Horn questions

So, I'm a bored college student with nothing to do but scheme about ways to improve my audio setup over the summer. I currently have a pair of 15" woofers from an old console unit tossed into 3 ft^3 boxes. I was thinking about ways to improve the efficiency of these, and am thinking about converting them to horns. However, after poking around a bit with google, I've found myself faced with a couple questions:

1. What influence does the choice of the curve defining the expansion of the horn have over the response?

2. How does the throat area influence the response?

3. Ideally, the horn should be 1/4 long as the lowest wavelength you're covering, correct? Or, am I thinking of a TL here? Because that gets some really big mouth sizes for an exponential horn.

Any pointers to specific resources would be appreciated.


Note: I do not know the T/S specs of these woofers, and do not have a test jig to measure them. So, any horn I design would be a general one, aimed mostly at increasing efficiency.
 
slackerbob said:

1. What influence does the choice of the curve defining the expansion of the horn have over the response?

2. How does the throat area influence the response?

3. Ideally, the horn should be 1/4 long as the lowest wavelength you're covering, correct?

Quick 'n dirty answers:
1. Different expansion curves have different lengths to get to the same mouth size, since 1/4 WL of the lowest freq you want to produce is a minimum, the expansion curves define the lowest freq your horn can safely reproduce for a given mouth size. (it's not that easy, mouth size itself is also a factor, but this is just to give you an idea.)
Also, the longer your horn, the greater your efficienty. A tractrix curve gives less eff than an exponential, and a hyperbolic curve gives even greater eff. Tractrix flares the most, hence is the shortest, hyperbolic was the lowest flare rate, hence the longest.
However, i read from several sources that a hyperbolic gives more distortion in the very low freq, due to air nonlinearities ... pick your posion. :apathic:

2. Lowpass filter. Smaller throat is less high freq, and more efficiency in the lowest freq. Try to overdo it, and instead of a relative flat responce, you'll get a bass-bump at the lowest freq your horn is designed for, and a rapid fallof above that.
Beware, the M. Leach paper's calcs give a way too small throat area, the only downside of this exellent paper!

3. Yes. A horn is like a pipe sealed at one end, and the first resonance point is at 1/4 WL. I recently looked this up again in my Physics course from my uni days. Guess my memory isn't what it used to be. :eek:

Tip: use Hornresp and play with the different parameters, you'll quickly come to the same conclusions.
 
Thanks for all the replies. I'm still reading through Leach's paper. I haven't read this sort of techinical literature for a couple months.

It looks like I'm going to figure out the parameters for a better idea of a proper horn.

For a mouth, should 1/4 WL be equal to the circumfrence, or the diameter of a circular mouth?
 
http://www.decware.com/imperial.htm
Here is a good horn if you want increased efficiency and good deepend extention.

"Along the journey we quickly found out that nothing man made that you could buy in any stereo shop could ever come remotely close to this cabinet as a subwoofer. It would make bass come out of almost anything. A 25 watt 10 inch pioneer woofer shook items off a neighbors shelf in the house across the street. Distance: 220 feet. It did that with one channel of a 22 watt class A receiver."

"The response of the woofer in the cube was 3 dB down at 120 cycles. Basically ZERO bass. We installed this cube into the Imperial with the woofer facing the back of the encosure. The original 15 inch speaker opening was sealed. Installed the response was a bit different. Starting at the same 120 cycles and measuring the SPL at 90 dB in the room we slowly started to sweep the frequencies down until at 28.5 cycles the SPL had risen to 118 dB. That's a 28 dB of gain at 28.5 cycles! I really can't begin to describe what happens at that frequency when you hit that SPL, but it's serious. We saw a mouse stagger out of a crack in the concrete floor and die. "
:eek: :D :D :D
 
at 28.5 cycles the SPL had risen to 118 dB. That's a 28 dB of gain at 28.5 cycles! I really can't begin to describe what happens at that frequency when you hit that SPL, but it's serious. We saw a mouse stagger out of a crack in the concrete floor and die.

I'm calling BS on that one....

Back loaded horn... I've heard bad things! And I've never heard of 28db gain period from a horn 10db sure, but 28 is insane
 
If you read the text carefully you will see that he has created a fourth order bandpass box with a rather large horn as a port. You can get some serious efficiency out of 4th order BP box if you just tune it correctly. In this example it probably has a rather pronounced peak at 28,5 hz and with the horn and roomgain you will get 28 dB of gain.