Preample
I was looking for a speaker build with point source mids and highs, so I looked for 12" coax with 12" woofer untill I learned that coaxial compression drivers exists. Having very positive experiences with horns and compression drivers, I'm now fixated on a BMS 4594HE, I just need a suitable horn and a 10-12" woofer.
The Beef
I've been scrolling through dozens and dozens of specsheets and learned that there are masses of good twelve inchers out there, so I haven't really looked at 10" drivers yet. What I don't know which ones would fit the sound quality of this CD and my existing sealed 14" subs (which is a secondary concern). For horn I was looking at RCF HF94, because of wide dispersion and loading down to 500 Hz and it's quite pretty. Still open to suggestions though. Needs to be pretty enough for other than blind tests ;-)
So technically I'm going for
Being relatively new to DIY I really don't know which parameters to look for in a driver for this application. All forums and articles are full of tips about how to extend bass response, which does not apply here in the same sense. Dedicated mid range drivers seem to have completely different EBP than sealed subwoofer drivers, but this is something in between. I'm just looking for top quality bass-to-upperbass woofer. Also the proper enclosure size is unknown to me for this frequency range, where I'm not looking for maximal extension. Simulator suggests that almost any box size that can house the driver goes for >100 Hz, is this actually correct?
Some pointers of candidates:
- Faital PRO 12PR320?? Top of the list right now. Also other models look good.
- Beyma, P80ND/V2?? 12MC700ND??
- Eighteen sound?? Many models.
- BMS 12N810 looks fine for upper frequencies, but doesn't go very low. There are many other models and this manufacturer seems to be quite respected, although their specsheets are not always so good. Just honest maybe? BMS is the only one that I remember posting their distortion graphs.
- Precision Devices, like PD.124NR1??
- Acoustic Elegance TD12M or TD12S, a bit too expensive with shipping to Europe
- I haven't yet even looked at B&C, Ciare, Eminence, Fane, JBL, RCF, SB Audience etc or any hifi-like manufacturers
- I have got diy speakers with "starting at" prices, but I would like to try something higher now and use them for many years.
Preliminary Cadding
Woofer sims. 2 quite different drivers:
Faital PRO 12PR320 in sealed 22,9 litre box.
BMS 12N810 in sealed 18.6 litre box.
(cyan in excursion if cone force)
What I learn from these sims is almost nothing. They both have unlimited performance for home use and can be EQed to any FR target. The BMS is slightly better around 120 Hz, not sure what causes the hump for the Faital. Probably a sim parameter thing.
Cabinet drafts in attachments.
I was looking for a speaker build with point source mids and highs, so I looked for 12" coax with 12" woofer untill I learned that coaxial compression drivers exists. Having very positive experiences with horns and compression drivers, I'm now fixated on a BMS 4594HE, I just need a suitable horn and a 10-12" woofer.
The Beef
I've been scrolling through dozens and dozens of specsheets and learned that there are masses of good twelve inchers out there, so I haven't really looked at 10" drivers yet. What I don't know which ones would fit the sound quality of this CD and my existing sealed 14" subs (which is a secondary concern). For horn I was looking at RCF HF94, because of wide dispersion and loading down to 500 Hz and it's quite pretty. Still open to suggestions though. Needs to be pretty enough for other than blind tests ;-)
So technically I'm going for
- coax 1.4" cd, xo @ 6.3 kHz, this is rated down to 300 Hz
- horn with wide dispersion, because I'm using them at couch and at a desk between the speakers, and good loading down to woofer range
- 10-12" woofer, xo from horn to woofer @ 500-900 Hz, depending on selected horn and woofer. With HF94 I'm looking at around 650 Hz.
- xo to sealed subs @ 60-100 Hz, depending on selected woofer and fine-tuning
- fully active with dsp-plates.
- sealed, since subs are also sealed
- doesn't need to go below 80 Hz, since will be crossed to subs!
- preferably (much) less than 30 litre cabinet
- I've been looking at >=8 Ohm speakers, because fully active.
- Sound quality is what I'm looking for here, not just "best value". If there is a reasonably priced driver with performance where any improvement wouldn't really matter anymore, that's my choice.
Being relatively new to DIY I really don't know which parameters to look for in a driver for this application. All forums and articles are full of tips about how to extend bass response, which does not apply here in the same sense. Dedicated mid range drivers seem to have completely different EBP than sealed subwoofer drivers, but this is something in between. I'm just looking for top quality bass-to-upperbass woofer. Also the proper enclosure size is unknown to me for this frequency range, where I'm not looking for maximal extension. Simulator suggests that almost any box size that can house the driver goes for >100 Hz, is this actually correct?
Some pointers of candidates:
- Faital PRO 12PR320?? Top of the list right now. Also other models look good.
- Beyma, P80ND/V2?? 12MC700ND??
- Eighteen sound?? Many models.
- BMS 12N810 looks fine for upper frequencies, but doesn't go very low. There are many other models and this manufacturer seems to be quite respected, although their specsheets are not always so good. Just honest maybe? BMS is the only one that I remember posting their distortion graphs.
- Precision Devices, like PD.124NR1??
- Acoustic Elegance TD12M or TD12S, a bit too expensive with shipping to Europe
- I haven't yet even looked at B&C, Ciare, Eminence, Fane, JBL, RCF, SB Audience etc or any hifi-like manufacturers
- I have got diy speakers with "starting at" prices, but I would like to try something higher now and use them for many years.
Preliminary Cadding
Woofer sims. 2 quite different drivers:
Faital PRO 12PR320 in sealed 22,9 litre box.
BMS 12N810 in sealed 18.6 litre box.
(cyan in excursion if cone force)
What I learn from these sims is almost nothing. They both have unlimited performance for home use and can be EQed to any FR target. The BMS is slightly better around 120 Hz, not sure what causes the hump for the Faital. Probably a sim parameter thing.
Cabinet drafts in attachments.
Attachments
Hi JukkaM,
is this for home use or for public address? If home use, the loading doesn't mean much. It is going to be loud enough nevertheless 🙂 More important property of a waveguide is the pattern control. Pattern control bandwidth is roughly determined by the device mouth size, or woofer cone size. Optimally you would want to match pattern control of the horn and the direct radiating woofer you are crossing it to around the crossover frequency. You'd want the woofer start to beam and waveguide to lose pattern control around the same bandwidth so that they both kind of have similar directivity around the crossover. This yields smooth DI and power response.
If the driver is different size than the waveguide, you don't get as good results (some power / DI response as some issues). Same goes if the pattern control is not symmetrical, which yields a pattern flip and again some ripple to the power and DI (90x90 no pattern flip, 90x40 bad pattern flip). In this case you'd want to have the horn wider than the woofer to get the crossover somewhere into the patternflip. It is all tied to the physical size of sound waves and how they interfere, you need to size the "radiators" accordingly. If you wish to crossover at 650Hz you should be looking at 15" waveguide and woofer. If you use 10" waveguide and woofer best place for crossover is somewhere above 1kHz. 1kHz is easily handled with 1" driver at home, no need for the coaxial driver. But, the design is yours so you can do it however you like. You can use the big driver with small horn, but you kind of lose the benefits, pattern control to lower frequency. 1" would be cheaper and two amplifiers less and would perform the same (can handle the SPL at home, same pattern since it is determined by the waveguide).
You could easily use 1" driver at home. Buy best constant directivity waveguide (without diffraction slot, smooth mouth transition) you find and then similar size woofer. If you wish to use the coaxial driver, I'd buy / fabricate big axisymmetric freestanding horn/wageguide with good rollback (see ATH4 thread on the multiway forum) and maybe a 15" or 18" woofer below. Big waveguides cost a lot more than the small PA stuff due to smaller market. Always a compromise somewhere 😉 Have fun!🙂
is this for home use or for public address? If home use, the loading doesn't mean much. It is going to be loud enough nevertheless 🙂 More important property of a waveguide is the pattern control. Pattern control bandwidth is roughly determined by the device mouth size, or woofer cone size. Optimally you would want to match pattern control of the horn and the direct radiating woofer you are crossing it to around the crossover frequency. You'd want the woofer start to beam and waveguide to lose pattern control around the same bandwidth so that they both kind of have similar directivity around the crossover. This yields smooth DI and power response.
If the driver is different size than the waveguide, you don't get as good results (some power / DI response as some issues). Same goes if the pattern control is not symmetrical, which yields a pattern flip and again some ripple to the power and DI (90x90 no pattern flip, 90x40 bad pattern flip). In this case you'd want to have the horn wider than the woofer to get the crossover somewhere into the patternflip. It is all tied to the physical size of sound waves and how they interfere, you need to size the "radiators" accordingly. If you wish to crossover at 650Hz you should be looking at 15" waveguide and woofer. If you use 10" waveguide and woofer best place for crossover is somewhere above 1kHz. 1kHz is easily handled with 1" driver at home, no need for the coaxial driver. But, the design is yours so you can do it however you like. You can use the big driver with small horn, but you kind of lose the benefits, pattern control to lower frequency. 1" would be cheaper and two amplifiers less and would perform the same (can handle the SPL at home, same pattern since it is determined by the waveguide).
You could easily use 1" driver at home. Buy best constant directivity waveguide (without diffraction slot, smooth mouth transition) you find and then similar size woofer. If you wish to use the coaxial driver, I'd buy / fabricate big axisymmetric freestanding horn/wageguide with good rollback (see ATH4 thread on the multiway forum) and maybe a 15" or 18" woofer below. Big waveguides cost a lot more than the small PA stuff due to smaller market. Always a compromise somewhere 😉 Have fun!🙂
Last edited:
Jeez, no time to edit. It is for home use, says the title 🙂 Sorry 😀 There was some other stuff in the text I tried to edit... mainly, not sure if the rcf 90x40 horn has too bad patternflip since the pattern is achieved by a diffraction slot. Main takeaway is that the RCF94 is optimized for compact PA use with high SPL application, this means the diffraction slot and the size is not optimal for home use. Anyway, try to think the design so that all compromises, more importantly the trade-offs included. Try to land all trade-offs into cost and not audio quality if you want the best. For example the high SPL capability is not needed at home and you should try to sacrifice that for other more important aspects on the design like the pattern control. Have fun!
Last edited:
Thanks for quick replies.
Thick blue is excursion, light blue is xmax, cyan is cone force. It barely moves at 2.84 V, which is the simulated input.
One design feature here is that front baffle width is limited to ~350 mm due to space limitations and finding low loading horns for within that limitation is difficult. I'm going need more suggestions for that. 12" can be housed in 350 mm if not much volume is required, that's the width of the cabinet drafts in pics. 10" woofer could be a better fit for pattern control or crossing higher, things which I didn't yet have time to investigate.
I'll look into these new CDs that extend lower, but I still have the coax CD fixation 😉 And it takes a lot more time to look for a lot more CDs. I was planning to use passive xo for the coax, one can be bought from the same manufacturer and maybe tuned with better components. This CD is actually so sensitive with a horn that it could probably be driven to deafening levels with a headphone amp. That would be fun to try!
I realize the world is full of 2-ways with horn, in which case I would have gone to a well known design. I have many times almost pulled the trigger on a 2-way kit. I have auditioned some and they sound mostly good, but there is usually something lacking in mid range compared to 3-ways. So instead of yet another 2-way I would really want to try a 3-way, which turns into 4-way when combined with subs. My current setup is 2-way 1" cd +horn and 8" midbass on the subs, that sounds quite good when optimized with dsp and Dirac and I consider this to be really good value, but as stated in the first post, I want to try something more. If you know what I mean 😉 The current system doesn't really qualify as full 3-way when combined with subs, because the subs don't extend as high as typical 3-way is crossed. It's just 2-way with proper bottom.
Thick blue is excursion, light blue is xmax, cyan is cone force. It barely moves at 2.84 V, which is the simulated input.
One design feature here is that front baffle width is limited to ~350 mm due to space limitations and finding low loading horns for within that limitation is difficult. I'm going need more suggestions for that. 12" can be housed in 350 mm if not much volume is required, that's the width of the cabinet drafts in pics. 10" woofer could be a better fit for pattern control or crossing higher, things which I didn't yet have time to investigate.
I'll look into these new CDs that extend lower, but I still have the coax CD fixation 😉 And it takes a lot more time to look for a lot more CDs. I was planning to use passive xo for the coax, one can be bought from the same manufacturer and maybe tuned with better components. This CD is actually so sensitive with a horn that it could probably be driven to deafening levels with a headphone amp. That would be fun to try!
I realize the world is full of 2-ways with horn, in which case I would have gone to a well known design. I have many times almost pulled the trigger on a 2-way kit. I have auditioned some and they sound mostly good, but there is usually something lacking in mid range compared to 3-ways. So instead of yet another 2-way I would really want to try a 3-way, which turns into 4-way when combined with subs. My current setup is 2-way 1" cd +horn and 8" midbass on the subs, that sounds quite good when optimized with dsp and Dirac and I consider this to be really good value, but as stated in the first post, I want to try something more. If you know what I mean 😉 The current system doesn't really qualify as full 3-way when combined with subs, because the subs don't extend as high as typical 3-way is crossed. It's just 2-way with proper bottom.
Hi, yeah capability to play loud is fun but deafening certainly is not 🙂 Just prepare to study and experiment if you are chasing best performance, might take few prototypes but you seem to have the curiosity needed so go for it. Audio is a journey, nobody ever finds the holy grail 😀
I predict few years ahead: You'll find more of the "sound" is in the horn/waveguide/the system design and not in the compression driver given that the driver is suitable for the application and not broken. I'd say that the waveguide and the complete system design is way more important than changing capacitor brand or driver model. Changing listening position, curtains or carpet have more effect on the sound.
The 2/3 way midrange quality difference can be from multiple reasons and the waveguide/woofer integration and crossover implementation might be the most important (smooth controlled directivity) one given that the woofer doesn't have bad breakup and tweeter doesn't distort. It all goes back to the system design anyway which goes back to the sound wavelengths and limitations of your design goal. 350mm max width defines it all. You can use the coaxial compression driver but in my view it is waste of money since most 1" drivers can handle 350mm wavelength with loud enough SPL, loading or not. I'm kind of trying to brainwash some engineering mindset into you right at the start 😀 Engineering mindset is certainly needed if you want to have real performance, marketing men will tell anything about performance to sell more.
As a tip I've observed the following issues that affect midrange "quality" and don't involve waveguide or compression driver so you might want to design them out as well if you are chasing out better midrange:
- reflex port leakage/resonance
- cabinet resonance
- woofer resonance/breakup
- edge diffraction
- crossover implementation
Another tip for the design mindset iceberg is that it is important to realize that only thing that matters is the acoustic performance of the loudspeaker (well, aesthetics as well). Multiple transducers and crossovers are needed to be able to play the desired bandwidth (say 30Hz - 15kHz) loud enough since wavelengths at the end of the bandwidth are vastly different. Purpose of the crossover is merely to divide (and equalize) the signal to different transducers electrically so that they combine back to whole signal again acoustically so acoustic issues are the main design problem. When you get the acoustic design right, crossover is not a hard thing to do. If acoustic design is not right only thing to improve performance is to fix the acoustic design 😉 Have fun!
ps. with the ath4 software you can design any size horn with directivity and loading you want, except there is relationship with all three so you have to choose the trade-off.
I predict few years ahead: You'll find more of the "sound" is in the horn/waveguide/the system design and not in the compression driver given that the driver is suitable for the application and not broken. I'd say that the waveguide and the complete system design is way more important than changing capacitor brand or driver model. Changing listening position, curtains or carpet have more effect on the sound.
The 2/3 way midrange quality difference can be from multiple reasons and the waveguide/woofer integration and crossover implementation might be the most important (smooth controlled directivity) one given that the woofer doesn't have bad breakup and tweeter doesn't distort. It all goes back to the system design anyway which goes back to the sound wavelengths and limitations of your design goal. 350mm max width defines it all. You can use the coaxial compression driver but in my view it is waste of money since most 1" drivers can handle 350mm wavelength with loud enough SPL, loading or not. I'm kind of trying to brainwash some engineering mindset into you right at the start 😀 Engineering mindset is certainly needed if you want to have real performance, marketing men will tell anything about performance to sell more.
As a tip I've observed the following issues that affect midrange "quality" and don't involve waveguide or compression driver so you might want to design them out as well if you are chasing out better midrange:
- reflex port leakage/resonance
- cabinet resonance
- woofer resonance/breakup
- edge diffraction
- crossover implementation
Another tip for the design mindset iceberg is that it is important to realize that only thing that matters is the acoustic performance of the loudspeaker (well, aesthetics as well). Multiple transducers and crossovers are needed to be able to play the desired bandwidth (say 30Hz - 15kHz) loud enough since wavelengths at the end of the bandwidth are vastly different. Purpose of the crossover is merely to divide (and equalize) the signal to different transducers electrically so that they combine back to whole signal again acoustically so acoustic issues are the main design problem. When you get the acoustic design right, crossover is not a hard thing to do. If acoustic design is not right only thing to improve performance is to fix the acoustic design 😉 Have fun!
ps. with the ath4 software you can design any size horn with directivity and loading you want, except there is relationship with all three so you have to choose the trade-off.
Last edited:
Thanks, I'll check out ath4 as soon as I can. After watching some kites fly at the beach I've come to think that I could go for 10" and possibly narrower dispersion for horn to allow more options, like the Eighteen sound XTs, or I could start experimenting with 2-way build and migrate to 3-way later. Any smaller than 10" would take me to conventional hifi.
For directivity at crossover, low bass is usually omni directional and so are horns of going low enough with small horn. I was kinda hoping that would work a sweetspot within my intented xo range.
Right now I'm sweating to much to think straight...
For directivity at crossover, low bass is usually omni directional and so are horns of going low enough with small horn. I was kinda hoping that would work a sweetspot within my intented xo range.
Right now I'm sweating to much to think straight...
Hi, I've built a few different "horn over 12" using the 4594he".
Horns were the xt1464, xr1464c, xr1496c, and rcf hf950 .
12" was rcf mb12n351 or mb12n405
The xt1464 is the best sounding horn to me, when the narrower pattern works
The 90 deg hf950 is nice and even, i think the veins in the throat help with vertical pattern.
I'd probably choose a faital or something that goes lower than the rcf 12"s, if doing it again. It's been hard getting the rcf's down thru 100Hz.
The bms coax will greatly outrun any single 12" made. Don't think i'd mate it with a single 10" even for home. Also, matching horn width and cone width makes sense to me. The CD is probably realistically good to about 650Hz with any of these smaller type horns.
I've used it down to 500Hz, but with considerably larger horns.
Hope this helped...
Horns were the xt1464, xr1464c, xr1496c, and rcf hf950 .
12" was rcf mb12n351 or mb12n405
The xt1464 is the best sounding horn to me, when the narrower pattern works
The 90 deg hf950 is nice and even, i think the veins in the throat help with vertical pattern.
I'd probably choose a faital or something that goes lower than the rcf 12"s, if doing it again. It's been hard getting the rcf's down thru 100Hz.
The bms coax will greatly outrun any single 12" made. Don't think i'd mate it with a single 10" even for home. Also, matching horn width and cone width makes sense to me. The CD is probably realistically good to about 650Hz with any of these smaller type horns.
I've used it down to 500Hz, but with considerably larger horns.
Hope this helped...
Thanks mark100, that's the kind of reply I wanted to hear, because it would verify my background work and basically give me confidense in the design I'm making. I'm going to use fully active dsp, so I can fine tune details later easier.
I'm going to consider all replies later and make a decision based on some more investigation. All ideas and suggestions are still welcome.
I'm going to consider all replies later and make a decision based on some more investigation. All ideas and suggestions are still welcome.
The 4594 sounds wonderful mounted to the SEOS 24...if you can find any. I currently have this configuration with two AE IB15 underneath.
Okey, I've been googling and thinking about this. Here are some key points:
Also if using coax cd on a horn and crossing to 12", the directivity would probably collapse at the crossover, but is that really dangerous? I can't really see any options how to cross that coax driver with smooth directivity change to any cone driver which isn't at least 60 cm (2') wide. The point is: it's eventually going to be omni-directional, so how fast can that transition happen?
I'm still open to just 2-way. Since I'm still a novice with horn design, what I want/need is that one kind person would model me a horn that has 350 mm wide mouth (height and depth less than 1000 mm please..) and dispersion of ~80x60 and can be crossed at around 500 Hz without the pattern exploding. If that is impossible or the end result is simply horrible, then I will ditch this coax cd idea from my (current) home for good and go for a 2-way. Also I would like to know what it takes to make a cd like this go low, probably quite enormous horn.
Other than that, I'm at the mercy of commercial readily available horns. I may be able to strech the width of the baffle to 380 mm, which is the exact width of XT1464 (60x40 degrees), which could be enough with toe in or just toing in more when needed. RCF HF950 looks fine on paper, but not to eye, it would need some trims the very least! Today I can't tell different horn designs apart. I cannot see the diffraction slot and there is no mention of it in its data sheet, so I would need more education on this subject. The ATH thread looks like the next best thing since T/S parameters, but it's hundreds of pages long and reading and understanding it would postpone my project by a far. So here I am asking for help 🙂
- I'm not going to use larger than 12" woofer in this project.
- Which ever woofer I'm going to use should be dispersing quite near or more than 180 degress at the desired bandwidth (xo fair bit less than 1 Khz).
- I'm a novice at horn design.
Also if using coax cd on a horn and crossing to 12", the directivity would probably collapse at the crossover, but is that really dangerous? I can't really see any options how to cross that coax driver with smooth directivity change to any cone driver which isn't at least 60 cm (2') wide. The point is: it's eventually going to be omni-directional, so how fast can that transition happen?
I'm still open to just 2-way. Since I'm still a novice with horn design, what I want/need is that one kind person would model me a horn that has 350 mm wide mouth (height and depth less than 1000 mm please..) and dispersion of ~80x60 and can be crossed at around 500 Hz without the pattern exploding. If that is impossible or the end result is simply horrible, then I will ditch this coax cd idea from my (current) home for good and go for a 2-way. Also I would like to know what it takes to make a cd like this go low, probably quite enormous horn.
Other than that, I'm at the mercy of commercial readily available horns. I may be able to strech the width of the baffle to 380 mm, which is the exact width of XT1464 (60x40 degrees), which could be enough with toe in or just toing in more when needed. RCF HF950 looks fine on paper, but not to eye, it would need some trims the very least! Today I can't tell different horn designs apart. I cannot see the diffraction slot and there is no mention of it in its data sheet, so I would need more education on this subject. The ATH thread looks like the next best thing since T/S parameters, but it's hundreds of pages long and reading and understanding it would postpone my project by a far. So here I am asking for help 🙂
The idea of the coaxial comps is to cross very low, 300hz is posible at PA volume, unfortunately you need a massive horn for this. E.G Suggested Design: 15SW100 / DCX464 / ME464 Point Source - YouTube
Have you thought about the bms triaxial transducers?
Neodymium Triaxial Transducers
Have you thought about the bms triaxial transducers?
Neodymium Triaxial Transducers
Last edited:
"The idea of the coaxial comps is to cross very low, 300hz is posible at PA volume, unfortunately you need a massive horn for this"
---
How does this work at PA volume? The air in the throat behaves in a non linear manner if you compress it hard enough.
"Past a certain limit (which varies with frequency vs. horn cutoff), supplying more power creates no increase in SPL, but simply creates more and more distortion. The maximum power must be reduced as frequency increases [...] According to a technical note from JBL [ 2 ], the situation is actually worse than the graph shows. A 200Hz horn at 10kHz can readily generate 48% second harmonic distortion, with as little as 2.5W (electrical) input - a mere 0.75 acoustical Watts. As noted in references 1 and 2, this information was first determined in 1954, but over time seems to have been lost"
PA Systems
---
How does this work at PA volume? The air in the throat behaves in a non linear manner if you compress it hard enough.
"Past a certain limit (which varies with frequency vs. horn cutoff), supplying more power creates no increase in SPL, but simply creates more and more distortion. The maximum power must be reduced as frequency increases [...] According to a technical note from JBL [ 2 ], the situation is actually worse than the graph shows. A 200Hz horn at 10kHz can readily generate 48% second harmonic distortion, with as little as 2.5W (electrical) input - a mere 0.75 acoustical Watts. As noted in references 1 and 2, this information was first determined in 1954, but over time seems to have been lost"
PA Systems
Yes by increasing the flare rate you reduce the harmonic distortion and increase the low frequency cut off. However 0.75 acoustic watts is very loud 😉 a level that would destroy a conventional dome tweeter. Harmonic distortion for horn+compression drivers at domestic listening levels is usually acceptable even with lower flare rate horns. It should also be pointed out the peak spl is often inside the compression drivers phase plug channels so flare rate may have less effect than expected.
I few things I need suggestions with:
If I find a horn that can do well to 800-1000 Hz, I can basically use regular 1" or 1.4" CD and not a coax (like FaitalPRO HF1440 or HF10AK), because coax would only double my budget and offer quite much more headroom, which I'm not going to need.
How will the all-important mid range sound when crossed like that to a 12"? (probably FaitalPRO 12PR320)
There is a magical combo to beat: Eighteen sound NSD1095N+XT1086 with 10NMB420, which is a reasonably priced package with fine sound quality. If I want better performance than that (specially the mids), what kind drivers should I be looking at? Can a 12" produce good enough mid range or should I be looking at a smaller driver if low end extension is not required past 100 Hz? Or just go with the Eighteen sound? 🙂
If I find a horn that can do well to 800-1000 Hz, I can basically use regular 1" or 1.4" CD and not a coax (like FaitalPRO HF1440 or HF10AK), because coax would only double my budget and offer quite much more headroom, which I'm not going to need.
How will the all-important mid range sound when crossed like that to a 12"? (probably FaitalPRO 12PR320)
There is a magical combo to beat: Eighteen sound NSD1095N+XT1086 with 10NMB420, which is a reasonably priced package with fine sound quality. If I want better performance than that (specially the mids), what kind drivers should I be looking at? Can a 12" produce good enough mid range or should I be looking at a smaller driver if low end extension is not required past 100 Hz? Or just go with the Eighteen sound? 🙂
Few remarks to help you: HF10AK might not bend below 1k (I have one, never tried that low thought since my horn is small), the newer HF108 might be more suitable (it is akin to HF1440). At least the hf10ak has more issues on the frequency response and crossover around 2k happens quite naturally with small horn. While the 1.4" throats can definitely play lower in frequency the throat size starts to sacrifice the top end compared to 1" so if at all possible by the SPL requirement, 1" drivers seem to be the balanced option for home use.
I'm not sure if you can get too much performance over the 18sound combo you have (is it the Finnish Ruutinassakka kit?🙂 ) for the space limitation you have. With 12" parts instead of 10" you'll get only a bit lower crossover frequency and chance to design enclosure and crossover yourself. While this is not too difficult it needs few prototypes to weed out as many issues as possible to gradually reach the performance you are trying to achieve.
So the same process could be done for the existing parts as well. I don't know your configuration but it might be possible to get little bit more performance of your current transducers if you have time and measurement gear. Anyway, it is similar kind of "tuning" and "shaving" process ahead with the old parts or with new parts to get the last inch of performance out of a system. "Better" transducers might give some performance as well, but not sure how much since 18sound seems to make quality products.
If you had more space there would be possibility to take the crossover lower and take the narrower directivity lower in frequency with bigger horn/woofer. If the mid issues are due to the crossover (frequency) there is not much more you can do than make the crossover out of the critical band. Or it is just some other issue(s) with the current system like resonance(s) or something, which the new system could have as well.
edit. some background: "midrange" has been problem for me as well, 8" woofer and similar sized horn on a three way system. I think it is combination of cabinet resonances and diffraction that make the mids not as defined or something, or the directivity widening above 1kHz. Multiple issues nevertheless. Haven't been able to build new protos for some time now, maybe later this year. I've got DSP so possible to tweak the crossover to hearts content. It sound pretty amazing, but not as good as I wished. There is always some uncertainty and lack of confidence on the design. Meanwhile I've loved to listen the thing and had loads of fun so it is not too important. The most important thing is to have fun. Newer rule out option of buying new commercial speakers either, life is too short to have bad time voluntarily 🙂
I'm not sure if you can get too much performance over the 18sound combo you have (is it the Finnish Ruutinassakka kit?🙂 ) for the space limitation you have. With 12" parts instead of 10" you'll get only a bit lower crossover frequency and chance to design enclosure and crossover yourself. While this is not too difficult it needs few prototypes to weed out as many issues as possible to gradually reach the performance you are trying to achieve.
So the same process could be done for the existing parts as well. I don't know your configuration but it might be possible to get little bit more performance of your current transducers if you have time and measurement gear. Anyway, it is similar kind of "tuning" and "shaving" process ahead with the old parts or with new parts to get the last inch of performance out of a system. "Better" transducers might give some performance as well, but not sure how much since 18sound seems to make quality products.
If you had more space there would be possibility to take the crossover lower and take the narrower directivity lower in frequency with bigger horn/woofer. If the mid issues are due to the crossover (frequency) there is not much more you can do than make the crossover out of the critical band. Or it is just some other issue(s) with the current system like resonance(s) or something, which the new system could have as well.
edit. some background: "midrange" has been problem for me as well, 8" woofer and similar sized horn on a three way system. I think it is combination of cabinet resonances and diffraction that make the mids not as defined or something, or the directivity widening above 1kHz. Multiple issues nevertheless. Haven't been able to build new protos for some time now, maybe later this year. I've got DSP so possible to tweak the crossover to hearts content. It sound pretty amazing, but not as good as I wished. There is always some uncertainty and lack of confidence on the design. Meanwhile I've loved to listen the thing and had loads of fun so it is not too important. The most important thing is to have fun. Newer rule out option of buying new commercial speakers either, life is too short to have bad time voluntarily 🙂
Last edited:
On the woofers to use, maybe take a look at the Beyma 12CMV2 also. You can use it without issues to a bit above 1kHz (after that it start beaming and resonating). So even with a one way compression driver it could work very well in a 30L sealed to about 80Hz.
Yes, the Ruutinassakka is now my reference 🙂 I don't own them, but I'm considering them as a 2-way option. I listened to them in Hifitalo listening room recently and they sounded very good, except that I wished something more for the mid range (I had to lower the treble quite much on the amp, it was way too bright, that might have left mid range uneven). Now that I got decent bass from sealed subs I've come to want better mid range. My current speakers are a combination of Eighteen sound 8" ferrite woofer (8MB500 IIRC), Faital horn and entry level BMS cd in a ~40 litre ported box. Alone they are ok, tuned to get best bass from that element, so mid range is lacking, treble is fine, I actually like that BMS. When I plug the ports and crossover to subs at 100 Hz, the mid range clears up remarkably and with some Dirac magic the speakers turns into a very good sounding system. I could go mad with my existing dsp and fien tune them until the world ends, but I think I can get much more detail by upgrading. The Eighteen sound combo, Ruutinassakka, if I made them a sealed cabinet, cross them to subs and fine tune with DSP, just like the current system, it could be enough for me. If I employ a dsp plate for them, the price goes up, but I could better match output levels (Ruutinassakka elements have something like 20 dB efficiency difference!) and fine tune to fit my taste. Does anyone have experience with these Eighteen sound drivers and how to go versus some competion?
If you look at the attachment here (I hope the text is well visible), you can see the approximated xo points and octave ranges of drivers of a few different style of builds. When looking at this while keeping in mind that I need the same speakers to work for nearfield and to the couch without moving them, wide horizontal dispersion is needed, I think that best overall sound quality for near field would be accomplished with point source for at least mids and highs, bass is not as important. To me horn loaded coax cd looks like very good option, since I like the sound of horns.
I was looking at Troels's Faital-3WC, but it looks like drivers are too wide apart. 3-way: horn, 8" and 12"? Maybe, but the 8" looks like too big for point source and smaller doesn't differ much from Faital-3WC. Also having 3 cones (mid, woofer, sub) below 1.5 kHz seems like an overkill. Higher xo to cd might endanger to point source and wastes cd+horn capability. I was also playing with a thought of getting a KEF LS50 for its UniQ and cross it as-is to a larger bass driver (imagine WCW with 12"ers and LS50 in the middle..) and also to the subs. Many conventional hifi drivers may also offer great alternatives, although power handling could become a problem 😉
Attachments
Hi, yeah definitely closed cabinet for the mid and try to make the baffle as small as possible, eg. as little baffle around transducers as possible to minimize diffraction interference in the listening window. Alternatively add slants or roundovers to minimize the flat front panel area. If your current mid driver has typical rising response and the box becomes a bit too small in volume, you can use DSP to boost a little (or cut the mids and attenuate the CD), there should be enough headroom in home use with pro drivers and enough amplifier power.
Never heard or looked very closely the Ruutinassakka but it seems to be constant directivity horn in it so if you listen straight on-axis it should be too bright. Toe in 45 degrees if you have the equilateral triangle listening position. This will yield listening axis at 15 degrees off-axis where the constant directivity devices are often tuned. This is done to achieve maximum listening area sweet spot and least early reflections from room walls.
The rationale for the toe in and listening axis is: when you move from the sweet spot to a side (on a 3 person sofa for example), the now closer speaker loses some SPL due to being more off-axis which compensates the closer distance while the further speaker is now more on-axis and again compensates for the distance = stable phantom center = big sweet spot.
You could build a point source with the coaxial compression driver and small horn with the extra cost of money, if you do please report how it went 🙂 If it wasn't worth it I think you can resell the drivers at least 50% of the purchasing price, maybe more since they are not too often seen on used market here in Finland.
I'm not sure it is that much better than not point source to compensate for the cost. Not being point source isn't too bad if you listen far field and the off-axis response is good (smooth power response and DI). It is pretty simple with DSP to compensate for delays and what not so the acoustical design of the system is the key for success I think. Few prototypes might be needed to check out the performance in any case. Have fun!🙂
Never heard or looked very closely the Ruutinassakka but it seems to be constant directivity horn in it so if you listen straight on-axis it should be too bright. Toe in 45 degrees if you have the equilateral triangle listening position. This will yield listening axis at 15 degrees off-axis where the constant directivity devices are often tuned. This is done to achieve maximum listening area sweet spot and least early reflections from room walls.
The rationale for the toe in and listening axis is: when you move from the sweet spot to a side (on a 3 person sofa for example), the now closer speaker loses some SPL due to being more off-axis which compensates the closer distance while the further speaker is now more on-axis and again compensates for the distance = stable phantom center = big sweet spot.
You could build a point source with the coaxial compression driver and small horn with the extra cost of money, if you do please report how it went 🙂 If it wasn't worth it I think you can resell the drivers at least 50% of the purchasing price, maybe more since they are not too often seen on used market here in Finland.
I'm not sure it is that much better than not point source to compensate for the cost. Not being point source isn't too bad if you listen far field and the off-axis response is good (smooth power response and DI). It is pretty simple with DSP to compensate for delays and what not so the acoustical design of the system is the key for success I think. Few prototypes might be needed to check out the performance in any case. Have fun!🙂
Last edited: