In Reply to: horn signature - whats left after proper tailoring? posted by freddyi on January 23, 2006 at 06:56:08:
Hi FreddyiI can answer in a general way.
Directivity, here is a hard one to picture. Horns usually not only increase the radiation efficiency of the driver but above some frequency also controls the radiation angle.
The larger the dimensions, the lower the frequency is where directivity control is finally lost. The wider the horn wall angle for a given size mouth, the lower the frequency where pattern control is lost. The larger the horn mouth, the less energy is radiated say 90 degrees of axis even at 20KHz..
Exponential and other heavily curved wall horns can have better low frequency loading than conical horns but have a directivity pattern which progressively narrows with increasing frequency.
The reason can be seen by keeping in mind that as the frequency climbs, the region of the horn that is controlling directivity retreats down the horn towards the driver.
What is rarely appreciated here(except by perhaps Earl) is that at say 20KHz with a 2 inch exit, the working portion of the “horn†and to significant degree, the portion that governs directivity, is within the driver itself. An abrupt or “too large†transition to a larger angle than that internally, produces the HOM’s. To minimize these effects, one must keep the exit diameter (acoustic size) consistent with the horn wall angle its connected to.The compression driver connected to the horn can normally only go up to about 2 to 4 KHz before beginning its hf roll off, they are not “flat†like a dome tweeter.
With a curved wall horn, If everything is just right, then much of this roll off is off set at least on axis because of the narrowing of the radiation pattern. If the roll off and narrowing are perfectly complementary, the on axis response is flat.
I think this is often part of the “magic†combinations of horns and drivers people have found, where these two issues are self compensating.
If one had a horn that had constant directivity, then one would still see the drivers inherent power roll off.
If one had two horns of identical response etc on axis, one being CD the other exponential etc *, then on axis, they would sound nearly exactly the same.
(* I am not sure this could be done actually.).Anywhere else in the room and the difference would be very clear.
For a horn which uses focusing to compensate for the power roll off id flat on axis but its radiated power still rolls off starting at 2 –4 KHz.
A CD horn, flat on axis, also radiates a flat acoustic power curve. Off axis, this system sounds nearly the same spectrally as on axis but of course there isn’t a stereo image off axis. If your working in a room with a CD horn system, I find it is very nice and listen able 90 degrees off axis. An Altec A-7 on the other hand can sound very dark and murky off axis (one I’m familiar with in this room) because of its collapsing directivity / falling power response.
For both kinds of horns one often sees a narrowing in the pattern just before pattern control is lost at the low frequency end. To combat this, it is normal to have a short mouth section which has about a 30% greater angle than the rest of the horn to combat this narrowing.
Mating to the adjacent drivers is a problem with horns, one has to have two sources less than about 1 / 4 wl apart in order to have unilateral addition. Greater spacing means a new radiation pattern is formed according to the dimensions and positions of the two sources. Here symmetry is your friend as in direct radiators where this is also an issue.
Hope that helps,Tom
Hey those curves look “realâ€, so that would be Karlson Kitchen Kurves?
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: horn signature - whats left after proper tailoring? - tomservo 01/23/0619:04:38 01/23/06 (36)
- Agreements and disagreements - EGeddes 12:38:17 01/24/06 (35)
- Re: Agreements and disagreements - tomservo 08:18:01 01/25/06 (3)
- Re: Agreements and disagreements - EGeddes 08:55:03 01/26/06 (2)
- Re: Agreements and disagreements - Mark Seaton 10:20:16 01/26/06 (1)
- I would still contend - EGeddes 10:52:55 01/26/06 (0)
- Re: Agreements and disagreements - Tom Dawson 22:57:23 01/24/06 (30)
- Re: Agreements and disagreements - EGeddes 05:39:22 01/25/06 (29)
- What if? - TDM 06:30:29 01/25/06 (28)
- There are other problems here - EGeddes 07:04:55 01/25/06 (27)
- Re: are there, other problems here - tomservo 18:23:34 01/25/06 (24)
- Re: are there, other problems here- fix that karlson - lol - freddyi 15:13:58 01/26/06 (0)
- Interesting - EGeddes 08:27:04 01/26/06 (22)
- Re: Interesting - tomservo 06:08:33 01/27/06 (21)
- I tried - EGeddes 06:30:58 01/27/06 (0)
- But its not an open standard - EGeddes 06:15:49 01/27/06 (19)
- Re: But its not an open standard - tomservo 09:37:19 01/27/06 (18)
- I know - EGeddes 09:48:00 01/27/06 (17)
- Re: I know - tomservo 12:35:08 01/28/06 (1)
- I can probably do it - EGeddes 12:23:55 01/29/06 (0)
- Re: I know - bzdang 13:59:47 01/27/06 (14)
- No - EGeddes 14:29:02 01/27/06 (11)
- Polar maps are great! - John Sheerin 15:46:54 01/27/06 (10)
- Re: Polar maps are great! - EGeddes 18:13:49 01/27/06 (8)
- Re: Polar maps are great! - John Sheerin 18:18:57 01/27/06 (7)
- OK so what was - EGeddes 18:53:41 01/27/06 (6)
- Re: OK so what was - John Sheerin 06:45:04 01/28/06 (5)
- One more thing - EGeddes 07:19:58 01/28/06 (0)
- FEA vs BEM - EGeddes 07:12:59 01/28/06 (3)
- Re: FEA vs BEM - John Sheerin 07:47:41 01/28/06 (2)
- Re: FEA vs BEM - EGeddes 08:44:52 01/28/06 (1)
- Re: FEA vs BEM vs "what do I want" - freddyi 12:13:46 01/28/06 (0)
- Thanks for the clarification - bzdang 17:38:46 01/27/06 (0)
- Re: I know - kool - freddyi 14:11:43 01/27/06 (1)
- Re: I know - kool - bzdang 14:13:49 01/27/06 (0)
- Re: There are other problems here - Mark Seaton 08:30:18 01/25/06 (1)
- I will wait - EGeddes 07:57:23 01/26/06 (0)