I am planning a new diy project and I would like to know your thoughts!
On a previous 2-way I have come to like working with active crossovers, and after finishing the speakers at hand I plan to move on with a 6-channel diy power-amp.
I will use these speakers in an open-area living room with hardwood floors and a roll-out-for-listening rug. Not the best of conditions, but that is my reality. I listen to all kinds on music, but when I have an evening sit-down with my gear I tend to dig out acoustic music most of the time.
My design goals: I want a speaker that wants me to sit and listen to music for hours, hours and hours. I want good imaging and musicality. Coloration of the sound is not a goal, but a perception of "warm" sound is not necessarily bad.
Also, I do not want to implement a finished kit, simply because that takes away all of the fun for me. I want it to be my own design, at least to some extent.
These are my thoughts on design:
- Three way floorstander for full soundstage
- Narrow baffle design for WAF
- Transmission line for bass driver
- Separate sealed enclosures for midrange and tweeter
- Tilted baffle for time alignment
- Chamfered baffle for low diffraction
- Active LR4 three-way cross-over based on Rod Elliots P09 (Linkwitz-Riley Electronic Crossover)
These are my thoughts on driver selection:
- The level of each driver will be fine tuned using the active cross-over, so neither sensitivity nor impedance is crucial.
- Since I will be using steep filters, drivers need to be a close match in terms of "sound character".
- I do not want to bother with notch filters, so drivers with flat SPL is important
- 4" midrange for good dispersion.
- 1" ring radiator tweeter for exended frequency response and low distortion
- 7" long-excusion woofer
My budget with this is not unlimited, so highly exotic drivers like the C-Quence and stuff like that will not be possible. I think Scan-speak Illuminators are just about as far as I am prepared to go budget-wise. Spending top-dollars for exotic drivers is no self-purpose.
I am currently struggling with driver selection. It is hard to know which combinations will sound natural. I think especially the midrange and tweeter combination is important for the integration and sonic character of the system, especially as I do not want to be messing around too much with the crossovers.
Tweeter ideas:
- SB Acoustics SB29RD
- SB Acoustics TW29R
- Scan-Speak 6600
Midrange ideas:
- SB Avoustics SB12NRXF25-8
- Scan-speak 12MU/4731T00
- Scan-speak 12W/8524G00
Woofer ideas:
No ideas here, I really do not know which TS parameters to look for when setting up a TL box. Any help is appreciated!
Your help is appreciated!
On a previous 2-way I have come to like working with active crossovers, and after finishing the speakers at hand I plan to move on with a 6-channel diy power-amp.
I will use these speakers in an open-area living room with hardwood floors and a roll-out-for-listening rug. Not the best of conditions, but that is my reality. I listen to all kinds on music, but when I have an evening sit-down with my gear I tend to dig out acoustic music most of the time.
My design goals: I want a speaker that wants me to sit and listen to music for hours, hours and hours. I want good imaging and musicality. Coloration of the sound is not a goal, but a perception of "warm" sound is not necessarily bad.
Also, I do not want to implement a finished kit, simply because that takes away all of the fun for me. I want it to be my own design, at least to some extent.
These are my thoughts on design:
- Three way floorstander for full soundstage
- Narrow baffle design for WAF
- Transmission line for bass driver
- Separate sealed enclosures for midrange and tweeter
- Tilted baffle for time alignment
- Chamfered baffle for low diffraction
- Active LR4 three-way cross-over based on Rod Elliots P09 (Linkwitz-Riley Electronic Crossover)
These are my thoughts on driver selection:
- The level of each driver will be fine tuned using the active cross-over, so neither sensitivity nor impedance is crucial.
- Since I will be using steep filters, drivers need to be a close match in terms of "sound character".
- I do not want to bother with notch filters, so drivers with flat SPL is important
- 4" midrange for good dispersion.
- 1" ring radiator tweeter for exended frequency response and low distortion
- 7" long-excusion woofer
My budget with this is not unlimited, so highly exotic drivers like the C-Quence and stuff like that will not be possible. I think Scan-speak Illuminators are just about as far as I am prepared to go budget-wise. Spending top-dollars for exotic drivers is no self-purpose.
I am currently struggling with driver selection. It is hard to know which combinations will sound natural. I think especially the midrange and tweeter combination is important for the integration and sonic character of the system, especially as I do not want to be messing around too much with the crossovers.
Tweeter ideas:
- SB Acoustics SB29RD
- SB Acoustics TW29R
- Scan-Speak 6600
Midrange ideas:
- SB Avoustics SB12NRXF25-8
- Scan-speak 12MU/4731T00
- Scan-speak 12W/8524G00
Woofer ideas:
No ideas here, I really do not know which TS parameters to look for when setting up a TL box. Any help is appreciated!
Your help is appreciated!
Hi,
"All the fun" is actually all the hard work required, and if you
don't wan't to be suggested stuff where all the hard work
has been done for you, feel free to have "fun", but I
suspect your pretty clueless, so feel free to have "fun".
You don't want to be helped, you want to self indulge.
FWIW your list is pretty clueless regarding any reality.
People suggest good designs because they simply are,
to people who have no hope of ever getting there.
rgds, sreten.
http://audio.claub.net/Simple Loudspeaker Design ver2.pdf
http://www.rjbaudio.com/Audiofiles/FRDtools.html
"All the fun" is actually all the hard work required, and if you
don't wan't to be suggested stuff where all the hard work
has been done for you, feel free to have "fun", but I
suspect your pretty clueless, so feel free to have "fun".
You don't want to be helped, you want to self indulge.
FWIW your list is pretty clueless regarding any reality.
People suggest good designs because they simply are,
to people who have no hope of ever getting there.
rgds, sreten.
http://audio.claub.net/Simple Loudspeaker Design ver2.pdf
http://www.rjbaudio.com/Audiofiles/FRDtools.html
Last edited:
Curt Campbell and Jim Holtz used the Peerless NE149 in the Finalist 3-Way for midrange duty.
Here is a brief quote from Speakerdesignworks.com;
" We chose the NE149 based on feedback from others that have compared it favorably to its high end siblings from Scan-Speak. The NE149 was an unknown quantity, so I even went so far as to size
the mid cut out to accommodate SB Acoustics, Scan Revelator and Seas Excel 5 ½” drivers in case we were not satisfied with the NE149 performance. The NE149 exceeded my expectations. It tested great and sounded better."
I seriously considered building the Finalists until I got a hold of a pair of beautiful 10" 3-Way Technics.
The RS225 may be worth a look also for your design in woofer duty, or a pair of them for a "slim"ish floorstanding vented enclosure.
Here is a brief quote from Speakerdesignworks.com;
" We chose the NE149 based on feedback from others that have compared it favorably to its high end siblings from Scan-Speak. The NE149 was an unknown quantity, so I even went so far as to size
the mid cut out to accommodate SB Acoustics, Scan Revelator and Seas Excel 5 ½” drivers in case we were not satisfied with the NE149 performance. The NE149 exceeded my expectations. It tested great and sounded better."
I seriously considered building the Finalists until I got a hold of a pair of beautiful 10" 3-Way Technics.
The RS225 may be worth a look also for your design in woofer duty, or a pair of them for a "slim"ish floorstanding vented enclosure.
Hi,
I'm helping. If you want to design an advanced speaker
and can't be *rs*d to analyse other designs yourself but
want other people to do it for you, based on explaining
what is wrong with your set of very lazy premises.
It doesn't work like that. For every error you'll get
a clueless poster saying its OK and do it like that.
A thread is about totally the worst way to approach
any sort of sensible design, and usually gets hijacked
by someone knowing a little more the poster, but not
much more, severely agreeing with the poster too much.
However FWIW :
You cannot build a high quality speaker with the P09, period.
If you think you can you know nowhere near enough to do so.
That is the quandary, and you won't get the answers in a thread.
This isn't car audio, and I'm just making that explicit.
If you are not capable of analysing and understanding all
the designs out there, active, passive, or a combination,
then there is very little chance you can do any better.
Expensive drivers also need very good cabinets, an
area where the majority know next to nothing about.
rgds, sreten.
I'm helping. If you want to design an advanced speaker
and can't be *rs*d to analyse other designs yourself but
want other people to do it for you, based on explaining
what is wrong with your set of very lazy premises.
It doesn't work like that. For every error you'll get
a clueless poster saying its OK and do it like that.
A thread is about totally the worst way to approach
any sort of sensible design, and usually gets hijacked
by someone knowing a little more the poster, but not
much more, severely agreeing with the poster too much.
However FWIW :
You cannot build a high quality speaker with the P09, period.
If you think you can you know nowhere near enough to do so.
That is the quandary, and you won't get the answers in a thread.
This isn't car audio, and I'm just making that explicit.
If you are not capable of analysing and understanding all
the designs out there, active, passive, or a combination,
then there is very little chance you can do any better.
Expensive drivers also need very good cabinets, an
area where the majority know next to nothing about.
rgds, sreten.
Last edited:
Also, I do not want to implement a finished kit, simply because that takes away all of the fun for me. I want it to be my own design, at least to some extent.
If you want to use active crossover, that would be your own design of course.
It's very hard to achieve this with analog crossover. And shallow slopes (2nd order) tend to give better result in that direction. But okay you already know what drivers to have. Paper only please.
For enjoyment, I would prefer sealed for the woofer. Midrange OTOH is okay (actually better) with TL or even open baffle (see the Statement or Finalist).
Mid and tweeter doesn't have to be separated if the issue is vibration or box resonance from the bass box. Besides, you may want the midrange to be as close as possible with the tweeter. And you may want to "time align" the bass box with mid and tweeter.
Don't worry too much about dispersion, etc. In open space, big diameter midrange will give you the "authority". But I think the paper coned mid used in the Finalist can be added to your list.
Yes, with active equalization and unknown drivers SPL at least it is save to have big Xmax for the woofer. TL box for the low frequency driver is very big. You can use bigger diameter woofer instead, with average to low Sd.
Using active crossover will simplify your design effort, tho you cannot expect great result
Or study Linkwitz speakers and copy or learn from that!
Last edited:
A single 7" driver is a bit small for a floor standing 3 way without subwoofers. A pair of 8" drivers is likely to do a significantly better job and given the effort that is likely to be involved may be worth considering.
Since you mention budget as a concern then expensive drivers like top of the range Scan-Speaks are unlikely to be a wise choice. SB Acoustic drivers offer substantially higher value for money.
If you are new to speaker design then a flexible active crossover is very likely to be wise: where to crossover, which crossover slope, how much baffle step will be needed in the room, diffraction ripples, bumps in the driver response, etc... With a tunable active crossover you can do this based on measurement and listening. You could build a fixed one after you know what you want but it is unlikely to consist of two pairs of 24 dB/octave filters if you want to do a reasonable job.
Transmission line speakers are bigger and require more effort compared to, say, a vented cabinet. If you want to make a reasonable job of designing, building and sorting out one then it is hard to see an alternative to finding out about how they work and what they require in terms of driver parameters, length, absorption,... If this isn't fun for you then a rethink may be wise.
Words like musicality suggest you may currently have a less than full grasp of what is involved in designing a speaker that performs well. No problem as a starting point but you are opting for an expensive, complex and inflexible initial design. This is likely to be a problem if you expect it to turn out well. Everybody gets alls sorts of things wrong with their first attempt even if they do not realise it at the time. This doesn't mean don't go for a complex and difficult design but plan for it not to turn out that well and how to move forward if/when this comes to pass.
Yep, you have to be the kind of person who doesn't give up easily.
It's very easy to be deluded and to think that you've created something amazing. A good reference speaker helps and also to try your speakers in someone else system. Trying to find someone who will give you an honest opinion is one of the hardest parts I find. Most people don't want to be rude.
Hi - I was fairly clueless compared to the majority on this forum, but managed to build my own 3 way active speakers, which I am incredibly happy with.
They use some of the drivers you have mentioned and measure/sound very good (see pages 7 and 8 for measurements):
http://www.diyaudio.com/forums/mult...r-anyone-spot-any-probs-before-i-start-7.html
I use them with stereo sub-woofers to extend the bass, but they will go happily to 40 hz without. I have had a succession of speakers over the years and moved to high end active studio monitors some years back. These have replaced my previous Focal monitors.
Regardless of views, it is possible to use standard electrical LR4 crossovers to implement excellent acoustic results, but - you have to very well behaved drivers with a good 1-2 octaves of relatively flat overlap to be successful.
I have just completed construction of a crossover network, designed around OPA2134 op amps at 240hz and 2100 hz with even better measured results than the results shown in my thread - the crossover is very similar to Rod Elliott's design you have proposed - standard LR4 electrical slopes. I have no eq, but have just chosen my crossover points to coincide with baffle step and diffraction along with the measured driver parameters in their enclosures, and adjust for driver sensitivity and baffle step losses with my power amplifiers. Doing it this way has enabled reasonably flat frequency response, good even power response, and minimal electronics in the crossover network - minimising distortion and noise.
If you are going to do this though I would suggest at the very minimum a UMIK-1 or similar and some measurement software to ensure you get the crossover right - this is what is fundamental to a good design.
They use some of the drivers you have mentioned and measure/sound very good (see pages 7 and 8 for measurements):
http://www.diyaudio.com/forums/mult...r-anyone-spot-any-probs-before-i-start-7.html
I use them with stereo sub-woofers to extend the bass, but they will go happily to 40 hz without. I have had a succession of speakers over the years and moved to high end active studio monitors some years back. These have replaced my previous Focal monitors.
Regardless of views, it is possible to use standard electrical LR4 crossovers to implement excellent acoustic results, but - you have to very well behaved drivers with a good 1-2 octaves of relatively flat overlap to be successful.
I have just completed construction of a crossover network, designed around OPA2134 op amps at 240hz and 2100 hz with even better measured results than the results shown in my thread - the crossover is very similar to Rod Elliott's design you have proposed - standard LR4 electrical slopes. I have no eq, but have just chosen my crossover points to coincide with baffle step and diffraction along with the measured driver parameters in their enclosures, and adjust for driver sensitivity and baffle step losses with my power amplifiers. Doing it this way has enabled reasonably flat frequency response, good even power response, and minimal electronics in the crossover network - minimising distortion and noise.
If you are going to do this though I would suggest at the very minimum a UMIK-1 or similar and some measurement software to ensure you get the crossover right - this is what is fundamental to a good design.
I have already implemented a nice design based on the P09 with OPA2604, the Scanspeak 8545k midwoofer and the scanspeak 9300 tweeter. It measures pretty well, but it probably doesn't qualify as a hi-end speaker on this forum. It is by no regards a low quality speaker system, that much I know. There are certain benefits of active crossovers, one being that I happen to enjoy working with them. I am not in this to challenge guys like you, sreten. So, there is no need to tell me that I am an ignorant fool. My premises are no lazier than anyone else seeking some input on a project. I am where I am in my learning curve just like everyone else and I try to be as humble as I can about it. I am however not alone in wanting to make something that no-one else has made before, even if it turns out less optimal than a proven Gravesen design or something like that. And even if it is simply a combination of pre-proven ideas. That was kind of the reason I asked here for a starting-point for driver selection. So sreten, please spare me any additional rants on my bottomless ignorance and realize that this is hobby of mine, not the quest of a lifetime.
Anyway, now that I have made sure I will be hated for all eternity on this forum for asking the wrong questions and by critisizing one of the forum veterans, let's move on with my hobby.
Thanks Einric and Jay for the tip on the NE149, I will include that on my list.
andy19191, I also Think that the SB acoustics drivers seem to offer more value for money. Maybe not quite as much "pearls for swine".
I think I will take your advice and steer clear from TL designs. Partly because they seem hard to get "just right" and hard to fine-tune, and partly because it would require a bigger box.
Bushmeister, I like your project and I will use your thread for input!
I'm going to do some reading and will be back!
Please see my comments in red. You have told him quite a bit of misleading or incorrect information. I do not know where you have learned it yourself, but I have heard such things many times before. I do not wish for another to begin believing said misconceptions.If you want to use active crossover, that would be your own design of course.
It's very hard to achieve this with analog crossover. And shallow slopes (2nd order) tend to give better result in that direction. But okay you already know what drivers to have. Paper only please.
The filter slope has zero bearing on the amount of coloration present if properly implemented. More importantly, low order filters do not provide superior imaging, soundstaging, or anything of the sort. The filter order depends on driver/enclosure behavior in order to get the most seamless integration and the amount of acoustic slope needed for protection.
For enjoyment, I would prefer sealed for the woofer. Midrange OTOH is okay (actually better) with TL or even open baffle (see the Statement or Finalist).
A transmission line loaded midrange makes little practical sense unless it is fully damped and non-resonant. In this case, simply choose a sealed enclosure volume so that the filter you select complements the rolloff of the driver. If you choose open baffle, then that is another area entirely. A midrange in a 3-way design needs an enclosure only to isolate the rear radiation. It should not be operating low enough such that it is affected by the volume of air behind it.
Mid and tweeter doesn't have to be separated if the issue is vibration or box resonance from the bass box. Besides, you may want the midrange to be as close as possible with the tweeter. And you may want to "time align" the bass box with mid and tweeter.
Keep the mid separated from everything else. Keeping the tweeter in the same enclosure gives the added benefit of closer center-center spacing as you said. Time alignment of mid to woofer in a 3 way is not necessary. At 400Hz, you're talking about a wavelength of 33.75" or 30msec. Such a period means that the phase change wrt the mid and woofer offset is small enough to not be a concern.
Don't worry too much about dispersion, etc. In open space, big diameter midrange will give you the "authority". But I think the paper coned mid used in the Finalist can be added to your list.
This is not true; dispersion is fundamental to successful tweeter-midrange integration. Large diameter midranges do not give any more "authority" than a 2" dome. They will have higher SPL in the lower frequency ranges, but must also be crossed over more carefully to prevent beaming and power response anomalies due to off-axis lobing. The listening environment has nothing to do with it.
Yes, with active equalization and unknown drivers SPL at least it is save to have big Xmax for the woofer. TL box for the low frequency driver is very big. You can use bigger diameter woofer instead, with average to low Sd.
Using active crossover will simplify your design effort, tho you cannot expect great result
Or study Linkwitz speakers and copy or learn from that!
What sreten is doing is saving you time, money and frustration. As I keep saying on this forum, when diving in with your first original design, a 3-way is not the way to do it. There is so much that could go awry, and a computer simulation will never tell you that. Computer sims are wonderfully smoothed, and unless you have true in-box measurements (yes, including the tweeter on the baffle), a ground up design is hard.I have already implemented a nice design based on the P09 with OPA2604, the Scanspeak 8545k midwoofer and the scanspeak 9300 tweeter. It measures pretty well, but it probably doesn't qualify as a hi-end speaker on this forum. It is by no regards a low quality speaker system, that much I know. There are certain benefits of active crossovers, one being that I happen to enjoy working with them. I am not in this to challenge guys like you, sreten. So, there is no need to tell me that I am an ignorant fool. My premises are no lazier than anyone else seeking some input on a project. I am where I am in my learning curve just like everyone else and I try to be as humble as I can about it. I am however not alone in wanting to make something that no-one else has made before, even if it turns out less optimal than a proven Gravesen design or something like that. And even if it is simply a combination of pre-proven ideas. That was kind of the reason I asked here for a starting-point for driver selection. So sreten, please spare me any additional rants on my bottomless ignorance and realize that this is hobby of mine, not the quest of a lifetime.
Anyway, now that I have made sure I will be hated for all eternity on this forum for asking the wrong questions and by critisizing one of the forum veterans, let's move on with my hobby.
Thanks Einric and Jay for the tip on the NE149, I will include that on my list.
andy19191, I also Think that the SB acoustics drivers seem to offer more value for money. Maybe not quite as much "pearls for swine".
I think I will take your advice and steer clear from TL designs. Partly because they seem hard to get "just right" and hard to fine-tune, and partly because it would require a bigger box.
Bushmeister, I like your project and I will use your thread for input!
I'm going to do some reading and will be back!
Speaker design is indeed both art and science. However, it is not guesswork. Drivers must be selected because they will work well with not just one another, but the enclosure design, baffle design, and listening environment. Low ceilings may favor a D'Appolito configuration to minimize vertical dispersion, whereas large rooms with floor-ceiling glass may favor horn loaded designs.
So many things go into designing a speaker from the ground up that you should not take it personally when someone advises you to slow down. More often than not, and I fall into the category, they spent a lot of money doing trial and error instead of learning first and then designing. There is a large difference. Best of luck.
What sreten is doing is saving you time, money and frustration. As I keep saying on this forum, when diving in with your first original design, a 3-way is not the way to do it. There is so much that could go awry, and a computer simulation will never tell you that. Computer sims are wonderfully smoothed, and unless you have true in-box measurements (yes, including the tweeter on the baffle), a ground up design is hard.
Speaker design is indeed both art and science. However, it is not guesswork. Drivers must be selected because they will work well with not just one another, but the enclosure design, baffle design, and listening environment. Low ceilings may favor a D'Appolito configuration to minimize vertical dispersion, whereas large rooms with floor-ceiling glass may favor horn loaded designs.
So many things go into designing a speaker from the ground up that you should not take it personally when someone advises you to slow down. More often than not, and I fall into the category, they spent a lot of money doing trial and error instead of learning first and then designing. There is a large difference. Best of luck.
lemans23, first of all, thank you for your input on the desing issues pointed out in red.
In my previous project, which has spanned over a decade (I'm not fast, but with three kids I tend to do this in bursts) I have been through various passive crossover topologies, simulation in LspCAD and I have also built my own measurement microphone to take measurements in REW (i know I can buy a microphone, but why spoil all the fun when you can diy). I agree that measurements are key, tuning by ear is simply too hard, if not completely impossible. It was only after I got my measurement gear working that I was able to understand the results and failures of the trial and error of my previous crossovers.
I realize that moving to a three-way will be harder htan a simple 2-way, but trial and error is not only a bad thing, you know. It is kind of like solving a puzzle. Adding a piece of information at a time, tuning, scrapping and re-doing is part of the process. But... of course, I take and appreciate the point you are making! The more I know on beforehand, the more accurate my initial design will be. I will take yours and sretens advice and slow down. I guess it is time to dust off that old Vance Dickason book. Thanks!
Yes, I agree, that would be possible, but after working for so long on those speakers I want to move on with something fresh. And, the 8545k/9300 is a pretty neat pair speakers now, so I plan to keep them intact.
Hi,
Nothing wrong with active x/o's if properly designed.
There is something wrong assuming electrical 4 order
L/R will do the job even with very good drivers which
most likely will work best with 2nd order L/R acoustic.
Your problems are all to do with cabinet diffraction effects,
and fixing issues simple issues in otherwise good drivers.
Skillful x/o design, passive or active, can yield good L/R*
acoustic slopes with high value decent drivers, and will
turn out properly, unlike off the shelf active x/o's.
Rod Elliot fudges all the real issues of that active x/o.
Using one is just ignoring the real issues to address.
rgds, sreten.
* They tend to be quasi L/R, i.e. approximate the
right acoustic slopes over the x/o region, and
slightly assymetrical for significant driver offsets.
Zaphaudio.com is good place to look at proper L/R,
and he has all the electrical slopes a good active
x/o should implement, e.g. for a two way :
SR71, result is asymmetrical 4th order linkwitz-riley acoustic at 1750 Hz
http://audio.claub.net/Simple Loudspeaker Design ver2.pdf
http://www.rjbaudio.com/Audiofiles/FRDtools.html
Nothing wrong with active x/o's if properly designed.
There is something wrong assuming electrical 4 order
L/R will do the job even with very good drivers which
most likely will work best with 2nd order L/R acoustic.
Your problems are all to do with cabinet diffraction effects,
and fixing issues simple issues in otherwise good drivers.
Skillful x/o design, passive or active, can yield good L/R*
acoustic slopes with high value decent drivers, and will
turn out properly, unlike off the shelf active x/o's.
Rod Elliot fudges all the real issues of that active x/o.
Using one is just ignoring the real issues to address.
rgds, sreten.
* They tend to be quasi L/R, i.e. approximate the
right acoustic slopes over the x/o region, and
slightly assymetrical for significant driver offsets.
Zaphaudio.com is good place to look at proper L/R,
and he has all the electrical slopes a good active
x/o should implement, e.g. for a two way :
SR71, result is asymmetrical 4th order linkwitz-riley acoustic at 1750 Hz
http://audio.claub.net/Simple Loudspeaker Design ver2.pdf
http://www.rjbaudio.com/Audiofiles/FRDtools.html
Last edited:
JAY: It's very hard to achieve this with analog crossover. And shallow slopes (2nd order) tend to give better result in that direction. But okay you already know what drivers to have. Paper only please.
LEMANS23: The filter slope has zero bearing on the amount of coloration present if properly implemented. More importantly, low order filters do not provide superior imaging, soundstaging, or anything of the sort.
JAY: Technically, with low order slope such as 2nd order, the phase tracking along frequencies is better etc etc. Non-technically, ask those with good ears...
LEMANS23: The filter order depends on driver/enclosure behavior in order to get the most seamless integration and the amount of acoustic slope needed for protection.
JAY: The OP already knew why he has opted for his drivers. Smooth drivers SPL as he mentioned it. And he has mentioned about using LR4 crossover. I was not writing a book (about how to choose crossover slope for example), I was responding to him and his needs.
JAY: For enjoyment, I would prefer sealed for the woofer. Midrange OTOH is okay (actually better) with TL or even open baffle (see the Statement or Finalist).
LEMANS23: A transmission line loaded midrange makes little practical sense unless it is fully damped and non-resonant. In this case, simply choose a sealed enclosure volume so that the filter you select complements the rolloff of the driver. If you choose open baffle, then that is another area entirely. A midrange in a 3-way design needs an enclosure only to isolate the rear radiation. It should not be operating low enough such that it is affected by the volume of air behind it.
JAY: I didn't make a suggestion to use TL for midrange. It is not practical. But for his understanding I mentioned that it is actually better but is feasible only if you're looking for the best sound possible (which is not the case here). My best speaker ever, uses TL for the midrange. Fully damped and non-resonant you said? The enclosure is very heavy and cannot be lifted by a single man!
JAY: Mid and tweeter doesn't have to be separated if the issue is vibration or box resonance from the bass box. Besides, you may want the midrange to be as close as possible with the tweeter. And you may want to "time align" the bass box with mid and tweeter.
LEMANS23: Time alignment of mid to woofer in a 3 way is not necessary. At 400Hz, you're talking about a wavelength of 33.75" or 30msec. Such a period means that the phase change wrt the mid and woofer offset is small enough to not be a concern.
Note how I used "may" (In the future he may want to change the woofer, the Fxo etc etc). My "best" (representing life performance) speaker ever, the woofer is 1 meter behind the midrange tower. It is not "just" about relationship (e.g. time alignment) between the drivers. But with the room as well.
Project like this (refer to the OP), should be very very flexible.
JAY: Don't worry too much about dispersion, etc. In open space, big diameter midrange will give you the "authority". But I think the paper coned mid used in the Finalist can be added to your list.
LEMANS23: This is not true; dispersion is fundamental to successful tweeter-midrange integration.
JAY: I didn't say dispersion is not fundamental. I didn't treat the OP like a newbie, but I didn't treat him as if he has many experience as I do either. If I treated him like a newbie, I would have said "Speaker design is difficult, you need to know a lot of Physics, you have to have degree as I do, you better build other people design instead of designing stupid speakers"
LEMANS23: Large diameter midranges do not give any more "authority" than a 2" dome. They will have higher SPL in the lower frequency ranges, but must also be crossed over more carefully to prevent beaming and power response anomalies due to off-axis lobing. The listening environment has nothing to do with it.
JAY: I didn't want him to think that for better dispersion he will need to choose dome midrange. I didn't want him to worry about such issue. I suggested a 4" midrange and implicitly stated that even bigger is "better" (relatively of course).
Size of the midrange will determine who (which driver) will handle the lower mid and whether they will have perfect phase (all the relevant frequency band handled by a single driver without crossover issue). And don't for get that this will be in an open field. In a bigger room (such as open field) you need bigger/louder drivers.
JAY: Yes, with active equalization and unknown drivers SPL at least it is save to have big Xmax for the woofer. TL box for the low frequency driver is very big. You can use bigger diameter woofer instead, with average to low Sd. (I meant Qt here)
Using active crossover will simplify your design effort, tho you cannot expect great result
Or study Linkwitz speakers and copy or learn from that!
Please see my comments in red. You have told him quite a bit of misleading or incorrect information. I do not know where you have learned it yourself, but I have heard such things many times before. I do not wish for another to begin believing said misconceptions.
Hopefully it is not misleading. As I can see it, he is quite smart and with experience.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.