La Scala bass horn length

[Islander]

Does anyone know the actual length of the La Scala bass horn? I'm assuming one should measure from the woofer itself, perhaps at the dust cover. Does the sound, for measurement purposes at least, stay in the middle of the horn, rather than travelling along the walls? Since the horn is 3 inches wide and makes two turns, one after exiting the woofer slot and one at the rear corner (mirrored at the other rear corner), that's 1 1/2 inches times 2, so if the sound travels in the middle of the horn, that would make the horn 3 inches shorter in effective length than would be measured using the sides of it.

Has anyone seen any figures for the effective length? Also, for purposes of setting delay for time alignment of drivers, is the driver's cone or diaphragm the proper measuring point, or is there an "acoustic center" part-way along the horn that should be used instead?

[JohnA]

I don't think there is significant error in 36" for the horn length. If you want to be exact, the throat is the 3" x 9" rectangle in front of the splitter in front of the woofer. Then, the air column turns in the back corner, and runs beside the doghouse. Subtract 1/2 of the width of the slot beside the doghouse from 36".

The acoustic center would be the voice coil of the woofer. I'd measure from the gasket surface to the edge of the dust cap and add that to the path length calculated above.

[Islander]

I figured the total sound path was around 35 1/2", but just wanted to know if I was overlooking anything. Since I'm not sure of the distance from the doghouse slot to the dust cover and I can't see into there, I didn't want to go poking blindly around the woofer with a tape or a ruler. I assumed the distance from the speaker's mounting surface to the dust cap would be about 3 1/2". The sound path of the 510 horn/K69 driver combo is about 6", added to its present 1 1/2" backset from the front of the bass bin, for a total difference of 28".

Earlier, I'd somehow come up with a 35" difference. When I convert that to milliseconds, that was 2.63 milliseconds of treble delay, whereas the 28" setting gives 2.06 milliseconds. When I listened for a difference a few minutes ago, there may have been a subtle improvement, but it's not obvious. That's okay, a few barely-audible subtleties add up to a noticeable improvement. The Dx38 allows for a degree of precision in settings that really appeals to me.

Thanks, John.

[djk]

I Just figure it is about 36" to the dustcap, but try a little longer because the acoustic center may actually be a bit behind the driver.

[Islander]

I'm not clear on the concept of acoustic center. Is that the spot in the air where the sound actually forms? How can it be behind the cone?

[djk]

Because it has mass and inertia.

[Don Richard]

I think all of the Klipsch folded bass horn designs' final flare continues past the wood to a point in front of the physical structure of the loudspeaker, creating a larger horn mouth than is apparent. The air velocity is high enough that the pressure wave "doesn't realize" that it is no longer bounded and continues for a certain distance in front of the cabinet resulting in a sort of virtual final flare.

[PrestonTom]

Ultimately you are interested in the difference in propagation delay between the two drivers. So there is an old fashioned way of figuring this one out.

1. Invert the the polarity on one of the drivers.

2. feed a pure tone at the crossover frequency (I'll guess - 700 Hz).

3. Adjust the time delay (to the midrange) until the amplitude (from both drivers playing simultaneously) is minimized.

4. Now you have the time difference (or you are off by a full wavelength of about 18inches).

Good Luck,

-Tom

[Islander]

Thanks for the tips, Tom, but I don't have a tone generator. Resetting the treble delay to 2.06ms/28"/71cm, based on tape measure figures, seemed to dial in the sound nicely and I like what I'm hearing.

The crossover frequency with the JubScalas is around 445Hz.

[Islander]

I think all of the Klipsch folded bass horn designs' final flare continues past the wood to a point in front of the physical structure of the loudspeaker, creating a larger horn mouth than is apparent. The air velocity is high enough that the pressure wave "doesn't realize" that it is no longer bounded and continues for a certain distance in front of the cabinet resulting in a sort of virtual final flare.

Isn't the high velocity area at the throat of the horn, where it meets the compression driver? The large size of the horn's front opening would result in very low airspeed combined with large volume, wouldn't it?

[Don Richard]

Isn't the high velocity area at the throat of the horn, where it meets the compression driver?

The throat of a horn is a high pressure area. The shape of the horn acts as an acoustical transformer converting the high pressure/small volume at the throat to a low pressure/ larger volume at the mouth. I have not measured the air velocity but it is sufficient to extinguish a candle at the mouth at high SPLs.

[DrWho]

The volumetric flow rate is the same at both ends of the horn, so where the cross-sectional area is smaller you have higher velocity.

[Don Richard]

The air velocity is high enough that the pressure wave "doesn't realize" that it is no longer bounded and continues for a certain distance in front of the cabinet resulting in a sort of virtual final flare.

The acoustical length is nearly always longer than the physical length for a horn. When designing a horn, end correction must be applied.

http://ldsg.snippets.org/HORNS/design.html

[DrWho]

bummer he didn't show any data...I almost wonder if it doesnt make more sense to think of end correction as a limitation of performance - like anything below perfect end correction yields less extension.

[djk]

End correction?

Always asserted without data.

Here is data from a respected company (EAW). It shows that there is no end correction factor, and that quadrupling the mouth area (X4 cabinets) only gives the theoretical 6dB of gain up around 200hz, and virtually nothing at the low cut-off point (maybe 1.5dB).

[WMcD]

I can see what djk is saying, or rather what the graph is saying.

It seem to me that the issue lurking in here is the effect of lack of lenght. Olson has a comparison of horns of different lengths. The throat impedance roll off as we get to the bass region gets worse as the horn is shortened. It is not a matter of mouth size alone.

The graph shows the effect of stacking finite horns which are probably short and have that issue. Grouping them should solve the problem of a small mouth. But it can't solve the length issue in this manner.

Gil

[Islander]

The graph shows the effect of stacking finite horns which are probably short and have that issue. Grouping them should solve the problem of a small mouth. But it can't solve the length issue in this manner.

You mean we can't get perfect sound with horns of finite length? But the infinite-length horns need such big amps! And an infinite number of sheets of plywood...

[Don Richard]

The volumetric flow rate is the same at both ends of the horn, so where the cross-sectional area is smaller you have higher velocity.

Dr. Eugene Patronis wrote in the third edition of Sound System Engineering:

"Horn surfaces define a bounded region whose cross-sectional area increases from the input to the output in a loudspeaker application. The acoustic power flowing through a cross section of area S, when the acoustic pressure and particle velocity are in phase, is the product puS. The product uS is called the volume velocity and is denoted by U, therefore the acoustic power flow is pU. At the input end of the horn where S is small and the acoustic pressure is large, the volume velocity for a given acoustic power is small. At the output end of the horn where S is large, the volume velocity is large and the acoustic pressure is small for the same acoustic power."

[Don Richard]

I almost wonder if it doesnt make more sense to think of end correction as a limitation of performance - like anything below perfect end correction yields less extension.

Shortening a horn does affect performance, making the frequency response irregular compared to a horn of "proper" length. The LaScala bass horn is certainly a shortened horn and, as a result, has less low end extension and a rougher frequency response than a Khorn. What allows it to work as well as it does is the fact that low frequency boundary reinforcement comes into play. The floor, wall, and even the ceiling act as extensions of the horn thus allowing operation at frequencies that the mouth size would not seem to support. Then high particle velocities at the mouth complete the flare to sufficiently horn-load the driver, further smoothing and extending the frequency response.

The Khorn also employs these techniques, except that the flare rates allow for lower frequency output and the corner horn is designed to load into 1/8th space, using the walls as the final flare.