Help me understand what is practical to achieve in terms of frequency response by coupling smaller woofers into an array.
Common examples that we use regularly would be a 2x18" subwoofer, or a 4x10 bass guitar cabinet. We understand that they not only handle more power and therefore go louder than a single driver, but they also go lower.
What kind of practical use is this coupling behavior for arraying drivers?
We see the column arrays of tiny 2" and 3" drivers. The small drivers enable effective HF response, and the long array extends the frequency response lower to some degree, but not typically to full-range. These long (tall) arrays are also typically used to narrow the vertical pattern.
Let's say we made an array of seven 5" woofers arranged in a circle:
.OO
OOO
.OO
Assuming we get the impedance, power-handling, and sensitivity (at 1000Hz) about the same as a single 15" driver. We can imagine such an array won't have the narrow vertical pattern of a column, but instead it will have a wide conical pattern with a lot of lobes and spikes and some effects of interference like comb filtering.
But how sensitive does it get at lower frequencies? Does the coupling of the adjacent drivers bump the LF response? Maybe not to as much as a single 15", but quantitatively?
Common examples that we use regularly would be a 2x18" subwoofer, or a 4x10 bass guitar cabinet. We understand that they not only handle more power and therefore go louder than a single driver, but they also go lower.
What kind of practical use is this coupling behavior for arraying drivers?
We see the column arrays of tiny 2" and 3" drivers. The small drivers enable effective HF response, and the long array extends the frequency response lower to some degree, but not typically to full-range. These long (tall) arrays are also typically used to narrow the vertical pattern.
Let's say we made an array of seven 5" woofers arranged in a circle:
.OO
OOO
.OO
Assuming we get the impedance, power-handling, and sensitivity (at 1000Hz) about the same as a single 15" driver. We can imagine such an array won't have the narrow vertical pattern of a column, but instead it will have a wide conical pattern with a lot of lobes and spikes and some effects of interference like comb filtering.
But how sensitive does it get at lower frequencies? Does the coupling of the adjacent drivers bump the LF response? Maybe not to as much as a single 15", but quantitatively?
Last edited:
Can you explain more? Are you saying that between a 1x18 and a 2x18, the 2x18 is more efficient, but the frequency response curve is basically the same?
use just one for mid-base, the rest as subwoofers below 200 Hz
the modern sealed box subwoofer trick is to EQ and throw lots of (cheap, Class D) power at long throw drivers in a small box to extend the lows
since you still need lots more displacement volume as you go down more drivers could help
better use of many sub drivers would be to make more sub boxes, distribute them around the room and individually EQ to sum to a even low frequency response at the listening position in the room's modal region
the modern sealed box subwoofer trick is to EQ and throw lots of (cheap, Class D) power at long throw drivers in a small box to extend the lows
since you still need lots more displacement volume as you go down more drivers could help
better use of many sub drivers would be to make more sub boxes, distribute them around the room and individually EQ to sum to a even low frequency response at the listening position in the room's modal region
Fully agree and add: some will claim it doubles efficiency, you get extra "free +3dB" , etc.
Don´t know where that misconception comes from, might apply to two pistons feeding a closed chamber or a looooonnngggg tube, not a realistic case.
In the real World, two side to side drivers will help each other somewhat, but never ever double efficiency.
I mean real double efficiency; people do listen a louder system, ignoring that the main effect is "focusing" the sound towards listener ... which also introduces phase differences, combing, etc.
Don´t know where that misconception comes from, might apply to two pistons feeding a closed chamber or a looooonnngggg tube, not a realistic case.
In the real World, two side to side drivers will help each other somewhat, but never ever double efficiency.
I mean real double efficiency; people do listen a louder system, ignoring that the main effect is "focusing" the sound towards listener ... which also introduces phase differences, combing, etc.
You really don't know where it comes from? You surely know that most all loudspeaker simulation software adds 3dB to the efficiency of a box when you double the amount of drivers in the box (provided you also double the enclosure volume).
While in the real world it may not add 3 whole decibels of sound pressure across the entire dispersion of the enclosure, it's close enough to use as a rule of thumb at low frequencies (and higher frequencies when arrayed vertically)
Ok. I am just going to take some copy from Danley as it is the most recent place I read of such a thing, but I don't think they're the only ones making such a claim.
"The Danley Tapped Horn Subwoofer.... are unlike conventional horns in that combining multiple TH boxes together results in more output, but not significantly more extension."
"The DBH subwoofer is a Danley take on a conventional design. The DBH subs feature a front loaded, folded horn design. The DBH subs behave conventionally in that adding boxes together not only increases the output, but extends the low frequency response downward."
But I don't think coupling the horns is what is extending the frequency response downward:
"It is basically false that multiple bass horns extend the low freq response. Yes they do-but only a Hz or so. What they DO do is to smooth out the low end response (which is typically ragged because the horn mouth is to small for the size wavlength. So it 'appears' as if they are going low-but in reality it is just the part of the response that is already being reproduced is just smoother and louder." - Ivan Beaver
So if multiple horns don't practically extend the response downward, is the notion regarding "conventional" subwoofers simply false also?
"The Danley Tapped Horn Subwoofer.... are unlike conventional horns in that combining multiple TH boxes together results in more output, but not significantly more extension."
"The DBH subwoofer is a Danley take on a conventional design. The DBH subs feature a front loaded, folded horn design. The DBH subs behave conventionally in that adding boxes together not only increases the output, but extends the low frequency response downward."
But I don't think coupling the horns is what is extending the frequency response downward:
"It is basically false that multiple bass horns extend the low freq response. Yes they do-but only a Hz or so. What they DO do is to smooth out the low end response (which is typically ragged because the horn mouth is to small for the size wavlength. So it 'appears' as if they are going low-but in reality it is just the part of the response that is already being reproduced is just smoother and louder." - Ivan Beaver
So if multiple horns don't practically extend the response downward, is the notion regarding "conventional" subwoofers simply false also?
Last edited:
The standardised LF assumptions are that assuming identical drivers, every doubling of Sd results in +3dB, but must be modified depending upon how they are wired, with parallel doubling the current draw relative to a single unit & thereby resulting in a nominal +6dB, & series halving the current draw & thereby providing no net gains relative to a single unit. In practice somewhat more complex, but as an ROT they serve.
They appear to be saying essentially the same thing, albeit with slightly different frames of reference & basis.
-There is little practical increase in extension by just chucking xyz number of identical subwoofers together.
-Horn subwoofers can be an exception. For practical reasons of size most are compromised types (i.e. a terminus < than optimal for a given tuning), so combining multiples, especially in a pro-audio environment where banks of the things are not uncommon, should help compensate, smoothing out the harmonic nulls & potentially lowering the LF end of the gain BW if this was attenuated with a single example through insufficient size. This latter is very much dependent upon the design specifics though so cannot be generalised.
-DSL's [double] tapped horns are a hybrid horn / BP style enclosure and essentially reliant upon resonant action rather than horn coupling in the usual sense, so will be largely unaffected. DSL specialise in producing horn variations for the pro-audio market, which out of necessity tends to be dominated by such enclosures, so their comments are likely meant to be read in that context.
Last edited:
Danley's statement is very misleading. Practically, multiple horns do extend the response downwards, and by a lot more than 1Hz. You can see the effect clearly for yourself in Hornresp or other modelling software.
Technically they don't extend it so much as they smooth it out, but because 'smoothing it out' naturally involves raising the lowest trough in the response (and sometimes that trough will be massive), you lower the F3 of the system substantially.
Direct radiating speakers don't extend their frequency response at all with multiples.
But direct radiators appear to get the same 'smoothing it out' benefit.
This is 1, 2, 3, 4 subs.
Image from Data-Bass
If you don't need as much eq to fill in nulls, you have more headroom, meaning you should be able to set the system to go a bit lower.
Attachments
The speakers don't. The in-room response can ('can' not 'will'). That is a different issue from a horn loaded loudspeaker response smoothing due to compensation for an individually undersized terminus when multiples are physically coupled together. The advantages of distributed multiple subwoofers has been demonstrated for years, most notably by Toole et al at Harman. https://www.harman.com/sites/default/files/multsubs_0.pdf
Last edited:
I forgot the 'distributed' bit.
Also: maybe in-room benefits are not relevant to the OP (if this is for outdoor gigs).
When drivers are <1/3 WL [actually WL/pi] apart they will acoustically couple near enough to +3 dB, so from this we see that since we're meshing together [2] hemispherical patterns that as frequency decreases there's precious little to be gained and as it increases, coupling gain decreases until at ~1 WL they are individual direct radiators.
Easy enough then to figure out how many of what size, specs driver/cab alignment is required to fit 'x' usable BW @ 'y' acoustic efficiency.
GM
Yeah, in either case be it Fb or Fp, this is as low as they go, but the horn starts as a single point source that 'grows' along it length enough to meet the desired HF corner frequency, 1/4 WL of Fp, so as already noted, an array will fill in all the lower end; whereas the 'mouth' area in a direct radiator or TH array is only coupling over a much narrower BW.
Different driver specs required for each to get the same low end response, but basically one needs properly designed horn arrays for wide BW apps and the others are fine for < 2-2.5 octave woofer apps.
GM
That was a very educating read with a scientific end result. Thx for posting I enjoyed it.
- Status
- Not open for further replies.