efficiency of horns

[xcortes]

I made a pair of horns using Fostex FE103E drivers in the Fostex BLH recommended enclosure. The driver efficiency is 89dB. Are my horns more efficient than the driver inherent efficiency or do I gain efficiency by placing the drivers in horns (or in other type of enclosures)?

I've read a lot about horns being very efficient but have not being able to find info on whether they are more efficient because they use more efficient drivers or because the speaker system is more efficient vis-a-vis other system using the same driver.

Thanks

[Swedish Chef]

Horns are more efficient than other box types because they match the acoustic impedance of the driver to air. So generally speaking ANY driver would become much more efficient when put in a horn, rather than in a T/S cabinet. BUT not all drivers work well in horns. In fact, only a few do. Generally speaking, strong motor structures, low Qts, high Fs (thus light cones) are the most suitable ones. This pretty much rules out low-efficiency drivers like Vifa, Peerless, Scan-Speak, Focal, Seas et al.

Wait for Bill Fitzmaurice to chime in. He could likely answer most of your horn questions.

You're on the right track though. Mother Nature wants us to horn-load. That's what she did when designing your ears!

/Magnus

[mgoedeke]

In the case of a rear-loaded horn like I presume you are referring to, the horn serves to load the rear of the driver and get more bass output for the same amount of excursion as more traditional enclosures. The rear-chamber of the horn acts as a low-pass filter and thus the mid-band sensitivity of the driver remains pretty much unaffected. Read up on Nelson Pass’s J-Low article on www.passdiy.com for a pretty good explanation. He uses an eq in combination with a bigger horn though.

Front-loaded horns when done right end up in the 25-45% efficiency range and sometimes even more, but they have limited bandwidth.

[Tim Moorman]

xcortes

Rear horn, right?

You gain in efficiency, but only for the bandwidth of the design
(basically a bandpass device). For example, the Fostex rear horn design is really using the horn to counter the normal low frequency roll-off of the driver at, say 150 Hz downward. By lifting the bass output using the back wave of the driver you extend and flatten the low end response. Rear horns are out of phase with the direct radiation front of the driver (like a bass reflex enclosure).

This is probably only effective for an octave or so. Above the point where the horn is working you just have the output of the driver working as a direct radiator, forward firing, where you gain perhaps a dB from baffle mounting over the rated efficiency, if that. Front horns would retain and even improve phase relationships of the raw driver, but that's another matter.

In most cases, a well designed horn will only cover perhaps a decade of the sound spectrum, if that, i.e., 50 - 500 Hz. There are exceptions.

Many manufacturers horn load lower efficiency tweeters because it's physically smaller and less demanding work to get the extra bump in efficiency. Plus, you gain the added benefit of dispersion control. If the normal dome tweeter is spraying sound around in the usual 180 degree + pattern of a 1" direct radiator dome, by adding the horn the pattern control may be reduced to, say, 90 degrees and the output will gain several dB on axis.

Try it yourself with a conical cardboard flare. Take the lowest frequency you want to cover with the tweeter and figure the wavelength by dividing the speed of sound by the frequency of interest (roughly 1132'/second = 13584"/2000 Hz= 6.79" approx. Make the cone around 6 3/4" - 7" in depth front to back for 2 KHz with about a 45 degree flare (90 degrees overall), and tape it on the front of the speaker. The opening for the driver should be no larger than the cone/dome of the driver. (For clarity sake, this is not a true horn, just a minimal waveguide.)
Most horns are 1/2 or even 1/4 wavelength designs, again for ease of manufacture, but usually at the cost of some irregularity in response.

Have a look at David McBean's "Horn Response" program if interested in designing your own.


Tim

[BillFitzmaurice]

Rear horns are out of phase with the direct radiation front of the driver (like a bass reflex enclosure).

That's not so in both cases; if it were they wouldn't work. The output of a port is in phase down to the Fb, due to the capacitive nature of the reflex box. It is only below Fb that the port output reverses in phase, which accounts for why reflex boxes roll off as 4th order filters and sealed boxes as only 2nd order filters below Fb.

Horns have two factors that control phase issues. One is the rear chamber, which also acts as a capacitor. The other is the sheer length of the horn which time delays the signal long enough so that by the time the front and rear waves meet they are in phase within the horn passband, in the same fashion that a plain baffle does.

[Tim Moorman]

Bill,

Correct. I'll rephrase. Both reflex enclosures and rear horns utilize the backwave(rareification) of the driver, which is out of phase with the usual forward cone movement(compression). In both cases the enclosure corrects the phase at output, but at the expense of some delay in time, depending on the pathway length.


Tim

[filgor]

Horn loading is one of those issues I have seen a lot of weird opinions on... most of which are based on misconceptions and some of which are based on the fact that you can describe the same thing from a different view point and sound like you're saying something completely different.

Both theory and practice say one thing is for sure: horn loading will increase the SPL at a given distance for the same input power regardless of the driver you are using. Horns only fail to boost SPL when the frequency is so low that the driver doesn't see the horn (this occurs where the wavelength of sound is much greater than the horns mouth width) and where the freqency is so high that the wave propagation from the speakers diaphram is naturaly very directional (this occurs at Fd and above which is around 1400Hz for a 4" driver and around 5600Hz for a 1" driver) the weight of the cone has NOTHING to do with this. In fact heavy coned loudspeakers have more to gain from horn loading than light coned loudspeakers due to a greater impedance mismatch. The physical theory of this is far too advanced for the scope of this particular query.

In short, a horn will boost the sensetivity of your drivers above nominal over a narrow but calculatable bandwidth of sound.
This is why most horn loaded systems are 3 or 4 way or more.

PS. The boost and resonances produced by horns is a complex issue. I would only carefully consider them for any use other than PA systems

[BillFitzmaurice]

Correct. I'll rephrase. Both reflex enclosures and rear horns utilize the backwave(rareification) of the driver, which is out of phase with the usual forward cone movement(compression). In both cases the enclosure corrects the phase at output, but at the expense of some delay in time, depending on the pathway length.

That's better. And in the real world the time delays involved are insignificant with long wavelengths; it would be a different story above 300 Hz or so.

In short, a horn will boost the sensetivity of your drivers above nominal over a narrow but calculatable bandwidth of sound

Actually a well designed horn can operate over a bandwidth of four to six octaves; even a folded sub horn with an intentional low-pass function can work quite well over a three octave range. A 3 way horn-loaded system covering from 16Hz to 20kHz isn't unusual, and direct radiators generally require three elements to properly cover the same range.