Member
Joined 2003
Different approach to large mid
Here is another approach to midrange pattern control to match with a large-waveguide tweeter.
The idea is to combine the better pure-midrange characteristics of relatively small drivers with the controlled off-axis response of a larger driver. Rather than exciting one large cone at the center, "drive" is distributed over a wide area. In this case, four RS-180s arranged in a square.
First, hi-res (long window) outdoor ground plane frequency response...
CSD...
Response at 0-15-30-45 degrees off-axis...
So, the concept works. On-axis response is smooth up to nearly 2k, CSD is so good it surprised me, and off-axis response is much like a 15". With very high midrange cone area, excursion related distortion will be low at any sane listening level.
I'll listen to these for a while as mids, but these particular speakers are actually intended for use as mid-bass modules in new side-channel speakers. With room gain from a floor-wall intersection, the (low-Q sealed box) bass response should work well.
It would be really interesting for someone to do this with true midrange drivers rather than the RS-180s shown here. Heck, even the RS-150s will go higher!
Or, maybe go the other way with 9 RS-180's in a two-way with a large WG compression driver...humm.
Here is another approach to midrange pattern control to match with a large-waveguide tweeter.
The idea is to combine the better pure-midrange characteristics of relatively small drivers with the controlled off-axis response of a larger driver. Rather than exciting one large cone at the center, "drive" is distributed over a wide area. In this case, four RS-180s arranged in a square.
First, hi-res (long window) outdoor ground plane frequency response...
An externally hosted image should be here but it was not working when we last tested it.
CSD...
An externally hosted image should be here but it was not working when we last tested it.
Response at 0-15-30-45 degrees off-axis...
An externally hosted image should be here but it was not working when we last tested it.
So, the concept works. On-axis response is smooth up to nearly 2k, CSD is so good it surprised me, and off-axis response is much like a 15". With very high midrange cone area, excursion related distortion will be low at any sane listening level.
I'll listen to these for a while as mids, but these particular speakers are actually intended for use as mid-bass modules in new side-channel speakers. With room gain from a floor-wall intersection, the (low-Q sealed box) bass response should work well.
It would be really interesting for someone to do this with true midrange drivers rather than the RS-180s shown here. Heck, even the RS-150s will go higher!
Or, maybe go the other way with 9 RS-180's in a two-way with a large WG compression driver...humm.
Paul,
interesting results. For sure the decay of many small drivers should have advantages over a big one. On the other hand the off axis response does not really hold up >900 Hz. Maybe another way to go would be to build a ring radiator out of many small drivers around a WG's HF unit... It would look like a huge ear I guess...
Well I finally got around to fiddle with ARTA to move ahead on producing CSD's (lest I find myself dishonoured due to prolonged inaction). For these tests I left the drivers in baffle, which has two drawbacks - 1. not a huge separation distance to the rear wave, ca. 50 cm path length difference at a measurement distance of 50cm, which limits gating to 1.5 ms or so, and even there I had to gate longer to get a minimal FFT sample size; 2. woofer to close to the floor, with euqal consequences for the gating. Basically, very restricted badwidth and lotsa artefacts. I post it anyway, for reference and comments, and hope to repeat in a proper large test baffle in the future.
Here is the SS8543 6.5", first.
interesting results. For sure the decay of many small drivers should have advantages over a big one. On the other hand the off axis response does not really hold up >900 Hz. Maybe another way to go would be to build a ring radiator out of many small drivers around a WG's HF unit... It would look like a huge ear I guess...
Well I finally got around to fiddle with ARTA to move ahead on producing CSD's (lest I find myself dishonoured due to prolonged inaction). For these tests I left the drivers in baffle, which has two drawbacks - 1. not a huge separation distance to the rear wave, ca. 50 cm path length difference at a measurement distance of 50cm, which limits gating to 1.5 ms or so, and even there I had to gate longer to get a minimal FFT sample size; 2. woofer to close to the floor, with euqal consequences for the gating. Basically, very restricted badwidth and lotsa artefacts. I post it anyway, for reference and comments, and hope to repeat in a proper large test baffle in the future.
Here is the SS8543 6.5", first.
Attachments
And here the 15ND930.
Unfortunately the range covers hardly anything in the passband, which ideally should be below 1.2k for the 15" and below 2k for the 6.5". Interestingly below 1.5k where the data start to matter wrt to a WG'd tweeter crossed low, both drivers show a fast initial decay (partially hidden by the following time slices) and then a rise, in the case of the 15" in fact two additional "waves". Now I am not sure if these are reflections (diffraction? floor?) or driver generated stored energy (especially those later peaks in the sub 1k region for the 15"). Anyway, to repeat with a more rigorous setup.
Unfortunately the range covers hardly anything in the passband, which ideally should be below 1.2k for the 15" and below 2k for the 6.5". Interestingly below 1.5k where the data start to matter wrt to a WG'd tweeter crossed low, both drivers show a fast initial decay (partially hidden by the following time slices) and then a rise, in the case of the 15" in fact two additional "waves". Now I am not sure if these are reflections (diffraction? floor?) or driver generated stored energy (especially those later peaks in the sub 1k region for the 15"). Anyway, to repeat with a more rigorous setup.
Attachments
Hi
The resonancies moving up in frequencie during decay are most likely due to reflections (floor, walls or cabinet).
Furthermore it seems you have set te time window in the time domain too late - cutting away for the first impulse rising slope and thus the FR completely.
For a wider frequencie range repeat your measurement with placing the speakers as far from floor and any wall as possible.
Use a a time window much longer than the 2.69ms now. ''I fYou use ARTA set the FFT points to the maximum of 1024 when sampling with 44100 Hz.
If you use a 24 bit soundcard you can easily increase the range of the diagramm to 30 dB and the decay time to about 10 -30 ms
If measuring only a single speaker you can place the mic very close - even 1 cm in front of the diaphragm.
This way you should be able to get results similar than here:
http://www.diyaudio.com/forums/showthread.php?postid=1202358#post1202358
"http://www.diyaudio.com/forums/showthread.php?postid=1202358#post1202358"
greetings
Michael
The resonancies moving up in frequencie during decay are most likely due to reflections (floor, walls or cabinet).
Furthermore it seems you have set te time window in the time domain too late - cutting away for the first impulse rising slope and thus the FR completely.
For a wider frequencie range repeat your measurement with placing the speakers as far from floor and any wall as possible.
Use a a time window much longer than the 2.69ms now. ''I fYou use ARTA set the FFT points to the maximum of 1024 when sampling with 44100 Hz.
If you use a 24 bit soundcard you can easily increase the range of the diagramm to 30 dB and the decay time to about 10 -30 ms
If measuring only a single speaker you can place the mic very close - even 1 cm in front of the diaphragm.
This way you should be able to get results similar than here:
http://www.diyaudio.com/forums/showthread.php?postid=1202358#post1202358
"http://www.diyaudio.com/forums/showthread.php?postid=1202358#post1202358"
greetings
Michael
Member
Joined 2003
MBK,
To use the 2x2 configuration (at least of RS-180s) as the mid in the main L&R speakers, they would be crossed below 1k to either a large ribbon or a larger format compression driver. The relatively quick decay we're seeing with the multiple small mids is the only thing I've been able to measure (other than directivity) that clearly differentiates big ribbons from other drivers...so the quick decay 2x2 mid is tempting. I'll know just how tempting after a few days of listening
The 1" compression drivers are solid performers but, at least for me, they just don't go low enough to mate well with a directional mid. So, I think you were on the right track at 1.4", or even 2". Sims show even a 2" isn't too bad at the 14k limit of my abused ears...plus I am much more critical of midrange quality than extreme top end dispersion. Dunno if it is true but all else equal, a 2" exit seems like it should offer lower distortion via lower air compression. Perhaps a very large 100 degree OS to control the low end, gradually handing off to a naturally narrowing 2" top end. Opinions? Scott?
I asked Radian for unsmoothed FR and distortion measurements but all they have said so far is "1/3 octave smoothing is the industry standard". BMS doesn't even respond to email. So, although the 18 Sound compression drivers are more expensive, the company is much more open with information...they are looking better and better all the time.
I hope you and Michael get ARTA sorted out. The 15ND930 should be quite interesting!
Paul
To use the 2x2 configuration (at least of RS-180s) as the mid in the main L&R speakers, they would be crossed below 1k to either a large ribbon or a larger format compression driver. The relatively quick decay we're seeing with the multiple small mids is the only thing I've been able to measure (other than directivity) that clearly differentiates big ribbons from other drivers...so the quick decay 2x2 mid is tempting. I'll know just how tempting after a few days of listening
The 1" compression drivers are solid performers but, at least for me, they just don't go low enough to mate well with a directional mid. So, I think you were on the right track at 1.4", or even 2". Sims show even a 2" isn't too bad at the 14k limit of my abused ears...plus I am much more critical of midrange quality than extreme top end dispersion. Dunno if it is true but all else equal, a 2" exit seems like it should offer lower distortion via lower air compression. Perhaps a very large 100 degree OS to control the low end, gradually handing off to a naturally narrowing 2" top end. Opinions? Scott?
I asked Radian for unsmoothed FR and distortion measurements but all they have said so far is "1/3 octave smoothing is the industry standard". BMS doesn't even respond to email. So, although the 18 Sound compression drivers are more expensive, the company is much more open with information...they are looking better and better all the time.
I hope you and Michael get ARTA sorted out. The 15ND930 should be quite interesting!
Paul
Michael,
I set the impulse window to start before the initial rise of the impulse. The FR when viewed as pure FR looked OK, but you are right that the t0 FR of the CSD had suspicious "holes". Maybe the CSD windowing compounded with the initial impulse windowing and ended up chopping away data.
To get a 1024 bit FFT CSD I would need at least a 10+ ms window: ARTA required a 2.667ms window to get 128 samples for a 256 FFT so this needs to be x4. But that's 3.43 m path length difference from direct to rear wave / reflected sound, i.e. a wall mounted speaker or mounted in a baffle of 3.43 m diameter... or I need to build a box to the backside (your beautiful graphs were done in box as I understand). I can't physically lift my speakers, that's for sure.
All of this is feasible but the most labour saving option (i.e. the most likely to actually get done) is to fill the space of a door, away from corners, with a door sized baffle open to the rear... I am studying the doors in my house right now
But first I'll try a measurement at cone instead of 50 cm, maybe the reflections can be ignored at such a short distance.
Paul: also still pondering about 1.4" or 2". My thoughts are similar - on one hand the less HF demands on the 15", the better, and if so why not choose the largest option in order to keep the LF stress on the compression driver low. Basically, once you go huge, why not go huge all the way . 2" vs. 1.4" doubles the Sd and I can't even hear beyond 14k. A supertweeter might be added easily, and cheaply, if so desired. On the other hand the quoted FR extension, at least for 18Sound drivers, is similar for 1.4" and 2" models (-10dB at 500 Hz) , so either option is unlikely to be useable below say 750 Hz or so (not to mention the horn size for say, 400 Hz x-o). Yet this is still enough to go below 900 Hz which according to one of Griesinger's papers seems to be the one of the various limits for directional clue mechanisms of the ear.
I set the impulse window to start before the initial rise of the impulse. The FR when viewed as pure FR looked OK, but you are right that the t0 FR of the CSD had suspicious "holes". Maybe the CSD windowing compounded with the initial impulse windowing and ended up chopping away data.
To get a 1024 bit FFT CSD I would need at least a 10+ ms window: ARTA required a 2.667ms window to get 128 samples for a 256 FFT so this needs to be x4. But that's 3.43 m path length difference from direct to rear wave / reflected sound, i.e. a wall mounted speaker or mounted in a baffle of 3.43 m diameter... or I need to build a box to the backside (your beautiful graphs were done in box as I understand). I can't physically lift my speakers, that's for sure.
All of this is feasible but the most labour saving option (i.e. the most likely to actually get done) is to fill the space of a door, away from corners, with a door sized baffle open to the rear... I am studying the doors in my house right now
But first I'll try a measurement at cone instead of 50 cm, maybe the reflections can be ignored at such a short distance.Paul: also still pondering about 1.4" or 2". My thoughts are similar - on one hand the less HF demands on the 15", the better, and if so why not choose the largest option in order to keep the LF stress on the compression driver low. Basically, once you go huge, why not go huge all the way
. 2" vs. 1.4" doubles the Sd and I can't even hear beyond 14k. A supertweeter might be added easily, and cheaply, if so desired. On the other hand the quoted FR extension, at least for 18Sound drivers, is similar for 1.4" and 2" models (-10dB at 500 Hz) , so either option is unlikely to be useable below say 750 Hz or so (not to mention the horn size for say, 400 Hz x-o). Yet this is still enough to go below 900 Hz which according to one of Griesinger's papers seems to be the one of the various limits for directional clue mechanisms of the ear.
Member
Joined 2003
MBK,
It seems our ears have same 14k cutoff frequency
. The 1.4" and 2" 18 Sound drivers all have 3" diaphragms, but I'm wondering if the lower air compression and turbulence of the 2" exit is a good trade-off vs the smaller source 
.
You might try ground-plane measurement of the woofer by tilting the baffle so both edges of the driver are equal distance to a microphone on the floor. The concrete pad I use has a pebble surface so, when looking at HF GP, I use a half-sheet of MDF to smooth the surface from the driver to beyond the mic...something like that might work over carpet.
Paul
It seems our ears have same 14k cutoff frequency
. The 1.4" and 2" 18 Sound drivers all have 3" diaphragms, but I'm wondering if the lower air compression and turbulence of the 2" exit is a good trade-off vs the smaller source .You might try ground-plane measurement of the woofer by tilting the baffle so both edges of the driver are equal distance to a microphone on the floor. The concrete pad I use has a pebble surface so, when looking at HF GP, I use a half-sheet of MDF to smooth the surface from the driver to beyond the mic...something like that might work over carpet.
Paul
Paul,
if it's any consolation, at least I have proof that it's from a young age, ears' FR first tested at 17 and before I ever went to clubs or anything loud (hum. I guess this unmasks the geek) - and yet, already the same as now aged 39.
1.4" vs 2": Duh. I could've used my brain, of course the Sd is the same. In which case the 1.4" should have an efficiency advantage and as you say, an un-guessable diffraction/compression/turbulence disadvantage. Dispersion falls off linearly with diameter, so the 1.4" should start beaming at 3200 and the 2" at 2200 Hz if one goes by the standard cutoff formula of F(beaming)=wavelength of 3x diaphragm diameter. This I think is more of a consideration than top end extension which by the graphs of 18Sound is not shabby at all for the 2". In a waveguide this matters less since the radiation is constrained, so the directivity increae is pushed higher. Still, they seem to tout their NSD1480N as the top high quality driver with all the technology. In the 2" range their favourite seems to be the ND2060A.
CSD: slowly getting there. The SS8543 plot starts looking believable, though it still has a funny "shoulder". And I still can't quite believe the woofer plot, it must be too close to the speaker's foot even when measrued at the cone. Good point about the groundplane technique, this is what I use for regular FR and EQ, all else is hopeless in the critical 100-400 Hz range. So that might work for the woofer.
One trouble I have though is that the exact position of the impulse window has way too much an influence on the CSD. Set the start a bit too close to the impulse start and the FR gets truncated for the first slices. Set it a bit too early and there is no decay. And between these extremes there are just a few FFT samples, that is, microseconds. Will the correct setting please stand up??
Something else that begs the question, is - I (we) might get a good CSD comparison for raw unimpeded drivers in the end, but does the measurement problem not show that the plain driver CSD might be a very unrealistic assumption? Just plain baffle effects seem to have a large influence on the CSD, and in the case of a dipole, the rear wave seamlessly joins in right after the direct sound has decayed. The smaller the baffle, the earlier, right in the time slice magnitude of typical CSD plots. Still, dipoles sound "transparent".
if it's any consolation, at least I have proof that it's from a young age, ears' FR first tested at 17 and before I ever went to clubs or anything loud (hum. I guess this unmasks the geek) - and yet, already the same as now aged 39.
1.4" vs 2": Duh. I could've used my brain, of course the Sd is the same. In which case the 1.4" should have an efficiency advantage and as you say, an un-guessable diffraction/compression/turbulence disadvantage. Dispersion falls off linearly with diameter, so the 1.4" should start beaming at 3200 and the 2" at 2200 Hz if one goes by the standard cutoff formula of F(beaming)=wavelength of 3x diaphragm diameter. This I think is more of a consideration than top end extension which by the graphs of 18Sound is not shabby at all for the 2". In a waveguide this matters less since the radiation is constrained, so the directivity increae is pushed higher. Still, they seem to tout their NSD1480N as the top high quality driver with all the technology. In the 2" range their favourite seems to be the ND2060A.
CSD: slowly getting there. The SS8543 plot starts looking believable, though it still has a funny "shoulder". And I still can't quite believe the woofer plot, it must be too close to the speaker's foot even when measrued at the cone. Good point about the groundplane technique, this is what I use for regular FR and EQ, all else is hopeless in the critical 100-400 Hz range. So that might work for the woofer.
One trouble I have though is that the exact position of the impulse window has way too much an influence on the CSD. Set the start a bit too close to the impulse start and the FR gets truncated for the first slices. Set it a bit too early and there is no decay. And between these extremes there are just a few FFT samples, that is, microseconds. Will the correct setting please stand up??
Something else that begs the question, is - I (we) might get a good CSD comparison for raw unimpeded drivers in the end, but does the measurement problem not show that the plain driver CSD might be a very unrealistic assumption? Just plain baffle effects seem to have a large influence on the CSD, and in the case of a dipole, the rear wave seamlessly joins in right after the direct sound has decayed. The smaller the baffle, the earlier, right in the time slice magnitude of typical CSD plots. Still, dipoles sound "transparent".
Member
Joined 2003
My hearing started out at about 20k, but too much time around race cars, rock, and other diversions took the toll! I really feel very lucky to still hear 14k.
The NSD1480N has their new "titanium nitride" treated diaphragm so, along with the 2.2T motor, it is their latest and greatest. The ND2080 also has the 2.2T motor but no nitride on the titanium dome. I asked when they were going to produce the nitride version in a 2" format...the response was that they would build them for me on special order! Of course I don't have a price yet
.
Today I received 2nd and 3rd harmonic Audio Precision measurements at 1W and 10W for the 2060, 2060A, 2080, and even the (non-existent) NSD2080N. Except for a weird glitch at about 670Hz, the NSD is top dog (I've asked about the glitch). The standard 2080 is next best. The 2060s are good at 1W but the titanium version has a little trouble on the bottom; the aluminum dome breaks up on the top end. Overall, 1W third harmonic performance for all of the 2" drivers is simply outstanding at less than 0.25% from 500Hz-10k! Oh, did I mention 0.25% is 110db @ 1M? Second harmonic performance is not nearly as impressive, but not so bad for 110 and 120db!
Since it appears they simply forgot the 1.4", I asked about those again and will let you know the result...unless my contact tires of the incessant questions
If the CSD shows a lot of interference from the baffle, I think we need to think about fixing the baffle
The NSD1480N has their new "titanium nitride" treated diaphragm so, along with the 2.2T motor, it is their latest and greatest. The ND2080 also has the 2.2T motor but no nitride on the titanium dome. I asked when they were going to produce the nitride version in a 2" format...the response was that they would build them for me on special order! Of course I don't have a price yet
.Today I received 2nd and 3rd harmonic Audio Precision measurements at 1W and 10W for the 2060, 2060A, 2080, and even the (non-existent) NSD2080N. Except for a weird glitch at about 670Hz, the NSD is top dog (I've asked about the glitch). The standard 2080 is next best. The 2060s are good at 1W but the titanium version has a little trouble on the bottom; the aluminum dome breaks up on the top end. Overall, 1W third harmonic performance for all of the 2" drivers is simply outstanding at less than 0.25% from 500Hz-10k! Oh, did I mention 0.25% is 110db @ 1M? Second harmonic performance is not nearly as impressive, but not so bad for 110 and 120db!
Since it appears they simply forgot the 1.4", I asked about those again and will let you know the result...unless my contact tires of the incessant questions
If the CSD shows a lot of interference from the baffle, I think we need to think about fixing the baffle
Hi Paul, just talked to Mike at Radian on the phone yesterday. Without breaking any NDA's (they sell drivers to several $80,000/pr high-end vendors), the 850-PB 2" is favored for its relaxed, open midrange and performance almost up the top of the range (it is frequently used with supertweeters in ultra-fi applications). When ultimate HF extension is more important, the smaller 1.4" 835-PB is favored for slightly sweeter HF at the expense of a bit less midrange power-handling.
Apparently, the real differences sonically are the compression ratio and diaphragm material - higher compression ratios are more "focussed" and intense in the midrange, and lower compression ratios are more relaxed sounding. With a 3" diaphragm and a 2" exit, the 850-PB has a lower compression ratio than the 835-PB, with its 1.4" exit. This alters the sonic presentation.
I asked specifically about the 950-PB (4" diaphragm & neo) vs the 850-PB (3" diaphragm & ceramic) and the implied much lower distortion for the 950-PB mentioned on the web-page, but Mike said the 950-PB is only about 2.5 to 3 dB lower than the 850-PB - and both are many DB lower than the competition from JBL and TAD. The larger diaphragm is why the 950-PB has a somewhat lumpier extreme HF compared to the smoothness of the 850-PB.
Mike mentioned that some Ti diaphragms have a distortion peak around 2 kHz, compared to aluminum. The people who have commented negatively on the sonics of Ti are reacting to this Ti abberation in the midrange. Reading between the lines, Ti is selected by CD manufacturers for extreme HF performance - at the expense of midrange, where high-purity aluminum is at its best. According to Mike, Beryllium is (marginally) best, but is notoriously difficult to fabricate and is apparently prone to sudden failure.
I didn't realize this, but the 5312 coax (the one Mike feels is their most advanced and best-sounding coax) actually uses a 2" exit CD from Radian's 651PB/760PB/850PB compression-driver series - not the smaller-exit CD's used by other coax vendors. Yes, Radian makes the Hemp Acoustics series of coaxes, which use non-Radian hemp cones and slightly different CD diaphragms.
Apparently, the real differences sonically are the compression ratio and diaphragm material - higher compression ratios are more "focussed" and intense in the midrange, and lower compression ratios are more relaxed sounding. With a 3" diaphragm and a 2" exit, the 850-PB has a lower compression ratio than the 835-PB, with its 1.4" exit. This alters the sonic presentation.
I asked specifically about the 950-PB (4" diaphragm & neo) vs the 850-PB (3" diaphragm & ceramic) and the implied much lower distortion for the 950-PB mentioned on the web-page, but Mike said the 950-PB is only about 2.5 to 3 dB lower than the 850-PB - and both are many DB lower than the competition from JBL and TAD. The larger diaphragm is why the 950-PB has a somewhat lumpier extreme HF compared to the smoothness of the 850-PB.
Mike mentioned that some Ti diaphragms have a distortion peak around 2 kHz, compared to aluminum. The people who have commented negatively on the sonics of Ti are reacting to this Ti abberation in the midrange. Reading between the lines, Ti is selected by CD manufacturers for extreme HF performance - at the expense of midrange, where high-purity aluminum is at its best. According to Mike, Beryllium is (marginally) best, but is notoriously difficult to fabricate and is apparently prone to sudden failure.
I didn't realize this, but the 5312 coax (the one Mike feels is their most advanced and best-sounding coax) actually uses a 2" exit CD from Radian's 651PB/760PB/850PB compression-driver series - not the smaller-exit CD's used by other coax vendors. Yes, Radian makes the Hemp Acoustics series of coaxes, which use non-Radian hemp cones and slightly different CD diaphragms.
Compression Drivers
The JBL 2441's I have sound better with the radian diaphragms then the JBL aluminum 'diamonds' - BUT in the modified Beyma 15" 15DX coaxial (built in 4" diaphragm, 2" exit, top plate shaved to get away from titanium -yech) the JBL's sound better. These coax's are absolutely fantastic sounding with big back loaded horns and either aluminum diaphragm = titanium is bad
TAD's are not better, just different IME- I have had several 4001, 4002, 2001s -
Radian is from Emilar - Emilar made the EC600 - (Six inch aluminum shown in the bass horns above!), 320 and sweet sounding 175's - NICE DRIVERS - all of them - all aluminum with the mylar surround Radian now uses. I don't have any experience yet with the new Radian
A four inch driver is going to be better in the midrange, the 3" is a compromise between the 4" midrange and 2" diaphragm's treble
I prefer the large format drivers with a tweeter to help above 10k
Phenolic can even be better..........
Aren't these 6" diaphragms beautiful?
The JBL 2441's I have sound better with the radian diaphragms then the JBL aluminum 'diamonds' - BUT in the modified Beyma 15" 15DX coaxial (built in 4" diaphragm, 2" exit, top plate shaved to get away from titanium -yech) the JBL's sound better. These coax's are absolutely fantastic sounding with big back loaded horns and either aluminum diaphragm = titanium is bad
TAD's are not better, just different IME- I have had several 4001, 4002, 2001s -
An externally hosted image should be here but it was not working when we last tested it.
Radian is from Emilar - Emilar made the EC600 - (Six inch aluminum shown in the bass horns above!), 320 and sweet sounding 175's - NICE DRIVERS - all of them - all aluminum with the mylar surround Radian now uses. I don't have any experience yet with the new Radian
A four inch driver is going to be better in the midrange, the 3" is a compromise between the 4" midrange and 2" diaphragm's treble
I prefer the large format drivers with a tweeter to help above 10k
Phenolic can even be better..........
An externally hosted image should be here but it was not working when we last tested it.
Aren't these 6" diaphragms beautiful?
Paul,
wow, this is quite impressive. Also, very nice of them to reply with such detailed data. I didn't dare to ask, who knows in the end one is always likely to enjoy the top $$$ model the most
So the 2060's have a bit more trouble at 10W... that would be above 110 dB The data on the NSD1480N should be interesting too, from the data sheets 3rd harmonic and top end smoothness should benefit.
All this is very funny if you take a step back and look at what we're discussing here - the much reviled HD numbers for instance, from metal cones, at extreme outputs etc. Not subjects that would appear in regular audiophile talk. But from experience this does seem to matter a great deal, especially in its IMD consequences. Over the years I more and more come to like solid and massive engineering. When I think that 15 years ago or so I was "experimenting" with different types of spikes and stands to put under store-bought bookshelf speakers in the hope to hear improvements, talk about the wrong tree to bark at. Ah well. In my defense I must say, I never heard significant differences and gave up on this line of thought
wow, this is quite impressive. Also, very nice of them to reply with such detailed data. I didn't dare to ask, who knows in the end one is always likely to enjoy the top $$$ model the most
So the 2060's have a bit more trouble at 10W... that would be above 110 dB
The data on the NSD1480N should be interesting too, from the data sheets 3rd harmonic and top end smoothness should benefit. All this is very funny if you take a step back and look at what we're discussing here - the much reviled HD numbers for instance, from metal cones, at extreme outputs etc. Not subjects that would appear in regular audiophile talk. But from experience this does seem to matter a great deal, especially in its IMD consequences. Over the years I more and more come to like solid and massive engineering. When I think that 15 years ago or so I was "experimenting" with different types of spikes and stands to put under store-bought bookshelf speakers in the hope to hear improvements, talk about the wrong tree to bark at. Ah well. In my defense I must say, I never heard significant differences and gave up on this line of thought
Also a big WOW! from me too, those pix and that post you wrote are great, many thanks, Magnetar. Wonderful to hear from someone that's "been there" with equipment on this plane of performance.
As Magnetar probably already knows - but it was news to me - Radian/Emilar was founded by a metallurgist, so I'd guess they must know a thing or two about metal diaphragms. Mike pretty much said the same thing about Titanium as Magnetar - yuk! It must be one of those measure-good, sound-bad things.
Magnetar, since you brought up supertweeters, which are the well-regarded ones for use with large-format CD's? I was thinking something like the RAAL ribbons, but maybe ribbons just don't have the dynamics to keep up with the kind of compression drivers you're talking about.
As Magnetar probably already knows - but it was news to me - Radian/Emilar was founded by a metallurgist, so I'd guess they must know a thing or two about metal diaphragms. Mike pretty much said the same thing about Titanium as Magnetar - yuk! It must be one of those measure-good, sound-bad things.
Magnetar, since you brought up supertweeters, which are the well-regarded ones for use with large-format CD's? I was thinking something like the RAAL ribbons, but maybe ribbons just don't have the dynamics to keep up with the kind of compression drivers you're talking about.
Lynn Olson said:
I have had good luck with ESG ribbons with high efficiency direct radiators but they just aren't enough for a horn loaded midrange system - too fragile and they tend to run out of excursion and sound brittle when pushed too hard or too low to have the ease of dynamics you get with a horn.
There are many good horn tweeters out there. My favorites that are not to expensive are the Beyma CP21 slot (aluminum) and the little Eminence APT80 (phenolic) - Normally these sound best with a 3 rd order network or if not pushed too hard a single cap. The Beyma has more sparke and the Eminence is smooooother
A good compression driver to look at that is readily available and requires no additional tweeter is the Beyma CP380 . These use a phenolic dome and are really nice in the upper mids for a small format driver. They sound very nice in the little 18 Sound XT120 horns. That combo is good 1500 to 16K with a little midrange padding. - I also have used them as low as 800 Hz with excellent results in 300 Hz round tractrix horns.
Smoothed CP380 response plot: (note low 3rd harmonic distortion- Sweet!
Member
Joined 2003
Lynn,
Yep, Mike is the guy I've been talking with about the 835 and 850. I received another email late yesterday saying they would provide the info after they run the tests in a day or two.
The 18S distortion plots support what Mike told you about aluminum vs titanium in the midrange. The titanium domes have a 2nd harmonic peak just below 1k...the aluminum domes also peak at the same spot but at a lower level. Elsewhere, the materials are either equal or titanium wins. I really don't care about the aluminum/titanium dome material...like you, the results are what counts.
Magnetar,
Those big domes are beautiful! However, they sorta remind me of a McLaren I have on my wall but will never buy
. How high will they perform well? Agree on the supertweeter. I'll make provisions to add one if necessary.
MBK,
In looking at the photos of the 14xx and 20xx drivers, it seems possible that the 20 series is simply a 14 with a throat adapter. If so, there is probably more flexibility in just buying the 14 and working the throat of the waveguide.
18S has now confirmed the NSD2080N uses the 1480 diaphragm...so they may have done a good job of platform planning by mixing a couple of motors with 3 domes and a throat adaptor. It seems we could switch between aluminum and the two titanium domes at will. Maybe you could check all of this with your local distributor? I'll try from my end as well.
Paul
Yep, Mike is the guy I've been talking with about the 835 and 850. I received another email late yesterday saying they would provide the info after they run the tests in a day or two.
The 18S distortion plots support what Mike told you about aluminum vs titanium in the midrange. The titanium domes have a 2nd harmonic peak just below 1k...the aluminum domes also peak at the same spot but at a lower level. Elsewhere, the materials are either equal or titanium wins. I really don't care about the aluminum/titanium dome material...like you, the results are what counts.
Magnetar,
Those big domes are beautiful! However, they sorta remind me of a McLaren I have on my wall but will never buy
. How high will they perform well? Agree on the supertweeter. I'll make provisions to add one if necessary.MBK,
In looking at the photos of the 14xx and 20xx drivers, it seems possible that the 20 series is simply a 14 with a throat adapter. If so, there is probably more flexibility in just buying the 14 and working the throat of the waveguide.
18S has now confirmed the NSD2080N uses the 1480 diaphragm...so they may have done a good job of platform planning by mixing a couple of motors with 3 domes and a throat adaptor. It seems we could switch between aluminum and the two titanium domes at will. Maybe you could check all of this with your local distributor? I'll try from my end as well.
Paul
Paul, good point about the diaphragms and throats. Nice stuff all of this. Checking out Magnetar's link, this store has a whole variety of models and brands in stock. The nicer ones are definitely in McLaren and Ferrari territory... BMS for once has a coaxial design that goes down to 300 Hz... Anyway I am quite fond of 18Sound, from technology to availability of data to my woofer experience. Sadly too they're not cheap, just realized the NSD1480 comes in at, ugh, 495 Euros a piece. And I thought their woofers were expensive . Then again looking at TAD's prices 18S looks positively cheap 
.
I can call up the 18S dealer here but I have no way to shell out 1000 Euros for drivers anytime soon - I'll have to stick with testing horn shapes with dome tweeters at first - so this makes me feel a bit shy (then again he already got his deal, he sold me the woofers). The same dealer also sells RCF, by data sheet the technology looks quite similar to what 18S uses, and a bit cheaper. No measurements though in catalogue or website.
Now for the bright side - looks like the 15" 120 degree 1.4" throat WG project discussed earlier is going to happen on the CNC machine. Next step will be to check if my Alibre software properly talks to the CNC's software or if I have to redesign it on ProEngineer
.
. Then again looking at TAD's prices 18S looks positively cheap .I can call up the 18S dealer here but I have no way to shell out 1000 Euros for drivers anytime soon - I'll have to stick with testing horn shapes with dome tweeters at first - so this makes me feel a bit shy (then again he already got his deal, he sold me the woofers). The same dealer also sells RCF, by data sheet the technology looks quite similar to what 18S uses, and a bit cheaper. No measurements though in catalogue or website.
Now for the bright side - looks like the 15" 120 degree 1.4" throat WG project discussed earlier is going to happen on the CNC machine. Next step will be to check if my Alibre software properly talks to the CNC's software or if I have to redesign it on ProEngineer
.
Paul W said:counts.
Magnetar,
Those big domes are beautiful! However, they sorta remind me of a McLaren I have on my wall but will never buy
I used one pair from 80 to 300 hz and a second pair in a smaller horn 300 to 2K - above that the 1" Emilar EA 175 's - It's one of a couple of systems I've built that would literally scare the s**t out of me - sort of like having a monster in the basement.
Member
Joined 2003
MBK,
You're right...there are several models in Magnetar's link that are not on the 18S website. Maybe they really are "build to order" with mix-n-match components. Since your guy knows you are a serious customer, he'll probably answer your questions if he can. The CDs may not be as expensive in Singapore as in Europe...VAT and all. The few available off the shelf in the US are priced lower than in Europe. Did you compare prices for your woofers?
Congratulations on moving the WG project forward! I bet once you've seen them, you'll pop for the drivers
Paul
You're right...there are several models in Magnetar's link that are not on the 18S website. Maybe they really are "build to order" with mix-n-match components. Since your guy knows you are a serious customer, he'll probably answer your questions if he can. The CDs may not be as expensive in Singapore as in Europe...VAT and all. The few available off the shelf in the US are priced lower than in Europe. Did you compare prices for your woofers?
Congratulations on moving the WG project forward! I bet once you've seen them, you'll pop for the drivers
Paul
MBK said:
Ha! LOL! Yeah, there is only so much you can do when you start out wrong. I guess we've all been there.
As we used to say in show biz: "You can take a turd, polish it up, dress it in a tuxedo and shine a spotlight on it - but it's still a turd."
But coming back down out the stratosphere of fancy drivers, my speaker buddy John just had a big hit with his OBs at the Dallas LSAF show. 12" mid.
The mid he used is a discontinued Eminence model that had been intended for an auto sound sub. Great big motor, the same as used on 18' drivers. According to John, a very lively, great sounding mid.
Not fair to bring up a discontinued driver, I know. But it could always be brought back. And it might be a nice choice for a less expensive rig.
BMS is available in the US at pretty reasonable prices at: Assistance Audio in Utah.
Now the question I've got is: the Radian 850PB/950PB and BMS 4592ND are about the same price, but completely different technologies. Hmm - wonder how they would sound different?
Now the question I've got is: the Radian 850PB/950PB and BMS 4592ND are about the same price, but completely different technologies. Hmm - wonder how they would sound different?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.