line array project
Jim, John, and Planet 10, thanks for the feedback-my situation is that I have a large record collection and I don't want it to run through A to D and D to A conversions so a passive crossover to start would be my choice. Later I could add the Digital crossovers and have both. I do listen to a large collection of SACD's and red books.
I took your suggestions and talked to creative sound about the WR125's and I am waiting for GR Research to finish testing his their new 6" mid bass. I also talked to CSS about the 5" Eton hexacone driver for use as a mid bass. I have heard this driver in the buggtussel line and I talked to Dr. Kevin Blair who designed that speaker line. He told me that I should be able to cross over to the Eton between 50 and 80 hz on the low end with my subs.
Dr. Blair also advised to stick with the smaller 5" drivers due to dispersion off axis at the crossover frequency.
I have gotten a good price for the Fountex ribbons from www.meniscusaudio.com and would use 10-6" drivers. Based on Dr. Griffin's paper and other hobbyist info, I am now trying to find a mid bass match that I can crossover to the ribbons and meet the line array design criteria but still get decent bass in the 40-50hz range and then crossover to the subs at 12 or 24db.
CSS advised that I should be able to squeeze a bit more db out of the eatons vs the WR125's but I may sacrifice a bit of midrange clarity. This is the spot I am in right now.
I have another friend who is horrified that I would crossover at 1.5 to 2.5k thinking that this is in the midrange and will cause design issues.
I have also read Selahs website and they use ribbons with a 7" metal cone driver, I do not know where they cross them over.
To minimize my risk based on my abilitities I am trying to pick drivers that I am familiar with hearing (Maggie ribbons and etons)
realizing that there may be better choices. thanks for any futher opinions.
I have ruled out the Silver Flute line but they are cost effective. dave
Jim, John, and Planet 10, thanks for the feedback-my situation is that I have a large record collection and I don't want it to run through A to D and D to A conversions so a passive crossover to start would be my choice. Later I could add the Digital crossovers and have both. I do listen to a large collection of SACD's and red books.
I took your suggestions and talked to creative sound about the WR125's and I am waiting for GR Research to finish testing his their new 6" mid bass. I also talked to CSS about the 5" Eton hexacone driver for use as a mid bass. I have heard this driver in the buggtussel line and I talked to Dr. Kevin Blair who designed that speaker line. He told me that I should be able to cross over to the Eton between 50 and 80 hz on the low end with my subs.
Dr. Blair also advised to stick with the smaller 5" drivers due to dispersion off axis at the crossover frequency.
I have gotten a good price for the Fountex ribbons from www.meniscusaudio.com and would use 10-6" drivers. Based on Dr. Griffin's paper and other hobbyist info, I am now trying to find a mid bass match that I can crossover to the ribbons and meet the line array design criteria but still get decent bass in the 40-50hz range and then crossover to the subs at 12 or 24db.
CSS advised that I should be able to squeeze a bit more db out of the eatons vs the WR125's but I may sacrifice a bit of midrange clarity. This is the spot I am in right now.
I have another friend who is horrified that I would crossover at 1.5 to 2.5k thinking that this is in the midrange and will cause design issues.
I have also read Selahs website and they use ribbons with a 7" metal cone driver, I do not know where they cross them over.
To minimize my risk based on my abilitities I am trying to pick drivers that I am familiar with hearing (Maggie ribbons and etons)
realizing that there may be better choices. thanks for any futher opinions.
I have ruled out the Silver Flute line but they are cost effective. dave
line array speaker project
Dr. Griffin, I have read your paper on line arrays, but I would be interested in what your current thinking is on the digital crossover. I was intending to do 24db in both directions and I like your idea of the super high crossover. What make and model of a digital system would you recommend (I have noticed that Behringer is popular on this site). I have plenty of home built class A amps to choose from and I was going to bi-amp anyway.
I may need level controls and what do you do about hot spots. thanks for your time. dave
Dr. Griffin, I have read your paper on line arrays, but I would be interested in what your current thinking is on the digital crossover. I was intending to do 24db in both directions and I like your idea of the super high crossover. What make and model of a digital system would you recommend (I have noticed that Behringer is popular on this site). I have plenty of home built class A amps to choose from and I was going to bi-amp anyway.
I may need level controls and what do you do about hot spots. thanks for your time. dave
Your problems are easily solved if you use an 8th order LR
crossover slope on problematic midwoofers and you can
take the ribbons lower in frequency. For more output at
40hz, find a midwoofer that lends itself well to porting.
If you don't like ported midwoofers, then find a good midwoofer
that has nice response down to 70-80hz and let use your
subwoofer do the work from here.
crossover slope on problematic midwoofers and you can
take the ribbons lower in frequency. For more output at
40hz, find a midwoofer that lends itself well to porting.
If you don't like ported midwoofers, then find a good midwoofer
that has nice response down to 70-80hz and let use your
subwoofer do the work from here.
thylantyr--advise well taken, I am going to use a ported design and I want to crossover steep for the reasons that you have given. I plan to start with a theorectical crossover and then test and measure and tweak from there. In my old age I am getting lazy and it is always easier to twist a dial then plug in values manually, but I will have the time. If you have any LR crossover design literature that you can share please advise. I have read the white papers on Rane Audio's web site for active LR -8th order networks and many of the formulas are there. I also have Walt Jung's op-amp cookbook which has interesting information.
I have ordered speaker building 201 which should have the passive formula's as well. dave
I have ordered speaker building 201 which should have the passive formula's as well. dave
Daly2K,
Let me try to give you some thoughts.
I favor the smaller 4 to 5.5" diameter drivers as you can get some mid-bass from them and then crossover around a wavelength c-t-c spacing. My experience with the woofer line in a line array is that the measured upper end performance indicates that the gain (or sensitivity) shows roll off even before one wavelength center to center spacing occurs. Hence, for example with a 4.5" driver this means that you'll likely have to cross over at or even before you reach 2000 Hz. The thought that you could reach much above a wavelength c-t-c spacing for your crossover to tweeters is just wishful thinking given the gain roll off of the mid-woofer line.
I prefer sealed boxes for my mid-woofers--better transient response and less energy storage--mean a snapper response. The down side is that you have a 3 dB rolloff on the low end in the 70-80 Hz area. Remember that a sealed box rolls off at a 12 dB/octave rate. You can go lower with a ported enclosure but I favor subwoofers to reach into the lower 1.5 octaves (30-80 Hz). Subs can do so much better covering down low vs. trying to over extend a small mid-woofer. Now you could equalize a line of small sealed mid-woofers to cover lower in frequency (say about 40 Hz) and be successful unless you have a large listening room and insist on high SPL listening.
Active crossovers are a very good way to go for a line array or any speaker system for that matter. Once you get beyond the cost of the crossover (a Behringer 2496 DCX is only $250) and the extra amps to bi or triamp, then you minimize the turnaround time to work out a passive crossover. With either a passive or active crossover you will need frequency measurement capability to work through the details. You can easily implement an eight order crossover (or a 16th order) in most active crossovers. You spend about as much money and much more time to tweak a passive crossover--consider the iterations and tiem to order and reorder capacitors and inductors that you need to home in on the right values. Finally, with a passive crossover you'll have to use a crossover optimization program while with an active crossover you can tweak and measure in close to real time.
BTW a 9 ribbon line (assuming the Fountek NeoCD2.0's) evenly fed is what I would recommend in your case. Don't taper the tweeter line.
In my new array I'm using 12 CSS WR125S per side crossed over to 9 Aurum Cantus G3i-130 ribbons at 1750 Hz. Neither line is tapered but fed in series groups of 3 each before paralleling. The overall impedance of the 12 WR's is 6 ohms in the array while the 9 G3i's are also 6 ohms in the array. The DEQX crossover slopes are set at 96 dB/octave. On the low end I second order cross (high pass) the mid-woofers at 100 Hz to the subs with a fourth order low pass at 100 Hz for the upper end of the subs.
One thing that is clear is that the ability to correct for room issues is critical to achieve good in-room performance with a line array or for that matter almost any speaker system. My new array is calibrated by the DEQX to ideal performance--you can measure and verify the performance--but in-room for a normal small listening room the sound is unlistenable unless can integrate the bass and cover some upper frequency peaks.
The Behringer DCX can help you with parametric filtering to accommodate bass roughness. and then smooth out the upper frequency peaks. A even better solution is to use the Behringer 2496 DEQ but you can get started with the DCX.
Good luck and keep thinking.
Jim
Let me try to give you some thoughts.
I favor the smaller 4 to 5.5" diameter drivers as you can get some mid-bass from them and then crossover around a wavelength c-t-c spacing. My experience with the woofer line in a line array is that the measured upper end performance indicates that the gain (or sensitivity) shows roll off even before one wavelength center to center spacing occurs. Hence, for example with a 4.5" driver this means that you'll likely have to cross over at or even before you reach 2000 Hz. The thought that you could reach much above a wavelength c-t-c spacing for your crossover to tweeters is just wishful thinking given the gain roll off of the mid-woofer line.
I prefer sealed boxes for my mid-woofers--better transient response and less energy storage--mean a snapper response. The down side is that you have a 3 dB rolloff on the low end in the 70-80 Hz area. Remember that a sealed box rolls off at a 12 dB/octave rate. You can go lower with a ported enclosure but I favor subwoofers to reach into the lower 1.5 octaves (30-80 Hz). Subs can do so much better covering down low vs. trying to over extend a small mid-woofer. Now you could equalize a line of small sealed mid-woofers to cover lower in frequency (say about 40 Hz) and be successful unless you have a large listening room and insist on high SPL listening.
Active crossovers are a very good way to go for a line array or any speaker system for that matter. Once you get beyond the cost of the crossover (a Behringer 2496 DCX is only $250) and the extra amps to bi or triamp, then you minimize the turnaround time to work out a passive crossover. With either a passive or active crossover you will need frequency measurement capability to work through the details. You can easily implement an eight order crossover (or a 16th order) in most active crossovers. You spend about as much money and much more time to tweak a passive crossover--consider the iterations and tiem to order and reorder capacitors and inductors that you need to home in on the right values. Finally, with a passive crossover you'll have to use a crossover optimization program while with an active crossover you can tweak and measure in close to real time.
BTW a 9 ribbon line (assuming the Fountek NeoCD2.0's) evenly fed is what I would recommend in your case. Don't taper the tweeter line.
In my new array I'm using 12 CSS WR125S per side crossed over to 9 Aurum Cantus G3i-130 ribbons at 1750 Hz. Neither line is tapered but fed in series groups of 3 each before paralleling. The overall impedance of the 12 WR's is 6 ohms in the array while the 9 G3i's are also 6 ohms in the array. The DEQX crossover slopes are set at 96 dB/octave. On the low end I second order cross (high pass) the mid-woofers at 100 Hz to the subs with a fourth order low pass at 100 Hz for the upper end of the subs.
One thing that is clear is that the ability to correct for room issues is critical to achieve good in-room performance with a line array or for that matter almost any speaker system. My new array is calibrated by the DEQX to ideal performance--you can measure and verify the performance--but in-room for a normal small listening room the sound is unlistenable unless can integrate the bass and cover some upper frequency peaks.
The Behringer DCX can help you with parametric filtering to accommodate bass roughness. and then smooth out the upper frequency peaks. A even better solution is to use the Behringer 2496 DEQ but you can get started with the DCX.
Good luck and keep thinking.
Jim
Hi Jim!
I'd like to chime in with an unsolicited testimonial. I had the opportunity to hear Jim's line arrays recently @ DIY Dayton. I believe they would hold their own against any speaker I've heard. Thanks Jim. Your creations are always a treat!
Jim's speakers are also one of the few I've seen to include REAL hardwood. Gotta luv em!
I'd like to chime in with an unsolicited testimonial. I had the opportunity to hear Jim's line arrays recently @ DIY Dayton. I believe they would hold their own against any speaker I've heard. Thanks Jim. Your creations are always a treat!
Jim's speakers are also one of the few I've seen to include REAL hardwood. Gotta luv em!
line array speaker project
Jim and BQ-thanks for the feed back--Jim I have been researching suppliers of the WR's and will see what I can do. Regarding the low bass, I am lucky that my two buggtussel subs are very versatile for integration with the main speakers. They have an adjustable slope 12 or 24db, a phase control, an adjustable turnover frequency from 35 to 200hz and two parametric notch controls for "help" in smoothing room peaks and dips. I believe I can get them to blend with the line arrays. I am using them for my maggie 3.6r's set at 50hz and they seem to keep up. They have built in 300w/ch amps. I was not contemplating to taper or equalize the line array. There was an interesting article in this months Audioxpress where the author used just one mid range driver and equalized it to get low bass and more treble. He used 5" drivers on close out from Partsexpress ($200). There was a past article written over a year ago about another builder of a line array using close out tweeters and midranges in a TL-Line array and he built it for $60. Neither one of these authors tapered their line arrays.
Now that I understand your design criteria better, what do you do about speaker protection. Blowing out 12 drivers because of a mistake or a leaky cap won't do. My amps have fuse protection for short circuits but the speakers would not be DC protected if I didn't use a passive crossover. I was thinking of some simple input fuses. I could also design a speaker relay protection circuit for my amps (relay would either open or close in the event of dc)
to remove the signal if dc is present.
Jim and BQ-thanks for the feed back--Jim I have been researching suppliers of the WR's and will see what I can do. Regarding the low bass, I am lucky that my two buggtussel subs are very versatile for integration with the main speakers. They have an adjustable slope 12 or 24db, a phase control, an adjustable turnover frequency from 35 to 200hz and two parametric notch controls for "help" in smoothing room peaks and dips. I believe I can get them to blend with the line arrays. I am using them for my maggie 3.6r's set at 50hz and they seem to keep up. They have built in 300w/ch amps. I was not contemplating to taper or equalize the line array. There was an interesting article in this months Audioxpress where the author used just one mid range driver and equalized it to get low bass and more treble. He used 5" drivers on close out from Partsexpress ($200). There was a past article written over a year ago about another builder of a line array using close out tweeters and midranges in a TL-Line array and he built it for $60. Neither one of these authors tapered their line arrays.
Now that I understand your design criteria better, what do you do about speaker protection. Blowing out 12 drivers because of a mistake or a leaky cap won't do. My amps have fuse protection for short circuits but the speakers would not be DC protected if I didn't use a passive crossover. I was thinking of some simple input fuses. I could also design a speaker relay protection circuit for my amps (relay would either open or close in the event of dc)
to remove the signal if dc is present.
Re: line array speaker project
Hi daly2K,
I would be very interrested in reading this article. Could you provide more details where to find it. Is it available online? If not where could I get ahold of it?
Thanks,
Exipnos
daly2k said:
Hi daly2K,
I would be very interrested in reading this article. Could you provide more details where to find it. Is it available online? If not where could I get ahold of it?
Thanks,
Exipnos
Daly,
I'd strongly suggest leaving the decision to power taper or not open until you listen. While I like arrays both ways depending on the application. Equal power gives you a bigger sound but can lead the audio image to be distorted in size. eg Barry White's voice makes him sound like his head is 4ft in diameter. The appeal that they put you front row center stage more than any speaker I've heard. With power tapering you still get a much "bigger" sound that a point source speaker with more accurate imaging.
I'd strongly suggest leaving the decision to power taper or not open until you listen. While I like arrays both ways depending on the application. Equal power gives you a bigger sound but can lead the audio image to be distorted in size. eg Barry White's voice makes him sound like his head is 4ft in diameter. The appeal that they put you front row center stage more than any speaker I've heard. With power tapering you still get a much "bigger" sound that a point source speaker with more accurate imaging.
line array speaker project
exipnos, the articles that you want to read were in www.audioxpress.com mag. They are in the April 2005 and November 2005 editions.
johnincr--I do not intend to power taper at first, I want to eventually get to a realistic soundstage similiar to my maggie 3.6r's. I agree with you suggestion to start. thanks to you both. dave
exipnos, the articles that you want to read were in www.audioxpress.com mag. They are in the April 2005 and November 2005 editions.
johnincr--I do not intend to power taper at first, I want to eventually get to a realistic soundstage similiar to my maggie 3.6r's. I agree with you suggestion to start. thanks to you both. dave
I'm considering a line array like Jim's for the three front speakers of a home theater setup (... there's a set on Raw Acoustics that look very promising: http://www.rawacoustics.com/ ... the Omega line arrays) I love the "big clean sound" of well designed arrays.
One issue I'd appreciate some advice on ...
I'm sensitive to "height" differences across the front channels. I've currently got "tower" lefts and rights, and a "matching" center with the same driver, tweeter, and crossover configuration (four 7" Seas mids and one Seas millenium per ported unit). When sounds pan left to right I hear a marked difference: image height and tonal balance shift quite a bit. I also notice a discontinuity when sounds are shared between a L or R and the Center channel. (The current horizontal axis mis match is approx 1 foot. Center channel is 3" below ear height.)
Wondering how a vertical line array for the Left / Right channels would integrate with a horizontal array with a single ribbon for the Center?
Does the use of a line array configuration make my height issues less of an issue or more?
Would I get better results if I could have the vertical center of the left and right arrays centered with the center channel (that is, us a shorter array for the left and rights) or does that not matter?
My alternative is three matching speakers across the fronts, but if I do that I can't use line arrays because of construction constraints with the screen placement.
One issue I'd appreciate some advice on ...
I'm sensitive to "height" differences across the front channels. I've currently got "tower" lefts and rights, and a "matching" center with the same driver, tweeter, and crossover configuration (four 7" Seas mids and one Seas millenium per ported unit). When sounds pan left to right I hear a marked difference: image height and tonal balance shift quite a bit. I also notice a discontinuity when sounds are shared between a L or R and the Center channel. (The current horizontal axis mis match is approx 1 foot. Center channel is 3" below ear height.)
Wondering how a vertical line array for the Left / Right channels would integrate with a horizontal array with a single ribbon for the Center?
Does the use of a line array configuration make my height issues less of an issue or more?
Would I get better results if I could have the vertical center of the left and right arrays centered with the center channel (that is, us a shorter array for the left and rights) or does that not matter?
My alternative is three matching speakers across the fronts, but if I do that I can't use line arrays because of construction constraints with the screen placement.
Just forget about the center channel. Arrays can easily handle the normal info for the mains and split the center channel info. You'll get better panning and a big audio image to match you large video image. IMHO an array pair is the one best match for a projected image.
The conundrum is that the off center listeners in the room really benefit from the center channel ...
Even if only in theory, is a line array more, less, or same subject to "center channel height differences" than more traditional box speakers?
Even if only in theory, is a line array more, less, or same subject to "center channel height differences" than more traditional box speakers?
brucemck2 said:
Bruce,
Off center listeners benefit greatly from line arrays vs point source speakers. Line arrays don't work well on their sides because you get very limited dispersion outside of the plane of the array.
With very large images, the general problem with a center channel is that it needs to be above or below the screen and the sound comes from too high or too low in relation to the image. The only fix is a supposedly "sonically transparent" screen. Then you could use a vertical array as your center, but the compromise is a less than ideal screen material for video.
The only compromise I find with using arrays as mains with no center channel is that you need to place the arrays closer to the screen than you might otherwise want. With some toe-in of the arrays, you'd have to sit well out past the speakers left or right and even then you are likely to be able to still hear both speakers. The area of prime listening for line arrays is magnitudes larger than for point source speakers because distance only has half the effect on SPL.
I built a horizontal center with a single tweeter to match a pair of arrays. It worked quite well. One thing to remember though is that the woofer array should go nearly the entire seated distance because you will have very limited horizontal dispersion in this array. You will have very good vertical dispersion though. One other thing to consider is that you will need to optimize the tweeter level for the listening distance since it will roll off much faster than the woofer array. So if you have people sitting at much different distances it could be a problem.
To all amigos, I am in construction, I chose the fountex 5" ribbons for the tweeters-12 per side and 12 - 5" eaton mid-bass drivers. Tests show that 1800hz-24db crossover was just about ideal. We first (working with Meniscus Audio in Grand Rapids) built a test encloseure out of MDF to test the crossover design (simulated on computer based on listening distance and trying to keep all frequencies in the near field) with the actual drivers mounted. Meniscus will keep this encloseure for future projects. This also allowed us to make mistakes before cutting any of the veneered walnut MDF.
We are using the politically correct wiring which is silver coater Ox free teflon coated - 14 gauge and gold platted connectors. The crossover is mounted externally in a matched veneer box. This will allow me to use a digital crossover in the future when I can afford it.
I expect to have the project done mid month and will post pictures. dave
We are using the politically correct wiring which is silver coater Ox free teflon coated - 14 gauge and gold platted connectors. The crossover is mounted externally in a matched veneer box. This will allow me to use a digital crossover in the future when I can afford it.
I expect to have the project done mid month and will post pictures. dave
For those who have constructed line arrays using the CSS WR125ST's, has anyone constructed their enclosure to be a stacked array of aperiodic enclosures, a la the box design on the Planet-10 site? Each unit volume could be kept to about 4 litres.
???
Doug
???
Doug
Last edited:
- Status
- Not open for further replies.