Hi,
There are lots of small drivers and I want to put them in a waveguide and see the polar response.
The drivers I am talking about are these, probably used in mobile phones.
https://www.farnell.com/datasheets/4197292.pdf
https://www.farnell.com/datasheets/4197300.pdf
https://www.farnell.com/datasheets/3999554.pdf
I have seen better round ones available locally and measured them to be much flat than above ones and Fs of 750Hz.
If a mobile can make sufficient noise then these should be able to act like a nice tweeter when several are put in a waveguide.
Please note Don Keele used used 48 AKG XXL‐V5 Headphone Drivers in one and Toshiba laptop drivers in another and Harman multimedia driver in his CBT prototypes.
https://www.linkwitzlab.com/Keele - Introduction to CBT Loudspeaker Arrays.pdf
Before I get to the actual dirty work with 3D printing the entire thing, can somebody help me please simulate them?
I would appreciate if someone could simulate 15mm sound sources put on a spherical cap in an 8 inch 90 degrees conical waveguide with good round over leading to the baffle.
The following scenarios could help understand whether the idea has merits
a) 7 sources ie one in center and 6 surrounding in a circle, concentric arrangement on spherical cap, the spherical section should perpendicularly mate with the wall of the waveguide, all driven by same signal.
b) Same 7 sources in concentric arrangement on spherical cap, the side ring shaded by -3db than the center one. To get 90 degrees wave front the spherical cap should 138 degrees as per CBT theory.
c) 19 sources ie one in center and 6 surrounding in a circle, and another 12 in a surrounding circle concentric arrangement on spherical cap, the spherical section should perpendicularly, all driven by same signal.
d) Same 19 sources in concentric arrangement on spherical cap, center one at full strength, the next circle of 6 at -3 dB and the outermost circle of 12 at -6dB. To get 90 degrees wave front the spherical cap should 138 degrees as per CBT theory.
The b) and d) are essentially small spherical CBTs in a waveguide and a) and c) are unshaded curved drivers in a waveguide.
Few more driver in Don's presentation
I will attempt a 3D printed version of whichever gives good polars and post here the measurements.
Thanks and Regards,
WonderfulAudio
There are lots of small drivers and I want to put them in a waveguide and see the polar response.
The drivers I am talking about are these, probably used in mobile phones.
https://www.farnell.com/datasheets/4197292.pdf
https://www.farnell.com/datasheets/4197300.pdf
https://www.farnell.com/datasheets/3999554.pdf
I have seen better round ones available locally and measured them to be much flat than above ones and Fs of 750Hz.
If a mobile can make sufficient noise then these should be able to act like a nice tweeter when several are put in a waveguide.
Please note Don Keele used used 48 AKG XXL‐V5 Headphone Drivers in one and Toshiba laptop drivers in another and Harman multimedia driver in his CBT prototypes.
https://www.linkwitzlab.com/Keele - Introduction to CBT Loudspeaker Arrays.pdf
Before I get to the actual dirty work with 3D printing the entire thing, can somebody help me please simulate them?
I would appreciate if someone could simulate 15mm sound sources put on a spherical cap in an 8 inch 90 degrees conical waveguide with good round over leading to the baffle.
The following scenarios could help understand whether the idea has merits
a) 7 sources ie one in center and 6 surrounding in a circle, concentric arrangement on spherical cap, the spherical section should perpendicularly mate with the wall of the waveguide, all driven by same signal.
b) Same 7 sources in concentric arrangement on spherical cap, the side ring shaded by -3db than the center one. To get 90 degrees wave front the spherical cap should 138 degrees as per CBT theory.
c) 19 sources ie one in center and 6 surrounding in a circle, and another 12 in a surrounding circle concentric arrangement on spherical cap, the spherical section should perpendicularly, all driven by same signal.
d) Same 19 sources in concentric arrangement on spherical cap, center one at full strength, the next circle of 6 at -3 dB and the outermost circle of 12 at -6dB. To get 90 degrees wave front the spherical cap should 138 degrees as per CBT theory.
The b) and d) are essentially small spherical CBTs in a waveguide and a) and c) are unshaded curved drivers in a waveguide.
Few more driver in Don's presentation
I will attempt a 3D printed version of whichever gives good polars and post here the measurements.
Thanks and Regards,
WonderfulAudio
So essentially it's the source configuration you want to test. The waveguide appears to need to be conical, since you'll provide an approximation of a spherical source. The properties of the waveguide will be primarily determined by the radiation requirements, not by the source...
Yes, Allen, you put it correctly. Thanks.
By studying Don's papers I realized that the polar smoothness is sensitive to phase (curvature) but not SPL which means if we give the right phase by way of curvature we should get a smooth wave front from it and then on the lower frequency where the CBT loses directivity and starts to widen the the walls of the waveguide take over. We are using less steps of shading here and that seems to be not a bad compromise. The net result is we dont need so many drivers as a normal CBT and we dont lose treble as full sized CBT. We get directivity partially from CBT on the higher side and partly from the waveguide on the lower side. I believe the polars will be very good and need validation from simulation whether this is correct.
By studying Don's papers I realized that the polar smoothness is sensitive to phase (curvature) but not SPL which means if we give the right phase by way of curvature we should get a smooth wave front from it and then on the lower frequency where the CBT loses directivity and starts to widen the the walls of the waveguide take over. We are using less steps of shading here and that seems to be not a bad compromise. The net result is we dont need so many drivers as a normal CBT and we dont lose treble as full sized CBT. We get directivity partially from CBT on the higher side and partly from the waveguide on the lower side. I believe the polars will be very good and need validation from simulation whether this is correct.
Firstly, Keele talks about shading when a properly shaped array is open at the edges. In the case where you use a walled waveguide with a properly shaped array, you should not use shading. The level of all the drivers should be the same or in the case that different sized drivers are used across the array, their intensity should be the same. If not you will produce higher order modes.
Secondly, a properly shaped source should ride normal to the waveguide walls at all frequencies. The lows shouldn't be the exception here until you begin to terminate the waveguide.
Secondly, a properly shaped source should ride normal to the waveguide walls at all frequencies. The lows shouldn't be the exception here until you begin to terminate the waveguide.
In the high frequency it's a cbt alone that is holding the directivity but in the midtange it's same as a compression driver. The cbt will lose directivity just like a 1 inch throat compression driver does not have the directivity but the waveguide contains it. IMHO same holds good in what I propose
You have a choice. If you develop high frequency directivity within the open space between the walls of the waveguide, the diffraction that naturally develops afterward is going to colllide with those walls.
That's 20 equal strength (no shading) sources in a waveguide. Pls note, it may be a ribbon but it radiates as 20 discrete sources.
Of course, this one is in one plane, I am talking about two planes with proper curvature.
Yes it is.
As you mentioned about the array establishing it's own directivity, it's size is one of the things which gives it that level of control. Shading would stand it apart from the waveguide and not shading would be appropriate for going with it.
The sections from the ribbon you show are probably acoustically small. This is something you'd want to control with your array, since you will approximate one source that way.
We could discuss reflections off the back of the grille but I'm not sure it's as important as noting the source appears to be flat. This makes it dependent on the waveguide to 'tease it out' (widen the highest frequencies) to meet the waveguide walls.
As you mentioned about the array establishing it's own directivity, it's size is one of the things which gives it that level of control. Shading would stand it apart from the waveguide and not shading would be appropriate for going with it.
The sections from the ribbon you show are probably acoustically small. This is something you'd want to control with your array, since you will approximate one source that way.
We could discuss reflections off the back of the grille but I'm not sure it's as important as noting the source appears to be flat. This makes it dependent on the waveguide to 'tease it out' (widen the highest frequencies) to meet the waveguide walls.