I’m currently in the process of figuring out the replacement/upgrade for my living room home theater setup.
The room: I’ve got a 9 ft wide screen I’m projecting onto, hanging from the ceiling. The screen is about three feet from the back wall, and directly above a halfwall about 3.5’ high (there’s a stairway heading downstairs behind the screen). Acoustically, the room is really, really bad. Vaulted ceiling (peak perpendicular to line of sight to screen, thankfully), open to kitchen and dining room to left, open hallway to bedrooms and bath on right, huge volume to fill, assymetrical reflecting surface profile. The living room area is about 15’ x 25’; the back wall has a fireplace in one corner and a baby grand piano in the other, so the seats are ahead of those. Currently, I’ve got a set of Acoustic Research 312HO’s as mains, with an AR center and little bookshelf AR’s as surrounds. There’s suckouts in the room and the bass has serious nulls in various places and the waf of room treatments isn’t high, but I’m working on that. (luckily she wasn’t too upset with the 7 foot tall 24” diameter subwoofer…but she doesn’t like her miniature piano collection flying off the shelves in her music studio several doors down because of the vibrations…that kinda annoyed her when they broke).
The project concept: I want to have 7 identical speakers (sonically anyway) in the living room, and they don’t have to have extension below 70hz because they’ll be crossed over into the sub.
Some of the constraints include:
1. Form Factor: there’s about 12 inches below the screen and above the halfwall, and it’s right around ear height when seated for a movie. The part of the speaker with the drivers needs to fit in that space.
2. Output: Lots of it. Design goal is to get to 110db cleanly. I like dynamics. I like something that sounds good low AND high. The AR’s were good for that, with a high sensitivity, but the towers just don’t fit into the layout for the room.
3. Cost: Like most folks, I’m not made of money. So, >$1000 per speaker is out. I’m trying to get under $300 per speaker if possible for materials, not including amplification.
4. Clarity: I’m looking to use drivers that are quick, and to minimize anything that’ll make them muddy
5. Voice intelligibility: I’ve heard that crossing over in the 300hz to 3600hz range is to be avoided in a setup that does a lot of HT, however, it’s tough to do a 2way without violating that rule. And for the #3 criteria, a 2way is more attractive.
6. Dispersion/Sweet Spot: I’m looking to have excellent off axis response so that more than one person can have excellent imaging for movies and multi channel music. This also makes it tough to do a 2way.
7. Phase coherency: A must. I want this acoustic lens to be sharp.
Current direction for the design: Unless I go with three centers across the front, it’s almost impossible to satisfy the criteria stated above, and using centers on the walls to the side and rear is impractical. So, I need to roll my own, as it were. I’ve looked and looked and looked at various DIY designs, and one thing has become apparent…it’s silly to go with passive crossovers. With the Behringer DCX2496 as inexpensive as it is, active electronics is the best fit to the criteria. So…we’re going active. I’m planning on some fairly nonstandard enclosures as well.
1. Active crossover: Using multiple DCX2496’s for the seven speakers. Either 3 if 2ways, or 4 if 3way design is chosen. This will buy me phase coherency, time alignment, 48db/octave slopes, static and dynamic equalization, EQ profiles for various recordings, and the ability to iterate changes to the crossover to decide on best voicing. Downside is massive amount of time spent dinking with it, and the learning curve for using it. Cost is about a wash; good crossovers (passive) are similar in their cost per speaker especially if going with a 2way setup, where I can use one DCX per three speakers. If I’m able to implement one of my secondary goals (keeping the signal digital from source to amplifier…hard to do inexpensively) then it’s only 2 speakers per DCX, as that unit has one stereo AES input. Also, this solution will enable me to get a fairly flat FR in room, instead of just anechoic, which is important given that my room isn’t ideal by any means…
2. Enclosures: 12” diameter, probably 18” long (driven by driver requirements), and haven’t decided between sealed or vented. Won’t tolerate large group delay numbers, and will likely go sealed, but not locked into that yet. The cylindrical enclosure will be capped by wood on either end, and will have a tube either coming out the bottom with a 90deg mandrel bend the diameter of the woofer or normal to the surface of the tube. Tweeter (and mid, if 3way) will be mounted on a ring faceplate around the tube holding the woofer and be rotateable to accommodate various orientations of the main enclosure tube. I’ll model it in Pro/E and put up pics if anyone’s interested and can offer criticisms. Risks of standing waves will be mitigated. Material for tube will either be aluminum, pvc, cardboard (sonotube) or carbon fiber…haven’t looked into resonance yet.
3. Amplification: External, likely several 7 channel amps
Questions….man, I’ve got tons of them as I work through this exercise. Here’s a couple I’ve got right now that I’d love some feedback on:
1. Driver selection and orientation- I’ve looked at Peerless, Vifa, Hiquophon, Dayton, etc., etc. If there’s a two way setup that’ll work well, then that’d be my first choice. I just don’t know if that’ll work well…I need to settle that aspect before I can move on. And do I do dual woofers with the tweet in a 2way? Can I get away with a single woofer? Where do I put the crossover point….with the steep filtering the DCX provides, can I get away with putting that crossover point in the critical human voice range and not suffer deleterious effects? How about some open ended discussion on this?
2. Amplifier selection-I won’t need an uber expensive amplifier rack, as the load will be resistive due to the use of the active crossover. I’d love the LPA-1 from Emotiva if the last amp wasn’t split. Grr. For the tweets, I was thinking of something like a Panny receiver run in direct mode; they won’t need a ton of current to drive to high volumes cleanly. What are some suggestions for 7 channel amplification for low cost, given the modest requirements?
Once I settle on the drivers and layout, I can proceed with the rest of the design. I welcome any discussion on it; whether it’s in relation to the two questions posted, or anything else in the project described above. Thanks!
I may post this in several of the forums dedicated to DIY speaker design as well, so don’t be surprised if you see this more than once. Just trying to get some feedback and see if there's more I ought to consider from folks that have done this before...thanks for the help!
The room: I’ve got a 9 ft wide screen I’m projecting onto, hanging from the ceiling. The screen is about three feet from the back wall, and directly above a halfwall about 3.5’ high (there’s a stairway heading downstairs behind the screen). Acoustically, the room is really, really bad. Vaulted ceiling (peak perpendicular to line of sight to screen, thankfully), open to kitchen and dining room to left, open hallway to bedrooms and bath on right, huge volume to fill, assymetrical reflecting surface profile. The living room area is about 15’ x 25’; the back wall has a fireplace in one corner and a baby grand piano in the other, so the seats are ahead of those. Currently, I’ve got a set of Acoustic Research 312HO’s as mains, with an AR center and little bookshelf AR’s as surrounds. There’s suckouts in the room and the bass has serious nulls in various places and the waf of room treatments isn’t high, but I’m working on that. (luckily she wasn’t too upset with the 7 foot tall 24” diameter subwoofer…but she doesn’t like her miniature piano collection flying off the shelves in her music studio several doors down because of the vibrations…that kinda annoyed her when they broke).
The project concept: I want to have 7 identical speakers (sonically anyway) in the living room, and they don’t have to have extension below 70hz because they’ll be crossed over into the sub.
Some of the constraints include:
1. Form Factor: there’s about 12 inches below the screen and above the halfwall, and it’s right around ear height when seated for a movie. The part of the speaker with the drivers needs to fit in that space.
2. Output: Lots of it. Design goal is to get to 110db cleanly. I like dynamics. I like something that sounds good low AND high. The AR’s were good for that, with a high sensitivity, but the towers just don’t fit into the layout for the room.
3. Cost: Like most folks, I’m not made of money. So, >$1000 per speaker is out. I’m trying to get under $300 per speaker if possible for materials, not including amplification.
4. Clarity: I’m looking to use drivers that are quick, and to minimize anything that’ll make them muddy
5. Voice intelligibility: I’ve heard that crossing over in the 300hz to 3600hz range is to be avoided in a setup that does a lot of HT, however, it’s tough to do a 2way without violating that rule. And for the #3 criteria, a 2way is more attractive.
6. Dispersion/Sweet Spot: I’m looking to have excellent off axis response so that more than one person can have excellent imaging for movies and multi channel music. This also makes it tough to do a 2way.
7. Phase coherency: A must. I want this acoustic lens to be sharp.
Current direction for the design: Unless I go with three centers across the front, it’s almost impossible to satisfy the criteria stated above, and using centers on the walls to the side and rear is impractical. So, I need to roll my own, as it were. I’ve looked and looked and looked at various DIY designs, and one thing has become apparent…it’s silly to go with passive crossovers. With the Behringer DCX2496 as inexpensive as it is, active electronics is the best fit to the criteria. So…we’re going active. I’m planning on some fairly nonstandard enclosures as well.
1. Active crossover: Using multiple DCX2496’s for the seven speakers. Either 3 if 2ways, or 4 if 3way design is chosen. This will buy me phase coherency, time alignment, 48db/octave slopes, static and dynamic equalization, EQ profiles for various recordings, and the ability to iterate changes to the crossover to decide on best voicing. Downside is massive amount of time spent dinking with it, and the learning curve for using it. Cost is about a wash; good crossovers (passive) are similar in their cost per speaker especially if going with a 2way setup, where I can use one DCX per three speakers. If I’m able to implement one of my secondary goals (keeping the signal digital from source to amplifier…hard to do inexpensively) then it’s only 2 speakers per DCX, as that unit has one stereo AES input. Also, this solution will enable me to get a fairly flat FR in room, instead of just anechoic, which is important given that my room isn’t ideal by any means…
2. Enclosures: 12” diameter, probably 18” long (driven by driver requirements), and haven’t decided between sealed or vented. Won’t tolerate large group delay numbers, and will likely go sealed, but not locked into that yet. The cylindrical enclosure will be capped by wood on either end, and will have a tube either coming out the bottom with a 90deg mandrel bend the diameter of the woofer or normal to the surface of the tube. Tweeter (and mid, if 3way) will be mounted on a ring faceplate around the tube holding the woofer and be rotateable to accommodate various orientations of the main enclosure tube. I’ll model it in Pro/E and put up pics if anyone’s interested and can offer criticisms. Risks of standing waves will be mitigated. Material for tube will either be aluminum, pvc, cardboard (sonotube) or carbon fiber…haven’t looked into resonance yet.
3. Amplification: External, likely several 7 channel amps
Questions….man, I’ve got tons of them as I work through this exercise. Here’s a couple I’ve got right now that I’d love some feedback on:
1. Driver selection and orientation- I’ve looked at Peerless, Vifa, Hiquophon, Dayton, etc., etc. If there’s a two way setup that’ll work well, then that’d be my first choice. I just don’t know if that’ll work well…I need to settle that aspect before I can move on. And do I do dual woofers with the tweet in a 2way? Can I get away with a single woofer? Where do I put the crossover point….with the steep filtering the DCX provides, can I get away with putting that crossover point in the critical human voice range and not suffer deleterious effects? How about some open ended discussion on this?
2. Amplifier selection-I won’t need an uber expensive amplifier rack, as the load will be resistive due to the use of the active crossover. I’d love the LPA-1 from Emotiva if the last amp wasn’t split. Grr. For the tweets, I was thinking of something like a Panny receiver run in direct mode; they won’t need a ton of current to drive to high volumes cleanly. What are some suggestions for 7 channel amplification for low cost, given the modest requirements?
Once I settle on the drivers and layout, I can proceed with the rest of the design. I welcome any discussion on it; whether it’s in relation to the two questions posted, or anything else in the project described above. Thanks!
I may post this in several of the forums dedicated to DIY speaker design as well, so don’t be surprised if you see this more than once. Just trying to get some feedback and see if there's more I ought to consider from folks that have done this before...thanks for the help!
For the SPLs you need, I mainly see some small public adress drivers, or high efficiency.
Considering a 2-way design, I'd go for
- PHL2460 : P-A driver that has ultra clean CSD up to 3kHz, and very good distorsion
- Supravox 165GMF : high efficiency, excellent dynamics, good details
- Seas 27TBFC/G tweeter : inexpensive and high power handling
However, 70Hz seems to be a low crossover point to mount these in sealed boxes. And if you build bass reflex, you can just forget about phase coherency.
I'd rather go for a system with two subs, used as the 3d way of the speakers.
Considering a 2-way design, I'd go for
- PHL2460 : P-A driver that has ultra clean CSD up to 3kHz, and very good distorsion
- Supravox 165GMF : high efficiency, excellent dynamics, good details
- Seas 27TBFC/G tweeter : inexpensive and high power handling
However, 70Hz seems to be a low crossover point to mount these in sealed boxes. And if you build bass reflex, you can just forget about phase coherency.
I'd rather go for a system with two subs, used as the 3d way of the speakers.
I concur that PA drivers are a good choice, and at least a three way + sub is the way to go. The problem with trying to get 100 dB out of a two way is that you need a lot of displacement capability which tends to drive your tweeter crossover lower, limiting clean output.
I normally tend towards active speakers. But, given your budget constraint and that you apparently don't already own the amplification that you would need, I suggest going passive. Over in the HT guide forum you'll find these three way speakers http://htguide.com/forum/showpost.php4?p=203855&postcount=537 that should get you to your stated output requirement for around $600 each plus cabinet building expense. The actual cost will vary depending on how much you are willing to spend on caps. You might come in a little lower.
In the same forum is a center channel that should match the timbre quite well. Dayton drivers are generally excellent values, see Zaph Audio for test results.
Try to use some of the money that you don't spend on the speakers for acoustic treatments. A mix of absorptiona and diffusion should give a big bang for the buck improvement.
I normally tend towards active speakers. But, given your budget constraint and that you apparently don't already own the amplification that you would need, I suggest going passive. Over in the HT guide forum you'll find these three way speakers http://htguide.com/forum/showpost.php4?p=203855&postcount=537 that should get you to your stated output requirement for around $600 each plus cabinet building expense. The actual cost will vary depending on how much you are willing to spend on caps. You might come in a little lower.
In the same forum is a center channel that should match the timbre quite well. Dayton drivers are generally excellent values, see Zaph Audio for test results.
Try to use some of the money that you don't spend on the speakers for acoustic treatments. A mix of absorptiona and diffusion should give a big bang for the buck improvement.
I'm not sure passive filtering is less expensive. Simpler to build for sure, but not less expensive.
Big coils and good caps cost a loooooot !
Off course, DCX2496 crossovers cost even more, but if you buy only one to find the best crossover points, and then build the equivalent analog active filter, it could be less expensive than passive filtering.
Big coils and good caps cost a loooooot !
Off course, DCX2496 crossovers cost even more, but if you buy only one to find the best crossover points, and then build the equivalent analog active filter, it could be less expensive than passive filtering.
I don't know much about HT, but your room sounds exactly like mine- vaulted ceiling, open to the kitchen, and a horrendous bass suck-out right where I like to sit near the center of the room. The room eats power- no idea where it goes, but there must be a really loud place somewhere else in the house! Anyway, IMO the active cross is the only way to go. A 3-way with an active cross will give you a tremendous advantage on SPL without breaking the bank on amplifiers. 110 dB is no problem. The clarity on speech and vocals is excellent if you set it up right, and the active makes it much easier to do that. I'd never go back to passives for a serious system. Don't know what they make today, nor how much they cost, but I'm partial to Morel tweeters. My old MDT-28s are smooth and rugged, and match up well to my 5" Dynaudio midbass drivers. Avoid small dome midranges, as their low crossover region is usually in an undesirable location. Crossing the tweeter at 3600+ probably isn't realistic in a 3-way, but I haven't found it a problem. I like the 4-th order Linkwitz, but opinions vary.
If you start from scratch you could certainly spend a lot on development caps and coils going passive. I guess it all depends on what you find acceptable for amplifiers. Three channels of decent amplifier and half a DCX2496 per speaker is bound to break the <$1,000 per speaker budget, especially if you're looking for a fancy finish.
The designs I suggested are ready to go. The center channel would fit the space allotted and come in at around $300 for drivers and crossover components. One guy at the HTguide forum used three centers across the front successfully. I've got an active version in my living room.
By the way, goal #5 makes your 110 dB goal difficult to achieve. A driver with enough oomph to go below 300 Hz at 110 dB is likely to be far too big to cross above 3600 Hz. IMHO, trying to avoid these crossover frequencies should be a VERY low priority. My system crosses to the tweeter at 1500 Hz, using a pseudo 8th order Cauer-Elliptic filter. I go to the woofers at 400 Hz and am very happy with the results.
If I understand your intended enclosure design, you may want to reconsider. A driver centered at the end of a cylinder is the worst case for diffraction. You'll have all sorts of response irregularities.
The designs I suggested are ready to go. The center channel would fit the space allotted and come in at around $300 for drivers and crossover components. One guy at the HTguide forum used three centers across the front successfully. I've got an active version in my living room.
By the way, goal #5 makes your 110 dB goal difficult to achieve. A driver with enough oomph to go below 300 Hz at 110 dB is likely to be far too big to cross above 3600 Hz. IMHO, trying to avoid these crossover frequencies should be a VERY low priority. My system crosses to the tweeter at 1500 Hz, using a pseudo 8th order Cauer-Elliptic filter. I go to the woofers at 400 Hz and am very happy with the results.
If I understand your intended enclosure design, you may want to reconsider. A driver centered at the end of a cylinder is the worst case for diffraction. You'll have all sorts of response irregularities.
Why not use a waveguide on the tweeter and cross it at 1500Hz ?
The Seas 27TBFC/G can produce 103dB (Xmax limited SPL) @ 1500Hz.
Add it a waveguide that bumps the response by 10dB and you have your 110dB.
Then you can use a large midrange driver (8") with no directivity issues.
Off course this needs active filtering to have a steep slope.
The Seas 27TBFC/G can produce 103dB (Xmax limited SPL) @ 1500Hz.
Add it a waveguide that bumps the response by 10dB and you have your 110dB.
Then you can use a large midrange driver (8") with no directivity issues.
Off course this needs active filtering to have a steep slope.
I like the waveguide idea, but what waveguide did you have in mind? The one Zaph used isn't quite big enough for 1500 Hz, as it is starting to roll off. He got around 6 dB of boost from the WG, but it was down 3 dB at 1500 Hz. The Xmax limited 1500 Hz max SPL would be would be 106 dB, still plenty loud when you have potentially 3 of them working together (L, C, R)
If you're shooting for a 10 dB boost, it sounds like a narrow dispersion WG, and you might be OK transitioning to an 8" driver from a dispersion point of view. However, you might be running awfully close to its breakup modes. A 7" driver or two might be a better choice going that high.
I'd really like to find an affordable waveguide solution that would let me cross over down at 1KHz and still get 105+ dB. However, something that large would conflict with the original poster's 12" max height goal.
You can get 8th order initial rolloff Cauer-Elliptic response passively with just two more components than a conventional LR4 crossover, see the Modula MTM. Whether active or passive, I shoot to keep the bounce back of the C-E filter at least 50 dB down.
If you're shooting for a 10 dB boost, it sounds like a narrow dispersion WG, and you might be OK transitioning to an 8" driver from a dispersion point of view. However, you might be running awfully close to its breakup modes. A 7" driver or two might be a better choice going that high.
I'd really like to find an affordable waveguide solution that would let me cross over down at 1KHz and still get 105+ dB. However, something that large would conflict with the original poster's 12" max height goal.
You can get 8th order initial rolloff Cauer-Elliptic response passively with just two more components than a conventional LR4 crossover, see the Modula MTM. Whether active or passive, I shoot to keep the bounce back of the C-E filter at least 50 dB down.
Waveguide ? Well design it with Hornresp and build it yourself ! I have had this project in mind for long, and turned MDF or wood would work well IMHO.
You say 106dB for the 27TBFCG @ 1500Hz ? Loudspeaker Lab tells me 103.62dB... Don't know who is right, I don't know the formula for calculating this.
Edit : oh sorry, 106dB is with the waveguide, didn't understand well at first
The PHL2460 has breakup at 3kHz, and is ultra clean under this. Could be a good choice.
You say 106dB for the 27TBFCG @ 1500Hz ? Loudspeaker Lab tells me 103.62dB... Don't know who is right, I don't know the formula for calculating this.
Edit : oh sorry, 106dB is with the waveguide, didn't understand well at first
The PHL2460 has breakup at 3kHz, and is ultra clean under this. Could be a good choice.
I just took your number and added the 3 dB that the waveguide Zaph used added at 1500 Hz. We're getting a bit OT here, except that we are showing how difficult 110 dB is to achieve with "home" drivers. Not to mention that we haven't heard from Coby. I wonder what he thinks of our suggestions
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.