Hmmm - Acrylic
Certainly makes the panel "disappear" visually, no small consideration when you're talking about a 12 or 15-inch fullrange coax driver with one or two 15-inch bass drivers on the floor. Heat-treating the outer edges also gives a desirable curved rolloff to decrease diffraction, maybe a 3 to 6-inch radius would be about right, and it also stiffens the front panel as well.
Hmm, food for thought there ... you could even take a hybrid approach, a veneered Baltic Birch plywood cabinet with fairly deep side panels for the bass modules, and a transparent curved acrylic section for the top with gently curved "wings" to minimize diffraction. That would look nice, perform well, and be easier to move around, since the top and bottom would be in separate parts, with the acrylic top resting on a felt layer on top of the Baltic Birch bass units.
Also glad to see the specs for the Hemp Acoustics 8, 12, and 15-inch coax drivers. They look promising for the dipole project.
Certainly makes the panel "disappear" visually, no small consideration when you're talking about a 12 or 15-inch fullrange coax driver with one or two 15-inch bass drivers on the floor. Heat-treating the outer edges also gives a desirable curved rolloff to decrease diffraction, maybe a 3 to 6-inch radius would be about right, and it also stiffens the front panel as well.
Hmm, food for thought there ... you could even take a hybrid approach, a veneered Baltic Birch plywood cabinet with fairly deep side panels for the bass modules, and a transparent curved acrylic section for the top with gently curved "wings" to minimize diffraction. That would look nice, perform well, and be easier to move around, since the top and bottom would be in separate parts, with the acrylic top resting on a felt layer on top of the Baltic Birch bass units.
Also glad to see the specs for the Hemp Acoustics 8, 12, and 15-inch coax drivers. They look promising for the dipole project.
Still Thinking ...
I like that hybrid acrylic/wood cabinet approach, the more I think about it. The top portion is sort of an art object, whether you're talking about Lowther/AER, Jordan, or pro-quality coax drivers. These are expensive drivers and look the part - like a good tube amp or a custom motorcycle, you don't want to hide the technological beauty of the working parts.
The wood bass cabinet is interesting as well. I'm thinking of asymmetrical side walls to minimize standing waves. One wall would be short, maybe only 6 inches deep, and the other would be more like 15 inches deep. If there are two drivers, one would go on the front (close to the corner with the deep sidewall) and the other could go on the deep sidewall. Mirror-imaged pairs, of course.
I'm guessing the bass quality might be a little better if the side woofers fired towards the left and right sides of the room, which would improve the sensation of "air" and dimensionality in the bass region. People with stereo subwoofers know what I'm talking about.
Hmm, that's intereresting, swapping the bass cabinets left-to-right would actually swing the dipole pattern around. With the arrangement of the previous paragraph, the nulls of the dipoles are a bit closer to the listener, with side woofers facing the center, the nulls are steered towards the side walls of the room. One more way to tune the system to the room - steering the dipole nulls either towards the center or the side walls.
With a multi-tapped inductor for the bass drivers to control the overlap region, steerable polar patterns, and a pair of dedicated stereo subwoofers for the below-60~80 Hz region, there are a lot of ways to tune the system to the room. This is good, considering the wide variety of rooms out there.
I like that hybrid acrylic/wood cabinet approach, the more I think about it. The top portion is sort of an art object, whether you're talking about Lowther/AER, Jordan, or pro-quality coax drivers. These are expensive drivers and look the part - like a good tube amp or a custom motorcycle, you don't want to hide the technological beauty of the working parts.
The wood bass cabinet is interesting as well. I'm thinking of asymmetrical side walls to minimize standing waves. One wall would be short, maybe only 6 inches deep, and the other would be more like 15 inches deep. If there are two drivers, one would go on the front (close to the corner with the deep sidewall) and the other could go on the deep sidewall. Mirror-imaged pairs, of course.
I'm guessing the bass quality might be a little better if the side woofers fired towards the left and right sides of the room, which would improve the sensation of "air" and dimensionality in the bass region. People with stereo subwoofers know what I'm talking about.
Hmm, that's intereresting, swapping the bass cabinets left-to-right would actually swing the dipole pattern around. With the arrangement of the previous paragraph, the nulls of the dipoles are a bit closer to the listener, with side woofers facing the center, the nulls are steered towards the side walls of the room. One more way to tune the system to the room - steering the dipole nulls either towards the center or the side walls.
With a multi-tapped inductor for the bass drivers to control the overlap region, steerable polar patterns, and a pair of dedicated stereo subwoofers for the below-60~80 Hz region, there are a lot of ways to tune the system to the room. This is good, considering the wide variety of rooms out there.
Re: Hmmm - Acrylic
My existing proto baffles are BB in the centre (with removable driver module) and plexi-glass extensions (total ~1.2m x 1.1 m WxH)
Syncronystically, when i was just digging for some boxes to ship out a couple pair of FE12x, i saw & was reminded that i have a pr of unobtanium 3" neo cast-basket Tangbands in just such a baffle. Now if i can get the 4 x 12" push-push ripole modules off the paper and into wood i'd have some bass support to play around with.
Sorta like this -- different top...
dave
My existing proto baffles are BB in the centre (with removable driver module) and plexi-glass extensions (total ~1.2m x 1.1 m WxH)
Lynn Olson said:
Syncronystically, when i was just digging for some boxes to ship out a couple pair of FE12x, i saw & was reminded that i have a pr of unobtanium 3" neo cast-basket Tangbands in just such a baffle. Now if i can get the 4 x 12" push-push ripole modules off the paper and into wood i'd have some bass support to play around with.
Sorta like this -- different top...
dave
Re: Still Thinking ...
if you do something like this, how skiny can you make the front baffle. If you make it really narrow say 10" for a 8" full range speaker, you can use a piece of acrylic to go the entire length of the front baffle (48" maybe), and the two sides (6" and 15") can be veneered wood (I like Zebrawood, bubbinga, but you may like something else)
Obviously, a 15" woofer will not fitted onto a 6" sidewall. But a 20" sidewall will do it.
From the front you will see the 8" woofer and the magnet structures of the bass driver. I think it will look good.
Maybe, Hemp 12 coax on a 48" by 12" front clear baffle.
if you do something like this, how skiny can you make the front baffle. If you make it really narrow say 10" for a 8" full range speaker, you can use a piece of acrylic to go the entire length of the front baffle (48" maybe), and the two sides (6" and 15") can be veneered wood (I like Zebrawood, bubbinga, but you may like something else)
Obviously, a 15" woofer will not fitted onto a 6" sidewall. But a 20" sidewall will do it.
From the front you will see the 8" woofer and the magnet structures of the bass driver. I think it will look good.
Maybe, Hemp 12 coax on a 48" by 12" front clear baffle.
slightly off topic..or not ;
I'm really interested in "JamesD's " OB approach,especially suited for smaller rooms,or at least better suited than Bionor likes....
anyway-just today I tried row of different oval and round AlNiCo Iskra drivers ,all placed on topp of right cab on pic.
that right cab was disconected, and one by one eeny weeny diver was connected instead.
amp for test was veeeery old Pioneer ,around 15W/ch good for two reasons - easy on ear (even if not last word in resolution and neutrality) and more important-with cap in output ,so pretty forgivable for any short shortie with speaker wires
what most impress me is fact that(even if I'm long ago converted to ancient odd spks) all tried radio ,PA column and grammo spks were not much worse than entire cab on left side!!!
even if almost all of them sounds pretty mediocre in almost any sane cab.........
just for record: I also have several more than bad AlNiCo spks.......so-even than bad spk was bad
all this mumbling just in apologetic purpose for OB............any decent spk will sing on OB ,comparing it to situation when is closed in anything worse than perfect cab .........
hehe.....Eminence Alpha 15 is ~ 45 Euros a piece around me ......
I can see already one OB pair (Alpha 15 + some old AlNiCo) in every room
btw-interesting link.....not much theory but..........
http://www.lampizator.eu/SPEAKERS/PROJECTS/Speaker projects.html
I'm really interested in "JamesD's " OB approach,especially suited for smaller rooms,or at least better suited than Bionor likes....
anyway-just today I tried row of different oval and round AlNiCo Iskra drivers ,all placed on topp of right cab on pic.
that right cab was disconected, and one by one eeny weeny diver was connected instead.
amp for test was veeeery old Pioneer ,around 15W/ch good for two reasons - easy on ear (even if not last word in resolution and neutrality) and more important-with cap in output ,so pretty forgivable for any short shortie with speaker wires
what most impress me is fact that(even if I'm long ago converted to ancient odd spks) all tried radio ,PA column and grammo spks were not much worse than entire cab on left side!!!
even if almost all of them sounds pretty mediocre in almost any sane cab.........
just for record: I also have several more than bad AlNiCo spks.......so-even than bad spk was bad
all this mumbling just in apologetic purpose for OB............any decent spk will sing on OB ,comparing it to situation when is closed in anything worse than perfect cab .........
hehe.....Eminence Alpha 15 is ~ 45 Euros a piece around me ......
I can see already one OB pair (Alpha 15 + some old AlNiCo) in every room
btw-interesting link.....not much theory but..........
http://www.lampizator.eu/SPEAKERS/PROJECTS/Speaker projects.html
Hi Lynn,
sorry for hiijacking this thread.
The first thing I want to tell you is that I hope, that you are able to cope with the experiences you made. I had a nephritic holdup some years ago and I do not miss the pain. I did not have an accicent like yours right now - I am lucky. I wish you, that you are able to forget the shocking experience you made and that you are able to take your advantages of the moments you can not forget. May the music and the stuff we are constructing here help you.
Several posts ago you wrote, that a linar array would lead to problems because of the timing - the sonics from the outermost sources of the array reach the ear of the listener later. You proposed a geometry that is tilt towards the listener to avoid this effect.
I Think that you are not right. The effect of a line array is that every driver gets focused in the horizontal plane with the help of the other ones because delays of a half wavelength get canceled (though I have been studying physics for some years, I am not able to explain it a better way with some words - if you want a better eplanation read "QED" from Richard P. Feynman or send me a pm). When you are listening to a line array, you are listening to the speaker that is at the hight of your ears. The problem of a linearray is, that this effect is dependent of the frequency. As a consequence you get this "comb filtering" effect. In normal listening environments this effect is smeared by the room with the consequence of an inaccurancy in the time domain. But you will never hear the sound of the outermost speaker with a delay.
Best regards and best wishes
Floric
sorry for hiijacking this thread.
The first thing I want to tell you is that I hope, that you are able to cope with the experiences you made. I had a nephritic holdup some years ago and I do not miss the pain. I did not have an accicent like yours right now - I am lucky. I wish you, that you are able to forget the shocking experience you made and that you are able to take your advantages of the moments you can not forget. May the music and the stuff we are constructing here help you.
Several posts ago you wrote, that a linar array would lead to problems because of the timing - the sonics from the outermost sources of the array reach the ear of the listener later. You proposed a geometry that is tilt towards the listener to avoid this effect.
I Think that you are not right. The effect of a line array is that every driver gets focused in the horizontal plane with the help of the other ones because delays of a half wavelength get canceled (though I have been studying physics for some years, I am not able to explain it a better way with some words - if you want a better eplanation read "QED" from Richard P. Feynman or send me a pm). When you are listening to a line array, you are listening to the speaker that is at the hight of your ears. The problem of a linearray is, that this effect is dependent of the frequency. As a consequence you get this "comb filtering" effect. In normal listening environments this effect is smeared by the room with the consequence of an inaccurancy in the time domain. But you will never hear the sound of the outermost speaker with a delay.
Best regards and best wishes
Floric
Zen Mod, that link is to the "Kingston" guy. He is from Poland I think, had a great website a while back with about half as many projects. At that point he had transitioned to open baffles for his wide mid range drivers with a ribbon tweet He was a strong influence on me and made a lot of sense. Probably made me decide to do the Basszillas ... that have open baffle mid and ribbon highs.
I still have text from that site and it is very informative. Too bad he doesn't include it now. Maybe I could ask to post it here..
Now he seems to have go all the way with open baffles and also uses vintage drivers... Planet 10 will approve of that! Makes them hard to clone though.
I still have text from that site and it is very informative. Too bad he doesn't include it now. Maybe I could ask to post it here..
Now he seems to have go all the way with open baffles and also uses vintage drivers... Planet 10 will approve of that! Makes them hard to clone though.
Lynn Olson said:
Unfortunately, or fortunately, my body blocked most of the pain with shock (or whatever it is that the body does). I can remember getting up off the road and wonderring why only one of my arms would come up to my helmet. By the time my teeth started to grind (my body's way of telling me that I was feeling pain, I assume) I was already in the ambulance and on NO.
But, let's stop torturing ourselves and the others reading this thread.
Dave, interesting idea of yours (or is it already in practice), which reminds me a bit of Linkwitz' Orion in execution, but in a "it's taken steroids" fashion. Are those the FE12x's in the top?
Cloth Ears said:
It could be... that drawing was actually done for the Lyeco LY401F, and the previous incarnation, an FE108eS + AC ribbon (which i have). For that matter scottmoose won the Lyeco contest and the 401 is headed here 1st.
I have at least a dozen of the vintage Foster 12" woofers (was worth buying the boxes just to sell the tweeters) and they turn out to have T/S right in the sweet spot for the ripoles.
dave
Got a link? I searched a bit on the WD forum and couldn't seem to find the specifics on the xover.Scottmoose said:
Will watch this thread with interest. I have 8 Beta 15's NIB and have been trying various ideas for the >150Hz of so region as thought exercises only atm. The Hemp 15 coaxes or the TT12 and a suitable HF driver seem like good possiblilities.
I don't visit much any more, but thanks Lynn as your Amity article was what inspired me to move back into DIY. I've now built several versions, as well as Ariels and my business partner has the drivers for a pair now and is itching to start the build.
Best of luck with the leg Lynn. Your story of your trip up the drive brought back memories of a drive I had to make with 4 fractured vertebrae. 21 years later with full mobility, it brings a smile and a cringe at the same time.
Problems with Arrays
I look at it a little differently. Let's assume an array of 2" Jordans - the classic array that converts an ultra-low efficiency driver (82 dB/metre if I recall right) into something a lot more reasonable - with, say, 8 or 12 drivers.
These drivers, due to their small size, are pretty close to hemispherical radiators up to about 10 kHz or so. Now imagine feeding the entire array with a narrow pulse. With one driver, you'll get a very clean return at the 2-meter microphone position, limited only by the energy storage of the driver itself (I am assuming a perfect enclosure here with no diffraction or internal energy storage).
But - with 8 or 12 drivers in a vertical array, the arrival times at the microphone must be non-coincident due to different time-of-flight for each impulse. And this is exactly what you measure with large vertical arrays, whether composed of discrete dynamic drivers or large electrostatic or magnetic-planar panels. This happens with theoretically perfect drivers of zero size - the time dispersion (time smearing) with real-world drivers is worse, since the off-axis HF rolloff of the topmost and bottom-most drivers results in the acoustic centers moving backwards in space, which worsens the time dispersion for the most severely off-axis drivers.
(Anything that rolls off the HF response of the driver, whether electrical, mechanical, or dispersion limiting, always results in the center of radiation moving backward in space. This is why woofers have acoustical centers that are farther back than you might expect, thanks to the crossover lowpass filtering and the natural rolloff of the woofer.)
The greatest number of drivers that will give a synchronous arrival time (on a flat panel) is no more than two. Any more than that, you get non-synchronous arrival times, comb filtering in the frequency domain, and multiple lobes in the vertical polar pattern. In reality, all three domains are simply reflections of what is happening in the time domain (the real world).
The same thing appears in antenna theory - the most directional antennas have the narrowest frequency response, the worst pulse response, and the greatest number of sidelobes. Same for microphones, too - shotgun or parabolic microphones are a long way from hifi devices, and only really useful for surveillance, birdwatching, or newsgathering.
The simplest example of a vertical array, the MTM, are tricky to design thanks to very complicated vertical polar patterns in the crossover region. I found that out the hard way with the Ariel, where the measurements did not coincide with subjective impressions of frequency response. As it was, I compromised about halfway between direct-arrival response and total room response (energy into a sphere).
Which brings me to the other issue with vertical arrays - the power does not drop off according to the square law, thanks to the very narrow vertical dispersion. This means the acoustical joining to a radiator with spherical dispersion, such as a subwoofer or supertweeter, requires a listening-distance compensation. A system that is equalized flat at 2 meters will have excess HF energy at 4 meters, since the SPL for the array falls off less quickly with distance than the spherical radiator.
The same thing happens with horns joined to direct-radiator bass sections. The bass unit is omnidirectional, but the HF horn dispersion is typically 90 degrees or less. This means a system compensated for 2 meters will have too much HF energy at 4 meters, since the horn, like the vertical array, is not dropping off with distance as fast as the omnidirectional radiator.
I think you can see where this line of argument is going - one advantage of a coax dipole is a polar pattern that is fairly constant with frequency, thus minimizing the requirement for listening-distance compensation. The de facto polar pattern is about 120 to 90 degrees over most of the frequency range, and with reasonable care in design, free of the narrow vertical sidelobes of vertical arrays, MTM's, or conventional horns.
Floric said:
I look at it a little differently. Let's assume an array of 2" Jordans - the classic array that converts an ultra-low efficiency driver (82 dB/metre if I recall right) into something a lot more reasonable - with, say, 8 or 12 drivers.
These drivers, due to their small size, are pretty close to hemispherical radiators up to about 10 kHz or so. Now imagine feeding the entire array with a narrow pulse. With one driver, you'll get a very clean return at the 2-meter microphone position, limited only by the energy storage of the driver itself (I am assuming a perfect enclosure here with no diffraction or internal energy storage).
But - with 8 or 12 drivers in a vertical array, the arrival times at the microphone must be non-coincident due to different time-of-flight for each impulse. And this is exactly what you measure with large vertical arrays, whether composed of discrete dynamic drivers or large electrostatic or magnetic-planar panels. This happens with theoretically perfect drivers of zero size - the time dispersion (time smearing) with real-world drivers is worse, since the off-axis HF rolloff of the topmost and bottom-most drivers results in the acoustic centers moving backwards in space, which worsens the time dispersion for the most severely off-axis drivers.
(Anything that rolls off the HF response of the driver, whether electrical, mechanical, or dispersion limiting, always results in the center of radiation moving backward in space. This is why woofers have acoustical centers that are farther back than you might expect, thanks to the crossover lowpass filtering and the natural rolloff of the woofer.)
The greatest number of drivers that will give a synchronous arrival time (on a flat panel) is no more than two. Any more than that, you get non-synchronous arrival times, comb filtering in the frequency domain, and multiple lobes in the vertical polar pattern. In reality, all three domains are simply reflections of what is happening in the time domain (the real world).
The same thing appears in antenna theory - the most directional antennas have the narrowest frequency response, the worst pulse response, and the greatest number of sidelobes. Same for microphones, too - shotgun or parabolic microphones are a long way from hifi devices, and only really useful for surveillance, birdwatching, or newsgathering.
The simplest example of a vertical array, the MTM, are tricky to design thanks to very complicated vertical polar patterns in the crossover region. I found that out the hard way with the Ariel, where the measurements did not coincide with subjective impressions of frequency response. As it was, I compromised about halfway between direct-arrival response and total room response (energy into a sphere).
Which brings me to the other issue with vertical arrays - the power does not drop off according to the square law, thanks to the very narrow vertical dispersion. This means the acoustical joining to a radiator with spherical dispersion, such as a subwoofer or supertweeter, requires a listening-distance compensation. A system that is equalized flat at 2 meters will have excess HF energy at 4 meters, since the SPL for the array falls off less quickly with distance than the spherical radiator.
The same thing happens with horns joined to direct-radiator bass sections. The bass unit is omnidirectional, but the HF horn dispersion is typically 90 degrees or less. This means a system compensated for 2 meters will have too much HF energy at 4 meters, since the horn, like the vertical array, is not dropping off with distance as fast as the omnidirectional radiator.
I think you can see where this line of argument is going - one advantage of a coax dipole is a polar pattern that is fairly constant with frequency, thus minimizing the requirement for listening-distance compensation. The de facto polar pattern is about 120 to 90 degrees over most of the frequency range, and with reasonable care in design, free of the narrow vertical sidelobes of vertical arrays, MTM's, or conventional horns.