Acoustic wave canon

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
We could explain the theory to you, but the big question here is:

>>> Why? <<<

If you want a tubular subwoofer with a lot of output, try the SVS line of subwoofers.

http://www.svsubwoofers.com

Well-known in the Home Theater market for providing the best bass for the best price.

I'll admit that I don't know the theory behind the Wave Cannon but I do know that it takes a lot of equalization for it to give decent output. This means operating the driver closer to its maximum tolerances than usual.

The SVS subs, on the other hand, require no EQ and use no artificial bass boost, but still produce large amounts of earthmoving bass. They also are good at integrating, and they are known to be very "musical" as far as subs in their class are concerned.
 
There was once an article in Speaker Builder dealing with a waveguide subwoofer.
There was also an article in JAES about modeling of horns etc using cylindrical and conical sections where there is an example using the same topology that the B**e device uses.

I will try to find the articles.

This thingie is actually something like the TL equivalent to a double-chamber bass reflex enclosure. The ratio of the tube lengts is 1:3.
I think Mr. Bose has a strong affinity for using heavily resonant devices to get strong bass response cheaply - with questionable quality IMHO.

Regards

Charles
 
As far as I can recall the Bose cannon can be approximated by placing a long stroke driver (Peerless 8 ohm XLS 10 inch perhaps would be good) 1/4 of the way down a 6m long pipe.

I'll need to crunch the graphs again to look at the predicted response but from memory it was fairly flat (peaks and troughs less than +/-2dB ) from around 18Hz up to about 90.

Not sure what all this talk of eq is, my predictions based on phase summations of the outputs didn't seem to indicate that it'd be that bad (as resonant pipes go).

The cannon was developed more as a means of getting significant LF sound pressure into theatres more than as a high fidelity beast I think but I'm keen to have a go at it myself.

I"ll be back in touch once I've crunched some numbers.

drew
 
diyAudio Moderator Emeritus
Joined 2001
Couple of Bose patents. First, what looks like his latest version of his radio.

Note the patent number if you want details at the US Patent Office website. Don't forget to download one of their image viewers. IE 5 will drive you nuts trying to download one of them.

21A is where the speaker goes. 24 is the polyester stuffing the Line is stuffed with. I think 31 is where the other speaker is mounted, and the same thing is repeated beneath this drawing-two levels, as it were.
 

Attachments

  • bose radio.gif
    bose radio.gif
    13.4 KB · Views: 10,070
diyAudio Moderator Emeritus
Joined 2001
Now this might be for the thing you might think is the "Cannon". From the patent itself:

"United States Patent 4,628,528
Bose , et al. December 9, 1986

--------------------------------------------------------------------------------
Pressure wave transducing


Abstract
A loudspeaker driver has its front surface adjacent one end of a low loss acoustic waveguide and its rear surface adjacent to one end of a second acoustic waveguide that is one third the length of the first. The other openings of the waveguides face air and couple acoustical energy substantially uniformly over a relatively broad range of frequencies extenting into the bass frequency region. An equalizer includes a notch filter so that the frequency response of the equalizer below a bass cutoff frequency is sufficiently low to prevent audible distortion."

I think the long tube is between 1/4 and 1/2 the wavelength of the F3 frequency aimed for, and the short tube behind is 1/3 the length of the long tube. It might not be clear from ther illustration, but both ends are open to the air.
 

Attachments

  • bose cannon i think.gif
    bose cannon i think.gif
    20 KB · Views: 10,042
diyAudio Moderator Emeritus
Joined 2001
Here is another page from the patent. Again, note the patent number. There does indeed seem to be an equalizer involved, which schematic is included below.

Hope this helps.

Incidentally, in no way does the reproduction of these pictures mean that you can actually legally build this. You should check with Bose Corporation for legal permission. I include excerpts from this patent only for informational purposes-for you to gain some idea of how these things might work.
 

Attachments

  • bose cannon 2.gif
    bose cannon 2.gif
    10.2 KB · Views: 9,237
Actually, as a private individual and not for resale I'm fairly sure that there's nothing to stop you building whatever you want.

There was an article in a pro audio magazine (Connections I think it was) in about 1995 that dealt with the phase/wavelength theory behind the cannon (worked out independantly by them with no input from Bose based upon the length of the original tube) and provided plans for a DIY version.

Re eq, it looks like it's only a bandpass cct to stop really low end from overdriving the unit and a lowpass to tailor it into a system. Pretty much the same as any sub amp I've ever seen. Nothing new there.

Personally I think the long straight pipe is better than the labyrinthine version and the wave radio may well be stretching the physics to near breaking point. IMO Bose have always been pretty heavy on the EQ to try and get the response that is quoted. (Look at the output from a 901 equaliser) (Eeeuurrkk!)

A driver with pipes on either side with one that's 3 times longer than the other corresponds with my call of earlier that you put the driver 1/4 of the way along the 6m tube. Just a different way of saying the same thing. 3/4 is 3 times as long as 1/4.

Drew
 
The filter shown is of course a very ordinary Sallen-Key 2nd order highpass filter ( i.e. for subsonic protection).

If one wants to build such a device he really doesn't need to ask Mr. B**e if he is allowed to do so.

And there are more elegant foldings possible than Mr B@$* suggests (why do I never get this Name posted correctly !?) in his patents.

Regards

Charles
 
This is what I meant when I was talking about more elegant foldings. The length ratio is also 1:3 (i.e. driver moved in by 1/4 of the total line-lenght) as with the straight wave cannon.
It needs less space than a straight wave cannon but has less bends than the labyrinth proposed in the B@£& patent.

I have never tried it myself though I was interested in this woofer principle some years ago. This was the folding i wanted to use back then.
Although I think a straight cannon would work better if you have the space for it.

Regards

Charles
 

Attachments

  • wave_cannon.gif
    wave_cannon.gif
    3 KB · Views: 6,231
diyAudio Moderator Emeritus
Joined 2001
kelticwizard said:
I think the long tube is between 1/4 and 1/2 the wavelength of the F3 frequency aimed for, and the short tube behind is 1/3 the length of the long tube. It might not be clear from ther illustration, but both ends are open to the air.

You know, I have read that patent over, and I am still not exactly sure if the length of the two tubes added together amount to one wavelength of the lowest frequency to be reproduced well or one quarter the wavelength of that frequency.

The illustration shown uses a 4 1/2 inch driver similar to the one used in the Bose 802, and the back tube is about 4 feet or so.

If you add the length of the front tube, that comes to a total length of 5 1/2 feet. Which is about the wavelength of 200 Hz, but one quarter the wavelength of 50 Hz.

The drawing is just there to illlustrate a principle, but one must ask why anyone would go through all that to make a 4 1/2" speaker that goes down to only 200 Hz!

I shall reread the patent and try to come up with an answer as to length of the tube vs. frequency.

The patent also says that a good cross section of the tubes would be 2/3 to 1/2 the cross section of the speaker cone.
 
diyAudio Moderator Emeritus
Joined 2001
Is this the unit you were referring to? It is 12 feet long.

Twelve feet is one quarter the wavelength of 24 Hz, and the Cannon is rated down to 30 Hz or 25 Hz. However, I do not know if the tube is straight, or if there is a folded or even a spiral arrangement inside. The senstitivity is listed at 90 dB, with a peak of 115 dB.

The width is 17". There is a bracket on it-I don't know if they refer to the width including the bracket or just the tube. If they include the brack et in the specs, that would make the outside dimensions of the tube 14", I suppose.

Even with a figure of 14" outside diameter for the tube, that would still make the outside dimensions over 13 cubic feet. There are ways to produce 25 or 30 Hz at 90 dB sensitivity in enclosures much smaller than that.

For instance, an Adire Shiva could be put into a box about a third that size and produce 25 Hz at the rate of 88 dB/watt.

Anyway, I will reread the patent, unless someone else comes up with an answer as to the ratio of tube length to lowest wavelength reproduced.
 

Attachments

  • bose cannon picture.gif
    bose cannon picture.gif
    50 KB · Views: 7,178
diyAudio Moderator Emeritus
Joined 2001
DrewP said:
As far as I can recall the Bose cannon can be approximated by placing a long stroke driver (Peerless 8 ohm XLS 10 inch perhaps would be good) 1/4 of the way down a 6m long pipe.

I'll need to crunch the graphs again to look at the predicted response but from memory it was fairly flat (peaks and troughs less than +/-2dB ) from around 18Hz up to about 90.

Drew's configuration would appear to be consistent with a 12.5 foot pipe with an F3 of 27 Hz, which is what the Bose Cannon is. Drew's pipe is about 50% longer than the Cannon, but his F3 has a wavelength that is about 50% longer as well.

So it looks like there is no special routing inside the tubes-apparently they are just straight tubes. It looks like a good bet to just scale your F3 and the length of your cannon to Drew's and Bose's guidelines.

By the way, the promotional flyer that I got that picture from is located here, along with Bose's specs:
http://www.ccidaho.com/product/technical/Bose/pdfs/awcs2t.pdf
 
I made a quick test

I bought a 10 " speaker (pentivent PV-1030) with Fs of 19
2 sonotube and put the driver in the 1/4 of the tube (12 foot long)

the result is poor :bawling: .It have a resonance at one frequency (deep base!!) but not very loud. (not as loud of my energy es-18xl) The sound is terrible.... I don't know why :( My cross-over is set to 80 Hz 12 db/ octave

I don't have time this week to make other test.....

Any idea is welcome

René
 
Remember the graph of response in the Bose patent? I'm thinking that a 12 foot pipe will place the lower of the 2 peaks at just below 50Hz (possibly 45). If that's the case then your 80hz xo is likely missing the upper of the 2 peaks.

for deep bass I think you need longer than 19foot of pipe.

Could be that current setup corresponds with your room modes as well which wouldn't help matters any.

Drew
 
diyAudio Moderator Emeritus
Joined 2001
Bam:

I think the patent mentioned something about resonant frequency and wavelength. I will check.

The only thing that they mentioned was that the 4 1/2" speaker in the patent illustration was similar to the one used in the Bose 802. However, I was unable to locate Thiele-Small parameters on the 802 speaker.

PS: What an appropiate nickname in a thread about the Acoustic Cannon:D .
 
diyAudio Moderator Emeritus
Joined 2001
Here we go about the resonant frequency. This is from the patent:

"The free air resonant frequency of the loudspeaker driver may be chosen to be that at which the length of the longer of the tubes is a half wavelength and thereby lessen response irregularities that might be produced by resonances between reactive components of the loudspeaker driver and the tube. Preferably, the loudspeaker driver is overdamped to avoid undesired resonances between the loudspeaker and the tube."

Okay, if we do the math, (and if I'm doing the math, are we ever in trouble), that means the resonant frequency should have a wavelength 1 1/2 times the length of the tube, front and back together.

So a 12 foot tube should have an Fs wavelength of 18 feet, which means that Fs = 62.5. Uh-oh.

Overdamped means a low Qts, I believe. Since .38 seems ot be a dividing line of sorts when discussing Qts, I guess that means a speaker with a Qts of .30 or below. this is speculation on my part.

Maybe we shuld look among speakers designed for bass guitar?

Perhaps that line from the patent, "The free air resonant frequency of the loudspeaker driver may be chosen to be that at which the length of the longer of the tubes is a half wavelength" might mean that it should be at least twice the wavelength of the longer tube, in which case the speaker with the Fs = 19 Hz might be fine.
 
diyAudio Moderator Emeritus
Joined 2001
The diameter of the tube also affects frequency response. I mentioned the ratio of the area of the tube to tarea of the cone as being perhaps 2/3 to 1/2. This is from the patent:

"For a given ratio of (B1).sup.2 /R.sub.e the ratio of tube to cone areas (ATCR) typically controls the size of the system response peaks at the frequencies where the tube length is an odd multiple of a quarter wavelength for a single tube. For some typical speakers and an ATCR of 1 these peaks are relatively large. For ATCR of 0.5, the system response is relatively smooth. For ATCR less than one half, system response decreases because the tube provides increased load on the loudspeaker cone."

A 12" louspeaker typically has a cone area of 86 sq. inches. An 8" tube has an area of 50 sq inches. That is 58% of the 12 incher's cone area. Perhaps putting an 8" tube sown each 12" tube, held in place by 2 12 inch round brackets, would smooth the response.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.