Member
Joined 2003
Thanks Paul. Odd indeed. BTW none of the three AIC drivers mention shorting rings, that wouldn't make much sense anyway. The 520 models, 10 and 12", are similar in voice coil size and lower BL, probably same motor. The 10NDA610 is different in its larger voice coil and much higher BL (20ish vs. 14ish for the 520 models), and half the xmax.
I not only know drivers.. I also know distributors/resellers (well.. some US anyway).
Your best bet in the US:
http://www.loudspeakersplus.com/html/18sound.html
Another place that is usually overpriced:
http://www.usspeaker.com/homepage.htm
Some of the prices seem toooo high though (considering dollar vs. euro), on the other hand I wouldn't expect shipping to be cheap from somwhere in Europe. I'd still check to see what they would charge though from these two resellers before purchasing abroad.
BTW, that 12NDA520 price (130 euro) seems VERY attractive!
Why the 12" vs the 10" of the AI 18 sound drivers.. The 12 inch has greater surface area = greater spl's.. and its a good bit more eff. at lower freq.s. It also means that it has more dispersion lower in freq.s. - in fact it appears to have a more constant -3db from 600 Hz to 1.4 kHz (-45 degrees). I wouldn't recomend a lowpass any higher than 1.4 kHz with a LR 4th order.
Absolutely GORGEOUS ribbon! At some point I may take you up on the offer. At this time though, I have non-diy projects taking up my time (..and I'm still thinking about a lower midrange radial project I've got "going".. don't know if anything will come of it though, and likely not anytime soon.)
Note that for your waveguide driver selection.. I think I remember the woodhorn reseller mention that his favorite of all the BMSpro compression drivers was the 4540ND.. (..and strangely the lower freq. bandwidth of the driver isn't much different than the 4552ND.) Again, about 1.3-1.4 seems about right. There are of course good drivers from radian and B&C to consider as well.
Your best bet in the US:
http://www.loudspeakersplus.com/html/18sound.html
Another place that is usually overpriced:
http://www.usspeaker.com/homepage.htm
Some of the prices seem toooo high though (considering dollar vs. euro), on the other hand I wouldn't expect shipping to be cheap from somwhere in Europe. I'd still check to see what they would charge though from these two resellers before purchasing abroad.
BTW, that 12NDA520 price (130 euro) seems VERY attractive!
Why the 12" vs the 10" of the AI 18 sound drivers.. The 12 inch has greater surface area = greater spl's.. and its a good bit more eff. at lower freq.s. It also means that it has more dispersion lower in freq.s. - in fact it appears to have a more constant -3db from 600 Hz to 1.4 kHz (-45 degrees). I wouldn't recomend a lowpass any higher than 1.4 kHz with a LR 4th order.
Absolutely GORGEOUS ribbon!
At some point I may take you up on the offer. At this time though, I have non-diy projects taking up my time (..and I'm still thinking about a lower midrange radial project I've got "going".. don't know if anything will come of it though, and likely not anytime soon.)Note that for your waveguide driver selection.. I think I remember the woodhorn reseller mention that his favorite of all the BMSpro compression drivers was the 4540ND.. (..and strangely the lower freq. bandwidth of the driver isn't much different than the 4552ND.) Again, about 1.3-1.4 seems about right. There are of course good drivers from radian and B&C to consider as well.
Member
Joined 2003
Those were two of my US sources for the 10M600. Unfortunately, neither lists the 10NDA or 12NDA. I saw the ~130 price on two different sites in Europe so the price is real, just need to find out what is going on with the Le spec for the 12.
Loudspeakersplus does list the 10NDA610...and that is actually the one I was asking you about (not the 12"). It's a little feeble on the bottom end, but...
For compression drivers, I'm thinking 4552ND...though both the 4550 and 4540ND come highly recommended.
Don't faint, but I have a question in to DSS to see if waveguide is available in 1.5" for the 4555. I like the idea of a lower compression ratio down low and can't hear much above 14k anyhow. Do you have a view of whether the 1.5" exit would dominate top-octave polars or would the waveguide dominate?
Loudspeakersplus does list the 10NDA610...and that is actually the one I was asking you about (not the 12"). It's a little feeble on the bottom end, but...
For compression drivers, I'm thinking 4552ND...though both the 4550 and 4540ND come highly recommended.
Don't faint, but I have a question in to DSS to see if waveguide is available in 1.5" for the 4555. I like the idea of a lower compression ratio down low and can't hear much above 14k anyhow. Do you have a view of whether the 1.5" exit would dominate top-octave polars or would the waveguide dominate?
Paul W said:
The driver itself will be the determining factor.
The problem I have with the larger drivers and waveguides generally - is actually the dispersion pattern. Specifically it isn't constant and particularly between 3-11 kHz. What happens here is typical beaming and image shift. Also the soundstage is pushed forward and effectivly compressed in depth. At its worst you'll move a little off-axis and everything seems to be comming from the loudspeaker you are closest to. As you might imagin - it isn't to great for tonal balance either. The "cure" then is a wide dispersion super tweeter like the Beyma CP25 (..even it isn't rated from more than about 18 kHz, and even thats optimistic) - and here you can see that it isn't a matter of freq. extension per-se, but rather dispersion vs. freq. extension.
the 610 10" just doesn't have the dispersion pattern for anything but a fronthorn/waveguide and despite it being a purist midrange you'll still need to cross it over low (or with a shallow filter) to blend the dispersion patteren of the compression driver/waveguide to its upper freq. response. In otherwords its a LOT more work to get it right, but its likely that IF thats acomplished - that the driver will be even better (..particularly that the driver will almost certainly have a cleaner linear decay between 1-2.5 kHz).
Just browsing the 18Sound catalogue, the 12NMB420 looks almost more attractive than the AIC drivers: more extended constant off axis directivity, later breakup, and similarly low Le (0.2 mH). It has Faraday rings instead of AIC, and the same low mass, high eff., etc etc. Seems like a case of, you can design a good driver with a whole set of different technologies.
18Sound 12NMB420 data sheet
18Sound 12NMB420 data sheet
Member
Joined 2003
Scott,
An Acapella is the only horn I've ever liked, so it might seem strange that I'm trying one. This is just a pleasant journey, so it isn't the end of the world if the result is NG.
To your point, in looking at the different flares in Hornresp, it does seem nearly all of them are beamy on the top end, except the Obulate Spheroid. The OS isn't perfect and DSS doesn't specify the flare, but the photo looks very close. At any rate, it has a good reputation and should make a good baseline for comparison.
Agree on the 610...just wanted to hear your thoughts.
MBK,
Good catch! The 12NMB looks like another viable candidate.
What does everyone think of the JBL 2012H?
An Acapella is the only horn I've ever liked, so it might seem strange that I'm trying one. This is just a pleasant journey, so it isn't the end of the world if the result is NG.
To your point, in looking at the different flares in Hornresp, it does seem nearly all of them are beamy on the top end, except the Obulate Spheroid. The OS isn't perfect and DSS doesn't specify the flare, but the photo looks very close. At any rate, it has a good reputation and should make a good baseline for comparison.
Agree on the 610...just wanted to hear your thoughts.
MBK,
Good catch! The 12NMB looks like another viable candidate.
What does everyone think of the JBL 2012H?
The 12NMB does look very nice..(good catch MBK.) Note - looking at this vs. the AI version, you would have to conclude that the AI version has a misprint on Le and that its .09 not .9. I do like the slightly extended dispersion of the AI version better and I'm not sure how the lack of spl off axis will play-in with a steep filter between 1-1.5 kHz and the compression driver w/waveguide.
The JBL - overpriced (based on the original retail) but with overall superb engineering, it doesn't have the dispersion pattern you want, it isn't as eff., and the frame isn't exactly unobtrusive (considering a dipole). I had looked at this one before - but getting it in the US (new) can be difficult.
The JBL - overpriced (based on the original retail) but with overall superb engineering, it doesn't have the dispersion pattern you want, it isn't as eff., and the frame isn't exactly unobtrusive (considering a dipole). I had looked at this one before - but getting it in the US (new) can be difficult.
BTW Paul what are your reasons for wanting a large diameter driver - in open baffle the dispersion is quite predictable anyway, to match with a compression driver... and a 10" or 12" might have a much less favorable linear distortion / decay behavior, especially considering you've had ribbons and line sources (didn't you have this gorgeous B&G line with Seas Excel at some point? I thought I saw this on this site somewhere).
I say all this tongue in cheek because I have also been toying with options in this direction for my open baffles. Vested interests in this thread...
I now have double 10"'s crossed in the 100-200 range (depends on the mood of the week, with active x-o settings can be changes easily), 6.5" up til 1700 Hz and then a dome tweeter. The baffle is a 7.5" deep U-frame, 24" wide, total 48" high, that's where the measurements came from. The sound is great, the SS8543 6.5" is extremely clean, the (cheap) Seas tweeter is never harsh etc.
So why change then?
- the woofers are older Vifas and though Q=0.5 at Fs 23 Hz they always seem a bit boomy to me - I suspect at higher excursions 2nd and 3rd harmonics might play a role in this. I'd guess new designs are cleaner, plus maybe a bit more Xmax.
- the system sounds extremely clean up to say 90-93 dB in-room. Beyond that distortions set in - be it power limitations (I use 3x LM3886 at 38V rails per side) or tweeter overexcursion, hard to tell. It would be nice to actually have a true 20 dB *headroom* at say 90 dB in-room SPL and that means, efficient (pro) drivers and horns.
Now initially I wanted to simply add a 15" pro woofer - the B&C15PS76 is available here, looks good, and is OK price wise. Better may exist but I'd rather save that USD250 on shipping.
Then I thought, why not take power demands off the fairly inefficient SS8543 on the low end and aim at crossing the 15" high, say 300 Hz.
Then I saw Earl Geddes' Summa design, which is a boxed 15" (he uses B&C15TBX100 crossed at 900 Hz into a horn of his design, oblate spheroid that is) which gave me the idea to just use a larger driver higher up. Ironically the Summa has practically identical dimensions to my baffle, except for a bit more depth. Nothing new on the planet. But higher than 1k is daring with a 15", even in dipole, and what if I don't like the horn. So I thought of a 10" or 12" taken beyond 1k, maybe leave the dome tweere for now, and keep the horn option. In dipole a single 12" might do down to 60-80 Hz with sizeable SPL capability, below that a closed box woofer naturally comes to mind - even allowing for experimenting mixing dipole and monople with a shallow slope, for a smooth dipole-cardioid-monopole gradient. Especially, considering that my measurements show my design isn't exactly dipole in the mids anyway.
And now my trade offs:
- how likely is one to be disappointed with the transparency (read, linear distortion performance) of a 12" pro midbass vs an excellent 6.5" hifi woofer
- is dipole -> monopole transition experimenting worth the added complexity of two 12"ers, one in dipole, one in monopole, add an x-o and amp channel..., compared to a simple 15" up to 300 Hz in dipole
- horn tweeter or not - even if the radiation patterns and power response are made to match, the highs will then be very directional. If anything I'd actually want more dispersion in mids and highs than I have now, for depth and naturalness. A horn tweeter would possibly turn out a clinical monitor sound, great for microscopy, not so great for the typical off axis listening around the house that I do. OTOH you can't get high SPL headroom from a dome tweeter... save for exotics made of unobtainium.
Not to threadjack, I just suspect we have similar thoughts and trade offs in mind, and ScottG sure has some ideas on this as well.
I say all this tongue in cheek because I have also been toying with options in this direction for my open baffles. Vested interests in this thread...
I now have double 10"'s crossed in the 100-200 range (depends on the mood of the week, with active x-o settings can be changes easily), 6.5" up til 1700 Hz and then a dome tweeter. The baffle is a 7.5" deep U-frame, 24" wide, total 48" high, that's where the measurements came from. The sound is great, the SS8543 6.5" is extremely clean, the (cheap) Seas tweeter is never harsh etc.
So why change then?
- the woofers are older Vifas and though Q=0.5 at Fs 23 Hz they always seem a bit boomy to me - I suspect at higher excursions 2nd and 3rd harmonics might play a role in this. I'd guess new designs are cleaner, plus maybe a bit more Xmax.
- the system sounds extremely clean up to say 90-93 dB in-room. Beyond that distortions set in - be it power limitations (I use 3x LM3886 at 38V rails per side) or tweeter overexcursion, hard to tell. It would be nice to actually have a true 20 dB *headroom* at say 90 dB in-room SPL and that means, efficient (pro) drivers and horns.
Now initially I wanted to simply add a 15" pro woofer - the B&C15PS76 is available here, looks good, and is OK price wise. Better may exist but I'd rather save that USD250 on shipping.
Then I thought, why not take power demands off the fairly inefficient SS8543 on the low end and aim at crossing the 15" high, say 300 Hz.
Then I saw Earl Geddes' Summa design, which is a boxed 15" (he uses B&C15TBX100 crossed at 900 Hz into a horn of his design, oblate spheroid that is) which gave me the idea to just use a larger driver higher up. Ironically the Summa has practically identical dimensions to my baffle, except for a bit more depth. Nothing new on the planet. But higher than 1k is daring with a 15", even in dipole, and what if I don't like the horn. So I thought of a 10" or 12" taken beyond 1k, maybe leave the dome tweere for now, and keep the horn option. In dipole a single 12" might do down to 60-80 Hz with sizeable SPL capability, below that a closed box woofer naturally comes to mind - even allowing for experimenting mixing dipole and monople with a shallow slope, for a smooth dipole-cardioid-monopole gradient. Especially, considering that my measurements show my design isn't exactly dipole in the mids anyway.
And now my trade offs:
- how likely is one to be disappointed with the transparency (read, linear distortion performance) of a 12" pro midbass vs an excellent 6.5" hifi woofer
- is dipole -> monopole transition experimenting worth the added complexity of two 12"ers, one in dipole, one in monopole, add an x-o and amp channel..., compared to a simple 15" up to 300 Hz in dipole
- horn tweeter or not - even if the radiation patterns and power response are made to match, the highs will then be very directional. If anything I'd actually want more dispersion in mids and highs than I have now, for depth and naturalness. A horn tweeter would possibly turn out a clinical monitor sound, great for microscopy, not so great for the typical off axis listening around the house that I do. OTOH you can't get high SPL headroom from a dome tweeter... save for exotics made of unobtainium.
Not to threadjack, I just suspect we have similar thoughts and trade offs in mind, and ScottG sure has some ideas on this as well.
Member
Joined 2003
I like the "boxless" sound of OB, but the dipole backwave bouncing off the front wall interferes with imaging. The furthest distance achievable in my room is about 6' from drivers to the front wall. (The room has been rotated 90 degrees from this photo...see avatar for current orientation.) So, with OBs, I always end up with felt behind the drivers to attenuate the backwave. (BTW, that is a DIY ribbon, not a BG driver!)
An externally hosted image should be here but it was not working when we last tested it.
With felt across the backwave, the dipole figure-8 radiation pattern doesn't hold, especially on the top end of a drivers range. The felt will likely continue, so pattern control needs to come from the drivers front radiation. The Summa box and the felt loaded OBs are both monopole at xo...and that's why a large diameter driver to transition to a WG.
Your concerns about large pro-drivers are why I'm not as brave as Earl and will try a 10-12" rather than 15" like Summa. W26's would be great except that violent breakup is too close to a 1k-2k xo...and that's why the current search.
This system will be built as stackable modules until I find a combination that works at least as well as the Seas WTW pictured above...then "permanent" baffles will be built. The BMS1850 pictured up top is the woofer module. BTW, my excuse for this whole exercise is to find equal or better performance in a smaller package.
I find fewer room reflections help imaging, but a waveguide or horn may take directionality too far...just have to try it and live with it for a while to know for sure.
Paul
Ah I thought so - I well remembered seeing this photo of your gorgeous system. I assumed it was a triple RD-28 or so, didn't realize those were DIY ribbons. Respect!
Imaging and OB's: while I also noticed that rear space is essential for an "open" sound (gives the essential room effects for "natural" reflection and delayed sound images) I have no problem with imaging. I only have space for 4' or so behind the speakers, and I wish it was 6'+ because every foot adds to transparency. But in my system imaging is painfully accurate, to the point that I find various mastering idiosyncrasies annoying, such as always putting the main singer slightly off to one side (or on old stereo, of panning instruments to just one side). Maybe your positioning problem has to do with the problems SL imputes to large planar dipoles, i.e. impedance irregularities due to the sensitive ribbons being hit by the room reflection, plus combing effects.
The felt works as a flow resistor and lowpass, and likely makes the pattern cardioid-like, which fixes the backwave problem but possibly reduces depth and airiness... though I also had in mind experimenting with the supposedly superior feltmetal suggestion by SL if I can get that stuff somewhere. This is supposed to allow a perfect termination from driver to free air in a transmission line (read: U-baffle). Anyway I hope I soon have time to do a decent 360 degrees measurement of my baffles, that should give you an idea as to how much of the backwave a simple large U-frame w/o felt is capable of eliminating. Your baffles must be a complete nightmare to measure I assume, due to size ...
Horns - in your large room more directivity may be a plus. I sit 8-10' from the speakers in a 16x22' room and I feel the speakers want a bigger room, and *more* reflections. In any case the farther away I am the more "right" they sound especially off axis in an L-extension of my living room (!).
BTW since you live in the US you might try the USD 129 + free shipping Dayton 12" or even 15" reference drivers. Triple demodulation rings, low Le, and aluminium diaphragm. Low eff. though and 4 Ohms. Measurements are scarce with Daytons but paper T/S data and tech details look good. I was interested but sadly, outside the US the monstrous shipping charges on 2x30 lbs bring the price into B&C TBX range.
FYI here is a photo of a previous incarnation of my OB's, then with a Diatone fullrange. The form factor has remained the same.
Imaging and OB's: while I also noticed that rear space is essential for an "open" sound (gives the essential room effects for "natural" reflection and delayed sound images) I have no problem with imaging. I only have space for 4' or so behind the speakers, and I wish it was 6'+ because every foot adds to transparency. But in my system imaging is painfully accurate, to the point that I find various mastering idiosyncrasies annoying, such as always putting the main singer slightly off to one side (or on old stereo, of panning instruments to just one side). Maybe your positioning problem has to do with the problems SL imputes to large planar dipoles, i.e. impedance irregularities due to the sensitive ribbons being hit by the room reflection, plus combing effects.
The felt works as a flow resistor and lowpass, and likely makes the pattern cardioid-like, which fixes the backwave problem but possibly reduces depth and airiness... though I also had in mind experimenting with the supposedly superior feltmetal suggestion by SL if I can get that stuff somewhere. This is supposed to allow a perfect termination from driver to free air in a transmission line (read: U-baffle). Anyway I hope I soon have time to do a decent 360 degrees measurement of my baffles, that should give you an idea as to how much of the backwave a simple large U-frame w/o felt is capable of eliminating. Your baffles must be a complete nightmare to measure I assume, due to size ...
Horns - in your large room more directivity may be a plus. I sit 8-10' from the speakers in a 16x22' room and I feel the speakers want a bigger room, and *more* reflections. In any case the farther away I am the more "right" they sound especially off axis in an L-extension of my living room (!).
BTW since you live in the US you might try the USD 129 + free shipping Dayton 12" or even 15" reference drivers. Triple demodulation rings, low Le, and aluminium diaphragm. Low eff. though and 4 Ohms. Measurements are scarce with Daytons but paper T/S data and tech details look good. I was interested but sadly, outside the US the monstrous shipping charges on 2x30 lbs bring the price into B&C TBX range.
FYI here is a photo of a previous incarnation of my OB's, then with a Diatone fullrange. The form factor has remained the same.
Attachments
My thoughts are:
I prefer the inverse of a normal hi-fi speaker when considering horizontal dispersion:
i.e. broad dispersion at high freq.s and narrow dispersion at lower freq.s - specifically however ALL IN PHASE if possible.
Ideally then (with something that could actually be built), it would be something like this:
radial driver *without* waveguide (or even a baffle) from about 200 Hz to 20+kHz; below that a hypercardoid with the small bit of rear (out-of-phase) signal acoustically dampened. Naturally I'd want the transition to be smooth so the top of the passband of the hypercardoid would be approaching omni and would gradually progress to a hypercardoid as freq.s decreased.
Now that doesn't take into account the possitive attributes of a dipole when *not* considering dispersion. It also doesn't include the fact that I don't really perceive the rear out-of-phase of a dipole near 200 Hz and below (..when listening in the forward plane). All things considered then - I'd probably opt for a dipole (and in fact personally do) between the freq.s of about 220 Hz down to about 50 Hz.
I do NOT like a dipole above this 200 Hz freq. (of course its a matter of freq. - the higher the worse it is) - it gives a spacious but limited depth perspective that has a "sameness of sound" effectivly "stamped" onto every recording. This effect is entirely dependent on the room the loudspeakers are in. (..the bigger the room the better things are.)
Subjectivly the midbass dipole gives me good image placement of instruments and vocals that reach into this range. I particularly like the fact that drum kits (and pianos) seem to be placed properly with respect to the overall soundstage and other "images". (..most monopoles are absolutely horrible at this - i.e they sound "forward" and "bunched up".)
As to radial dispersion for higher freq.s... IF there isn't a waveguide, IF the "front wall" of the listening room is fairly well dampended (and/or a good distance away from the speaker), IF the radial driver has little acoustic resistance (and preferably not pressure driven), IF the loudspeakers are moderatly close together, and hopefully IF the listener is fairly close to the loudspeakers - THEN imaging and soundstaging both seem considerably better and actually alter from recording to recording more than any other design I've listened to.
It *is* different than what were discussing here, but to an extent it does have application.
Of course none of the above is to discourage Paul from his current search - I still think that he can create an excellent loudspeaker with the path he is pursuing - it will be different, but in some respects it should be superior (like low non-linear distortion, high spl's, incredibly dynamic, etc.).
Finally, w.r.t. the 15" vs. the 12", vs. the 10". As far as linear decay is concerned I think this is highly dependent on the driver itself rather than the diameter per-se. (..though naturally modes within the diaphram are allowed to move lower in freq. because of the distance for vc to surround - and that starts asking a LOT of the surround (to dampen) for larger drivers.) What I *do* think is dependent on the driver's diameter is the horn loading effect of higher freq.s that gives the limited dispersion. IMO the larger the diameter and the greater the cone depth profile - the more planar like the sound becomes (all else equal). When its operating well, imaging becomes more "pin-point" in character and less 3D overall. When its done poorly - its still pin-point, but can give a "hooded" horn sound that is tonally wrong. Additionally, IF the excursion of the driver starts becoming something you can actually see up close - then depth almost always starts to suffer - and loudspeaker localization increases. (..also, complex recordings start loosing imaging definition.) These effects get worse the higher in freq. you go with such a driver. Note however that with the exception of the "planar" sound (and even that will be limited due to the lower opertaing point for the waveguide) - I don't think you'll have a problem considering the limited passband for the driver and the low lowpass filter.
I prefer the inverse of a normal hi-fi speaker when considering horizontal dispersion:
i.e. broad dispersion at high freq.s and narrow dispersion at lower freq.s - specifically however ALL IN PHASE if possible.
Ideally then (with something that could actually be built), it would be something like this:
radial driver *without* waveguide (or even a baffle) from about 200 Hz to 20+kHz; below that a hypercardoid with the small bit of rear (out-of-phase) signal acoustically dampened. Naturally I'd want the transition to be smooth so the top of the passband of the hypercardoid would be approaching omni and would gradually progress to a hypercardoid as freq.s decreased.
Now that doesn't take into account the possitive attributes of a dipole when *not* considering dispersion. It also doesn't include the fact that I don't really perceive the rear out-of-phase of a dipole near 200 Hz and below (..when listening in the forward plane). All things considered then - I'd probably opt for a dipole (and in fact personally do) between the freq.s of about 220 Hz down to about 50 Hz.
I do NOT like a dipole above this 200 Hz freq. (of course its a matter of freq. - the higher the worse it is) - it gives a spacious but limited depth perspective that has a "sameness of sound" effectivly "stamped" onto every recording. This effect is entirely dependent on the room the loudspeakers are in. (..the bigger the room the better things are.)
Subjectivly the midbass dipole gives me good image placement of instruments and vocals that reach into this range. I particularly like the fact that drum kits (and pianos) seem to be placed properly with respect to the overall soundstage and other "images". (..most monopoles are absolutely horrible at this - i.e they sound "forward" and "bunched up".)
As to radial dispersion for higher freq.s... IF there isn't a waveguide, IF the "front wall" of the listening room is fairly well dampended (and/or a good distance away from the speaker), IF the radial driver has little acoustic resistance (and preferably not pressure driven), IF the loudspeakers are moderatly close together, and hopefully IF the listener is fairly close to the loudspeakers - THEN imaging and soundstaging both seem considerably better and actually alter from recording to recording more than any other design I've listened to.
It *is* different than what were discussing here, but to an extent it does have application.
Of course none of the above is to discourage Paul from his current search - I still think that he can create an excellent loudspeaker with the path he is pursuing - it will be different, but in some respects it should be superior (like low non-linear distortion, high spl's, incredibly dynamic, etc.).
Finally, w.r.t. the 15" vs. the 12", vs. the 10". As far as linear decay is concerned I think this is highly dependent on the driver itself rather than the diameter per-se. (..though naturally modes within the diaphram are allowed to move lower in freq. because of the distance for vc to surround - and that starts asking a LOT of the surround (to dampen) for larger drivers.) What I *do* think is dependent on the driver's diameter is the horn loading effect of higher freq.s that gives the limited dispersion. IMO the larger the diameter and the greater the cone depth profile - the more planar like the sound becomes (all else equal). When its operating well, imaging becomes more "pin-point" in character and less 3D overall. When its done poorly - its still pin-point, but can give a "hooded" horn sound that is tonally wrong. Additionally, IF the excursion of the driver starts becoming something you can actually see up close - then depth almost always starts to suffer - and loudspeaker localization increases. (..also, complex recordings start loosing imaging definition.) These effects get worse the higher in freq. you go with such a driver. Note however that with the exception of the "planar" sound (and even that will be limited due to the lower opertaing point for the waveguide) - I don't think you'll have a problem considering the limited passband for the driver and the low lowpass filter.
Member
Joined 2003
Interesting...three people with three very different perspectives on imaging. Guess that helps the world go 'round!
Scott, when you say radial, are you talking about something like the big MBLs? I had the opportunity to listen to them earlier this year and was both impressed and disappointed. Mechanical marvels and an impressive sound but, IMO, vague imaging. My impression could have come from the way they worked in that relatively small room but perhaps our differing preferences. What technology are you going to use?
MBK, for broader horizontal dispersion than you have now, you might consider a narrow ribbon and a smaller mid (4-way ?). How did you handle the resonances between the parallel side panels in the U-baffles? Especially in the midrange, my best results have come from non-parallel walls, or extremely careful damping.
I'm in acquisition mode now. Sent an RFQ to three suppliers in Europe and two in the US. So far, one response from Europe. In order of price high-low 12NDA520, 10NDA610, 10NDA520, 12NMB420; plus 80 Euro shipping. Leaning toward the 12NDA520 or 10NDA520 because of the relatively gradual overall decline in off-axis response.
Scott, when you say radial, are you talking about something like the big MBLs? I had the opportunity to listen to them earlier this year and was both impressed and disappointed. Mechanical marvels and an impressive sound but, IMO, vague imaging. My impression could have come from the way they worked in that relatively small room but perhaps our differing preferences. What technology are you going to use?
MBK, for broader horizontal dispersion than you have now, you might consider a narrow ribbon and a smaller mid (4-way ?). How did you handle the resonances between the parallel side panels in the U-baffles? Especially in the midrange, my best results have come from non-parallel walls, or extremely careful damping.
I'm in acquisition mode now. Sent an RFQ to three suppliers in Europe and two in the US. So far, one response from Europe. In order of price high-low 12NDA520, 10NDA610, 10NDA520, 12NMB420; plus 80 Euro shipping. Leaning toward the 12NDA520 or 10NDA520 because of the relatively gradual overall decline in off-axis response.
I think we're not so far off in our perceptions, the various details of implementation might explain a lot, plus the nature of the room and listening distances. For instance I once tried a rear tweeter and wasn't too impressed overall - it sounded intriguing at first, but too fuzzy in the long run. OTOH the spaciousness it conferred was interesting.
Scott: quite agree with much of what you say. One word about bass though, as much as I like the clarity and definition of dipole low mid and bass, and yes, drums are near perfectly natural - sometimes I miss the chest pressure of a sealed bass ;-) .
Paul: my baffle is undamped inside. I initially expected some 1/4 wave resonance and the need for a notch filter or dampening, but couldn't find a trace in measurements or in tonal (im)balance. I did leave the top open though, a result from previous experience with narrower U-frames. See pic. Still, I tried some dampening in the baffle but this only managed to take away some apparently essential HF: it invariably sounded dull.
What the baffle *does* need, as any dipole I suppose, is to notch out the hump in responce just before rolloff. SL describes this in detail, and it certanly exists and must be taken out. Unfortunately I find it hard to measure because it *just* occurs where floor reflections mess up my measurements, 200-300 Hz, and only shows up at higher measurement distances, say 5 ft. But a certain muddyness in the upper bass is noticeable w/o a notch here. Eventually I used 1/3 oct. pink noise in-room subjective balance to get a better guess of where to chime in with a notch filter, and on how to interpret my impulse measurements which otherwise would have been inexplicable. The only good option here is measurement at least 5 ft high up in the air, and I can't do that. This took a lot of careful educated guesses and try and error.
Smaller midrange: in hindsight this might have helped with tweeter power handling, especially since I now consider running a 15" to 200-300 Hz anyway. The initial plan was to carry the 6.5" down to 100 Hz with the help of a baffle so large that floor gain sets in just when dipole rolloff starts. This calculation actually worked out - but I noticed that below 200-300 Hz the smalle mid doesn;t really hace any advantage and might as well use the sub higher up, it's physically close anyway. Dispersion wise I think with the measurements as they are, I am already at least at a cardioid. And even a smaller mid would not have more rear radiation if the pattern is in fact determined by the baffle shape.
Ribbon: afraid of the poor vertical response ;-) But, i had another design idea, combining a 15" with a ribbon long enough toishort enough to cross in the 1-1.5k range, and short enough to be aline source only down to 2-3k. Rationale, get line source character where it matters, in my case for broader soft spot, and avoid cost, size and image height exaggeration of a true and full line source. HF being more attenuated than mids in real life rooms, the lesser SPL falloff with distance in such a HF line vs the mids-lows might just be compensated by the higher room absorption. But I disgress.
Scott: also unclear about the radial. Anything like the top firing omnis with deflecting spheres? The question is, how different would this be from a rear firing tweeter - in those frequency ranges in- or out of phase doesn't really matter anymore... Generally I'd agree on your preference for high dispersion in the HF and upper mids, and less in the bass. I suspect this has to do with room effects - below 200 Hz you get the various room resonances amplifying your bass - if not broadband, then at least enough to make up for the missing power response. And HF is typically most attenuated in the average domestic environment.
General remarks about balance and power response: if you want both radiation patterns in the forward plane and total power response to mtach, and I think this is very worth while, things get a bit complicated. For instance a monopole tweeter on a large baffle like mine should count as essentially half-space, thus -3dB in power response from a true omni. A dipole or cardioid is -4.8 dB so one would think my in room power balance is slightly on the hot side. But this is not the subjective case, again, if anything I'd like more diffuse HF energy, for instance by a baffle-less tweeter to make it more omni, but then the power handling would get even worse.
Now in Paul's plan, if you match a horn tweeter to a dipole or cardioid mid, you don't have much room for a waveguide, essentially the power responce would roughly match if you take an additional 1.8 dB off the -3 dB for half-space, in other words you can only restrict that half space by about an additional 1/3, i.e. 120 degrees coverage in all directions, or am I missing something. BTW Paul depending on your sub x-o point why not also use a smaller diameter mid and just cross in the horn at a higher point when the radiation pattern starts to suit?
One remark about the bass / Scott - for me the main attraction of a large mid is that in dipole it could run from say 1.4k down to 80 Hz with a baffle as large as mine. That would keep the mid bass unfragmented and might help with the "missing" chest thumping pressure in dipoles, compared to monopoles, and for additional potential benefit one could then cross in a sealed woofer at that point, with large overlap. Drawback is complexity and the chance of never getting the integration right, while a large dipole, 15 or even 18", could be used without much fuss up to 200 or even 300 Hz. Crossing at 300 Hz has the added advantage of having the frequencies most vulnerable to floor reflections played at a low incidence angle close to the floor. Not that I think they're crucially distracting but might as well.
Scott: quite agree with much of what you say. One word about bass though, as much as I like the clarity and definition of dipole low mid and bass, and yes, drums are near perfectly natural - sometimes I miss the chest pressure of a sealed bass ;-) .
Paul: my baffle is undamped inside. I initially expected some 1/4 wave resonance and the need for a notch filter or dampening, but couldn't find a trace in measurements or in tonal (im)balance. I did leave the top open though, a result from previous experience with narrower U-frames. See pic. Still, I tried some dampening in the baffle but this only managed to take away some apparently essential HF: it invariably sounded dull.
What the baffle *does* need, as any dipole I suppose, is to notch out the hump in responce just before rolloff. SL describes this in detail, and it certanly exists and must be taken out. Unfortunately I find it hard to measure because it *just* occurs where floor reflections mess up my measurements, 200-300 Hz, and only shows up at higher measurement distances, say 5 ft. But a certain muddyness in the upper bass is noticeable w/o a notch here. Eventually I used 1/3 oct. pink noise in-room subjective balance to get a better guess of where to chime in with a notch filter, and on how to interpret my impulse measurements which otherwise would have been inexplicable. The only good option here is measurement at least 5 ft high up in the air, and I can't do that. This took a lot of careful educated guesses and try and error.
Smaller midrange: in hindsight this might have helped with tweeter power handling, especially since I now consider running a 15" to 200-300 Hz anyway. The initial plan was to carry the 6.5" down to 100 Hz with the help of a baffle so large that floor gain sets in just when dipole rolloff starts. This calculation actually worked out - but I noticed that below 200-300 Hz the smalle mid doesn;t really hace any advantage and might as well use the sub higher up, it's physically close anyway. Dispersion wise I think with the measurements as they are, I am already at least at a cardioid. And even a smaller mid would not have more rear radiation if the pattern is in fact determined by the baffle shape.
Ribbon: afraid of the poor vertical response ;-) But, i had another design idea, combining a 15" with a ribbon long enough toishort enough to cross in the 1-1.5k range, and short enough to be aline source only down to 2-3k. Rationale, get line source character where it matters, in my case for broader soft spot, and avoid cost, size and image height exaggeration of a true and full line source. HF being more attenuated than mids in real life rooms, the lesser SPL falloff with distance in such a HF line vs the mids-lows might just be compensated by the higher room absorption. But I disgress.
Scott: also unclear about the radial. Anything like the top firing omnis with deflecting spheres? The question is, how different would this be from a rear firing tweeter - in those frequency ranges in- or out of phase doesn't really matter anymore... Generally I'd agree on your preference for high dispersion in the HF and upper mids, and less in the bass. I suspect this has to do with room effects - below 200 Hz you get the various room resonances amplifying your bass - if not broadband, then at least enough to make up for the missing power response. And HF is typically most attenuated in the average domestic environment.
General remarks about balance and power response: if you want both radiation patterns in the forward plane and total power response to mtach, and I think this is very worth while, things get a bit complicated. For instance a monopole tweeter on a large baffle like mine should count as essentially half-space, thus -3dB in power response from a true omni. A dipole or cardioid is -4.8 dB so one would think my in room power balance is slightly on the hot side. But this is not the subjective case, again, if anything I'd like more diffuse HF energy, for instance by a baffle-less tweeter to make it more omni, but then the power handling would get even worse.
Now in Paul's plan, if you match a horn tweeter to a dipole or cardioid mid, you don't have much room for a waveguide, essentially the power responce would roughly match if you take an additional 1.8 dB off the -3 dB for half-space, in other words you can only restrict that half space by about an additional 1/3, i.e. 120 degrees coverage in all directions, or am I missing something. BTW Paul depending on your sub x-o point why not also use a smaller diameter mid and just cross in the horn at a higher point when the radiation pattern starts to suit?
One remark about the bass / Scott - for me the main attraction of a large mid is that in dipole it could run from say 1.4k down to 80 Hz with a baffle as large as mine. That would keep the mid bass unfragmented and might help with the "missing" chest thumping pressure in dipoles, compared to monopoles, and for additional potential benefit one could then cross in a sealed woofer at that point, with large overlap. Drawback is complexity and the chance of never getting the integration right, while a large dipole, 15 or even 18", could be used without much fuss up to 200 or even 300 Hz. Crossing at 300 Hz has the added advantage of having the frequencies most vulnerable to floor reflections played at a low incidence angle close to the floor. Not that I think they're crucially distracting but might as well.
Attachments
Radials..
I've heard the MBL under good conditions - a bit bloated in the midbass ( effective acoustic center is to large for that particular lower freq. ribbon "pod"), otherwise quite good with respect to imaging. Requires a huge amount of current though (w/ a "broken-in" powersupply/amp), otherwise imaging becomes less distinct.
What I'm going through right now..
1. Practically speaking this means an upward firing driver.
2. NO waveguide, it just interfers with the quality of the sound and least when in the range of a fundamental.
3. Driver diameter - the smaller the better for image specificity. It also helps extend the response up to a higher crossover value (considering how much spl-loss occurs at such an off-axis angle which is usually anwhere from 70-90 degrees off-axis).
4. The more eff. the driver the better (for any number of reasons). Principally however, there is significant spl-loss because radiation in the forward plane is now a 4pi enviorment as opposed to the normal IEC baffle standard of a 2pi enviorment.
5. I like as shallow a filter as possible (and as extended a driver as possible into higher freq.s), and realistically it DOES need a filter (crossover) for less deviation on the "stand-up, sit-down" test for freq. response linearity at and near crossover. The lower filter with the wider-band driver puts out a bit more upper freq. energy in-room that sounds a bit more realistic.
Effectivly then I need an extended bandwidth midrange or a fullrange driver that is no more than 4 inches in actual diameter (i.e. 5.25"). It needs to eff.. It should extend up to almost 2kHz on and off-axis up to +/-80-85 degrees. Crossover to an upper mid/tweeter will then neccesarily be near 2kHz. Right now I've got *potentially* just such a driver, I'd like it to be about 6db more eff. though. My laptop that I was using for measuring is "broken" - and so it will be sometime before I'm able to measure again, so I couldn't really say if the driver is suitable yet. I can say that it does *sound* suitable based on preliminary subjective evaluations during its "break-in". Until I get some measurements though, I'll not say what driver it is. I will say though that it is quite in-expensive and appears to be very well made (..just to keep people guessing ). Ideally I'd prefer something like the co-axial BMS drivers (without horns or waveguides), but that poses other limitations. Also there is the MaxFidelity 4" that appears to be quite nice. (..and frankly the "field" is VERY limited.)
That of course leaves from 2 kHz up..
I could do a number of things there for a wide dispersion design.. it doesn't have to be radial. OF the DIY capable radial designs I've seen - the Bolzano Villetri is both the most spohisticated and unsophisticated (i.e. it looks pretty mundane, but technically its pretty complex and overcomes several limitations of normal drivers utilized in a radial configuration). There are however other exotic drivers that actually ARE radial drivers. Still, I'm not pursuing either route for several reasons (..sound and cost-wise). Instead this is what I'll be pursuing (..note the "odd" dispersion pattern I'm looking for - and the reasons why):
http://www.diyaudio.com/forums/showthread.php?s=&threadid=50570&pagenumber=2
On the dipole midbass..
I agree MBK - its a balancing act between "punch" and proper image placement. I have found however that high force driver compensates for this ..to an extent. ALSO, (..similarly to what you have suggested - only in the inverse), IF you use a cardiod "sub" or a monopole "sub" with a 1st order high pass - it tends to add-in the "punch" without messing up the possitive qualities of the midbass dipole.
I've heard the MBL under good conditions - a bit bloated in the midbass ( effective acoustic center is to large for that particular lower freq. ribbon "pod"), otherwise quite good with respect to imaging. Requires a huge amount of current though (w/ a "broken-in" powersupply/amp), otherwise imaging becomes less distinct.
What I'm going through right now..
1. Practically speaking this means an upward firing driver.
2. NO waveguide, it just interfers with the quality of the sound and least when in the range of a fundamental.
3. Driver diameter - the smaller the better for image specificity. It also helps extend the response up to a higher crossover value (considering how much spl-loss occurs at such an off-axis angle which is usually anwhere from 70-90 degrees off-axis).
4. The more eff. the driver the better (for any number of reasons). Principally however, there is significant spl-loss because radiation in the forward plane is now a 4pi enviorment as opposed to the normal IEC baffle standard of a 2pi enviorment.
5. I like as shallow a filter as possible (and as extended a driver as possible into higher freq.s), and realistically it DOES need a filter (crossover) for less deviation on the "stand-up, sit-down" test for freq. response linearity at and near crossover. The lower filter with the wider-band driver puts out a bit more upper freq. energy in-room that sounds a bit more realistic.
Effectivly then I need an extended bandwidth midrange or a fullrange driver that is no more than 4 inches in actual diameter (i.e. 5.25"). It needs to eff.. It should extend up to almost 2kHz on and off-axis up to +/-80-85 degrees. Crossover to an upper mid/tweeter will then neccesarily be near 2kHz. Right now I've got *potentially* just such a driver, I'd like it to be about 6db more eff. though. My laptop that I was using for measuring is "broken" - and so it will be sometime before I'm able to measure again, so I couldn't really say if the driver is suitable yet. I can say that it does *sound* suitable based on preliminary subjective evaluations during its "break-in". Until I get some measurements though, I'll not say what driver it is. I will say though that it is quite in-expensive and appears to be very well made (..just to keep people guessing
). Ideally I'd prefer something like the co-axial BMS drivers (without horns or waveguides), but that poses other limitations. Also there is the MaxFidelity 4" that appears to be quite nice. (..and frankly the "field" is VERY limited.)That of course leaves from 2 kHz up..
I could do a number of things there for a wide dispersion design.. it doesn't have to be radial. OF the DIY capable radial designs I've seen - the Bolzano Villetri is both the most spohisticated and unsophisticated (i.e. it looks pretty mundane, but technically its pretty complex and overcomes several limitations of normal drivers utilized in a radial configuration). There are however other exotic drivers that actually ARE radial drivers. Still, I'm not pursuing either route for several reasons (..sound and cost-wise). Instead this is what I'll be pursuing (..note the "odd" dispersion pattern I'm looking for - and the reasons why):
http://www.diyaudio.com/forums/showthread.php?s=&threadid=50570&pagenumber=2
On the dipole midbass..
I agree MBK - its a balancing act between "punch" and proper image placement. I have found however that high force driver compensates for this ..to an extent. ALSO, (..similarly to what you have suggested - only in the inverse), IF you use a cardiod "sub" or a monopole "sub" with a 1st order high pass - it tends to add-in the "punch" without messing up the possitive qualities of the midbass dipole.
So I did a full 180 degree measurement of my OB's with all filters and x-o's in place. I used 1.2 m distance (4 ft) in the hope that all baffle effects would show up. Unfortunately the results are inconclusive - there must be too much wall bounce adding sound where there shouldn't be any. Everything is in a 6 db range from 200 Hz up (below that no hope with this setup, and even here the floor bounce already shows). Even the monopole tweeter is just 3 dB down at 20k ... ... at 180 degrees (!). Any other wiggles might just be due to variations in wall bounce when rotating the speaker. See exhibited FR graph. 0, 90 and 180 degrees in bold.
On the bright sides:
- I now got a rotating platform I made for the purpose, to repeat the same thing outdoors some time at 2 am when traffic quiets down
- if I take the graphs as a token of power response / in-room balance, this actually IS a good result.
Scott:
this is actually what I meant. I should write shorter posts, for clarity
On the bright sides:
- I now got a rotating platform I made for the purpose, to repeat the same thing outdoors some time at 2 am when traffic quiets down
- if I take the graphs as a token of power response / in-room balance, this actually IS a good result.
Scott:
this is actually what I meant. I should write shorter posts, for clarity
Attachments
Member
Joined 2003
MBK,
Oops, forgot to answer your question...sorry!
Rather than pick a "convenient" upper xo frequency, the reason is to control the radiation pattern as low in frequency as practical, and for minimal midrange cone excursion with an xo to the BMS woofer at 200Hz.
On the bass issue I have noticed, even EQd to the same in-room response, the BMS drivers (heavy cones, high-powered motors) have more "impact" than the relatively small Seas W26s. Don't claim to know the reason for the difference, but it has caused me to look up from my reading on several occasions.
Scott,
only now did I look at the Bolzano Villetri. Interesting idea, although the benefits are not quite clear to me w.r.t. simple up-firing components. You gain in efficiency while keeping identical acoustical centers, you lose in increased diffraction on-axis (especially tweeter) and the distances must be adjusted carefully to avoid standing waves. Still, interesting idea to try that with two opposing dome tweeters, maybe not in "empty space" but in the center of a wide angle (120 degrees) bidirectional horn to match dipole mid. One would have a notch in the rear though at x-o when the rear tweeter wave is out of phase with the rear mid dipole wave, unless the rear lowpass effect of a conventional dynamic mid can be found to have an exact -180 degrees region in x-o vicinity... Then again one could do the same thing with a compression driver firing upwards into a horizontally oriented bidirectional horn...
Paul: impact seems to be kind of a mystery, associated with some features but not associable with physics ;-) Similarly I wonder why my Vifa 10" sound "soft" to me in spite of "correct" parameters and response data. But yes, they are low mass and low BL (54 g, 8.5).
Anyway in other news I found a Singapore 18Sounds supplier that actually stocks the 15ND930 , neo motor with demodulation rings and a very open basket, good looking radiation pattern... to me beating the B&C15PS76 in most aspects. Supposedly used in the Yorkville (of Unity horn fame). I now tend towards this solution, in dipole down to 30-ish, or lower, if sound quality holds up at and below resonance. I can still recycle the Vifas in a sealed box below that, maybe that's where they ought to be anyway...
only now did I look at the Bolzano Villetri. Interesting idea, although the benefits are not quite clear to me w.r.t. simple up-firing components. You gain in efficiency while keeping identical acoustical centers, you lose in increased diffraction on-axis (especially tweeter) and the distances must be adjusted carefully to avoid standing waves. Still, interesting idea to try that with two opposing dome tweeters, maybe not in "empty space" but in the center of a wide angle (120 degrees) bidirectional horn to match dipole mid. One would have a notch in the rear though at x-o when the rear tweeter wave is out of phase with the rear mid dipole wave, unless the rear lowpass effect of a conventional dynamic mid can be found to have an exact -180 degrees region in x-o vicinity... Then again one could do the same thing with a compression driver firing upwards into a horizontally oriented bidirectional horn...
Paul: impact seems to be kind of a mystery, associated with some features but not associable with physics ;-) Similarly I wonder why my Vifa 10" sound "soft" to me in spite of "correct" parameters and response data. But yes, they are low mass and low BL (54 g, 8.5).
Anyway in other news I found a Singapore 18Sounds supplier that actually stocks the 15ND930 , neo motor with demodulation rings and a very open basket, good looking radiation pattern... to me beating the B&C15PS76 in most aspects. Supposedly used in the Yorkville (of Unity horn fame). I now tend towards this solution, in dipole down to 30-ish, or lower, if sound quality holds up at and below resonance. I can still recycle the Vifas in a sealed box below that, maybe that's where they ought to be anyway...
MBK said:
The biggest thing about Bolzano Villetri design is maintaining sp-level for a radial..
Some things to note:
1. the lower midbass is lower in height than most radials - allowing for something quite a bit less than a full 90 degrees off-axis. The upper midbass further augments this - effectivly allowing for a MUCH more extended high freq. response for the driver. (..though how the two interact for combing and their combined response would be VERY interesting to see.)
2. the same can be said for the tweeters. Because the lower tweeter is up considerably higher, its off-axis behaviour will likely extend fairly flat up to almost 8 kHz (depending on how good the morel unit is). As far as fundamentals are concerned, MOST instrumentals will not have a phasey sound as a result of strongly dominate room sound (i.e. the direct sound will still dominate). Additionally, this will keep imaging more precisely located irrespective of listener position. (..note that I would expect something like triangles to suffer.)
3. then there is the precise manner of combing and what it adds and subtracts to the overall subjective sound. (..for instance the midbass's may not even need a lowpass filter; ..also "floor bounce" may be effectivly negated with this design.)
Now I'd still expect the nearfield response to be elevated as freq.s decrease - rather like a BBC compensation curve #1, but a bit more aggressive. But when combined with the overall response at greater distances (say 3+meters and a normal room enviorment), chances are the speakers are pretty flat for most listening. And all of this is achieved with a relativly simple design that does not utilize "horn" loading or suffer from serious diffraction.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.