Hi,
In a waveguide based system the directivity of the waveguide is matched by using a woofer of similar size. The frequency where this happens is about 1khz for a 15" waveguide/woofer.
Since the waveguide looses its directivity at 1khz, a small array (6 or 8 units of 6" or 5.25" drivers) could be used below 1khz to control the vertical directivity. The array would help smooth the ground reflection by using it to its advantage and also provide some directivity vertically. Since the array is small the listener is always in the far field, so there is no near field/far field transition. The drivers are acoustically small as compared to the wavelength so there would be no comb filtering too. Aesthetically, the array could be made to look like a stand for the waveguide.
Would this work to reduce the ground/ceiling reflection, aid directivity (below the waveguide's limit) and reduce the overall size/bulk?
Thanks and Regards,
Goldy
In a waveguide based system the directivity of the waveguide is matched by using a woofer of similar size. The frequency where this happens is about 1khz for a 15" waveguide/woofer.
Since the waveguide looses its directivity at 1khz, a small array (6 or 8 units of 6" or 5.25" drivers) could be used below 1khz to control the vertical directivity. The array would help smooth the ground reflection by using it to its advantage and also provide some directivity vertically. Since the array is small the listener is always in the far field, so there is no near field/far field transition. The drivers are acoustically small as compared to the wavelength so there would be no comb filtering too. Aesthetically, the array could be made to look like a stand for the waveguide.
Would this work to reduce the ground/ceiling reflection, aid directivity (below the waveguide's limit) and reduce the overall size/bulk?
Thanks and Regards,
Goldy
Attachments
It reminds me of JBL cinema designs. However, you will have strong comb filtering in the vertical plane because of the multiple drivers. This will strongly limit or degrade the vertical listening window.
This is hard to avoid if you want to increase vertical directivity, since the directivity increase is caused by destructive interference. If you want to decrease the floor reflection without the problematic comb filtering, you would have to make a CBT or line array from floor to ceiling.
This is hard to avoid if you want to increase vertical directivity, since the directivity increase is caused by destructive interference. If you want to decrease the floor reflection without the problematic comb filtering, you would have to make a CBT or line array from floor to ceiling.
Here is a spec sheet for a typical stacked pair of 15s as used in JBL theater products. Note the second page gives horizontal and vertical response. You can see that below 1000 Hz there is a little evidence of lobing in the 45 degree curve nulling and coming back (that is a lobe). Using 4 mids would give similar response if the line was the same length.
http://cn.jblpro.com/BackOffice/ProductAttachments/4638.pdf
It isn't hard to get directivity to match between woofers and horn, but I would put more emphasis on getting the lateral response well matched than worrying too much about power response.
The secret with short lines is to use variable crossover pints for the various drivers. I've done a lot of symmetrical arrays (so called di Appolito types) with no lobing by careful rolling off of the response to the outer units.
If you want to pursue dealing with the floor bounce you should try modeling the system and its reflected virtual system (the image due to the floor bounce). "The edge", which models edge diffraction can also be set up to model multiple sources, such as your 4 woofers and their 4 reflections.
Regards,
David S
http://cn.jblpro.com/BackOffice/ProductAttachments/4638.pdf
It isn't hard to get directivity to match between woofers and horn, but I would put more emphasis on getting the lateral response well matched than worrying too much about power response.
The secret with short lines is to use variable crossover pints for the various drivers. I've done a lot of symmetrical arrays (so called di Appolito types) with no lobing by careful rolling off of the response to the outer units.
If you want to pursue dealing with the floor bounce you should try modeling the system and its reflected virtual system (the image due to the floor bounce). "The edge", which models edge diffraction can also be set up to model multiple sources, such as your 4 woofers and their 4 reflections.
Regards,
David S
In an array (and all multi ways are arrays), your crossover point will have an impact on lobing.
Here's a graph of the on and off-axis response of my reference speakers. (Gedlee Summa) We see that directivity narrown in the octave from 450hz to 900hz. (The waveguide is 15", the woofer is 15", and the crossover is at one wavelength.)
One way that you can shrink the footprint of a speaker is by using a D'Appolito type array to flank the waveguide. JBL does that in the speaker above, and 'speakerdave' aka Dave Smith has written some AES papers on doing something similar. Mr Smith (Dr?) has a few of these designs in production, notably from Snell and McIntosh.
One of the reasons that we get lobing in an array is that the driver in the center and the drivers adjacent to it are playing the same frequency spectrum, but it's out of phase due to pathlength differences. One incredibly easy way to reduce these effects is to simply curve the baffle. That's what JBL does in their Control Now array, which I think is one of the best inexpensive speakers I've heard in my 40 years.
Here's a graph of the on and off-axis response of my reference speakers. (Gedlee Summa) We see that directivity narrown in the octave from 450hz to 900hz. (The waveguide is 15", the woofer is 15", and the crossover is at one wavelength.)
One way that you can shrink the footprint of a speaker is by using a D'Appolito type array to flank the waveguide. JBL does that in the speaker above, and 'speakerdave' aka Dave Smith has written some AES papers on doing something similar. Mr Smith (Dr?) has a few of these designs in production, notably from Snell and McIntosh.
One of the reasons that we get lobing in an array is that the driver in the center and the drivers adjacent to it are playing the same frequency spectrum, but it's out of phase due to pathlength differences. One incredibly easy way to reduce these effects is to simply curve the baffle. That's what JBL does in their Control Now array, which I think is one of the best inexpensive speakers I've heard in my 40 years.
Last edited:
Someone already did what you are thinking at: The Octagon
Pay attention to the smart crossover of that array, which basically let's the signal pass to the center speakers in the array as the frequency increases.
Some more info on a simpler array you can find in this thread, near the end of it: http://www.diyaudio.com/forums/multi-way/90804-large-midrange-ob-scott-g-40.html#post1953221
Pay attention to the smart crossover of that array, which basically let's the signal pass to the center speakers in the array as the frequency increases.
Some more info on a simpler array you can find in this thread, near the end of it: http://www.diyaudio.com/forums/multi-way/90804-large-midrange-ob-scott-g-40.html#post1953221
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.