angles for particular drivers:

18sound:
NSD1095N, ND1090 = 27°
NSD1480N = 10°


B&C

1"
DE250 14.6°
DE10 7.7°
DE12 24°
DE400TN 20.7°
DE400 31°
DE500 17°
DE200 9.9°

2"
DE85TN 34.5°
DE750TN 22°
DE950TN 17°

BEYMA:

CP750Nd – 24º
CP850Nd – 7º
CP755Nd – 12º60’
SMC65Nd – 15º
CP385Nd – 16º30’
SMC225Nd – 13º1’


Which is nicer?

http://img340.imageshack.us/i/leclach400hz1calcopy.gif/

[Frank]

I like the smaller one.

[jzagaja]

And what is important KIT KINOTECHNIK Austria proved that Gaussian optimization of LeCleach profile leads to constant directivity horn:

http://img230.imageshack.us/i/dihogaussexaaple.png/

[MethMan]

But I see no prove about CD. What's on the picture?

[jzagaja]

Parallel curves are not a prove of CD horn?

[MethMan]

What's on the picture? Measurement? Full hemisphere is not available?
What is gaussian optimization of LeCleach horn leading to CD? What is optimized? Horn or source?

[jzagaja]

On the picture there's BEM simulation up to 30deg off axis. As far as I know contour is optimized - will ask Michael from Germany. Source is always flat piston, usually 1".

Michael writes he tries bend LeCleach toward min-phase behavior (constant directivity). If you make the same simulation for OS profile then results is worse so there's a room for an improvement:
http://img28.imageshack.us/i/osexample.png/

Example of CD waveguides designed by Mr. Feredikis who works on Balanced Mode Radiators known from Naim Ovator 600.
http://img207.imageshack.us/i/edc82bda50.jpg/

[MaBat]

I've made an utility for OS WG calculation for Windows command line.

Usage:
oswg.exe <throat_diameter[cm]> <full_coverage[deg]> <transition_diameter[cm]> <roundover_radius[cm]> resolution[cm]> [<front_length[cm]> [<mount_length[cm]>]]

transition_diameter - a diameter at which OS profile is beginning to be curved to baffle mounting.
roundover_radius - a minimum radius for the above transition
front/mount_length (optionals) - lenghts of straight contours to be added to the throat/mouth

It outputs the profile coordinates to the file "oswg_profile.txt" in mm, (can be easily imported to Excel), also prints the wall angles for some common (desired) throat diameters.
Here is the resulting text file for the command line parameters 2.519 90 29 5 0.05 1 1:



In a case of the true OS WG (which I hope) we could agree on some numbers / make a finall decision...


PS. There can be bugs in the code - maybe for some higher resolutions, I don't know (didn't try).

[jzagaja]

One of the user of above horn writes it is much better than Stereo-Lab 400Hz Tractrix. Main reason is stereophony. His temporal setup shown here:

http://img97.imageshack.us/i/lecleachownerssystem.jpg/

[MethMan]

Michael writes he tries bend LeCleach toward min-phase behavior (constant directivity). If you make the same simulation for OS profile then results is worse so there's a room for an improvement:

http://img28.imageshack.us/i/osexample.png/

No, there is no room for real-world improvement unless you improve the source. No simulation, no hype, no fancy super-balanced-holy-whatever can change it, nor impress others. It is proved by math and work as intended in reality. It is not the best imaginable, but OS is the state-of-art in CD and is really existing. If somebody does not feel comfortable with it and like something else ok, but it doesn't prove anything still.

[jzagaja]

Nevertheless here's an example of min-phase LeCleach for 3" flat drivers:

http://img70.imageshack.us/i/gauss1503.png/

[MaBat]

Guys, let's leave this topic to Jean-Michel and Earl (we all know the story and background) and concentrate on our new OS WG countour.

- I, so as MethMan, really do believe in OS (and paramount importance of CD), there is no chance to convert anyone, I'm sorry. We want just that

[jzagaja]

If you want baffle mountable version (<90deg) then it should be cheaper. What compression drivers you would like to use?

[MaBat]

"If you want baffle mountable version (<90deg) then it should be cheaper"

That's what I wished to hear That's great, really.

DE250, probably? (why to complicate things...)
I don't know yet. Sure, we should agree on that - because of the throat.

[jzagaja]

Newer drivers for waveguides shift toward wider angles (20-30deg). Best would be without this conical part (0deg)?

[MaBat]

Well, that part is just for a technological purpose (if you mean the throat) - by cutting it away it will be possible to attach the driver more precisely (we hope). The initial throat diameter is made a little bit smaller for this purpose and the WG has the right wall angle at 1".

Question: How precise will the resulting shape be? In other words, what resolution for the profile coordinates makes even sense?

[jzagaja]

No problem with grinding throat to desired diameter but I can't guarantee throat angle. How could we measure such thing without electronic probe? With a bar gauge?

[MaBat]

No, we would do that. You would just be provided with the coordinates of the whole thing (i.e. that's what you've offered in post #9). No other cutting or measuring work for you...

My question was this: When I finally send you the coordinates, what resolution will be sufficient for maximum quality obtainable by your technology (say, how many decimal places of numbers in milimeters)? (Just informative - I don't have an idea.)

[jzagaja]

I think for CNC machine <=0,1mm is ok.

[MaBat]

Thanks, that's exactly what I wanted to know.

Bye for now and thank you again for your patience.