I thought what ScottG said made a lot of sense.
I think he was referring to the filler driver(s) as the woofer and the super tweeter that fill in where the full range driver is reaching the limits of its capability. Crossovers very low and very high. This is the path I took with my Lowther OB system and I am very happy with the results. I am also taking this path with a new smaller OB design.
To me this option seems significantly more forgiving when compared to a crossover between a large diameter woofer (probably beaming) and a coax compression driver in the 1 to 3 kHz frequency range.
I think he was referring to the filler driver(s) as the woofer and the super tweeter that fill in where the full range driver is reaching the limits of its capability. Crossovers very low and very high. This is the path I took with my Lowther OB system and I am very happy with the results. I am also taking this path with a new smaller OB design.
To me this option seems significantly more forgiving when compared to a crossover between a large diameter woofer (probably beaming) and a coax compression driver in the 1 to 3 kHz frequency range.
Hi All,
Go and take a quick look at the frequency plots at the tail end of the white paper on EnABL Specifically the initial curve trace of untreated Vs treated cone tweeter.
http://www.positive-feedback.com/Issue21/standingwaves.htm
That sort of "smoothing" up in the harsh top end of cone drivers is very typical of treated cones. So you may be able to have less than perfect cone activity here and not suffer from it. I suspect that a Mamboni process also applied to the cone driver bottom back cone skirt will help smooth things out, as these frequencies, at the top of a drivers range are where the worst sounding of the offending standing waves are created.
Bud
Go and take a quick look at the frequency plots at the tail end of the white paper on EnABL Specifically the initial curve trace of untreated Vs treated cone tweeter.
http://www.positive-feedback.com/Issue21/standingwaves.htm
That sort of "smoothing" up in the harsh top end of cone drivers is very typical of treated cones. So you may be able to have less than perfect cone activity here and not suffer from it. I suspect that a Mamboni process also applied to the cone driver bottom back cone skirt will help smooth things out, as these frequencies, at the top of a drivers range are where the worst sounding of the offending standing waves are created.
Bud
Lynn Olson said:
..there isn't a single horn there I'd recommend. Horizontal dispersion just isn't enough, and the cut-off freq.s are too high even with a steep crossover. (i.e. you would "hear" the horn.)
DDS had some horns that might "cut it" but they were all about 20 inches in width and not inexpensive. Who knows, maybe assistance audio has some in stock? Purchase one and use it to form a silicone mold for a cement horn?
It would also be MUCH better if there were 2 drivers/horns per loudspeaker operated as a dipole. (..don't remember if that was discussed or not.) Again, another reason to make a silicone mold.
Beyond that would be a custom horn, either radial or bipolar to "graft" radial dispersion.
As to the compression drivers.. Radian (after a good deal of "break in"). (..unless you like ultra clean and "pin point" in which case I'd say BMS.)
MJK said:
I was also referencing that.. but most of what was written was specifically directed at the B&O filler driver approach that Lynn "shot down".
BTW, I've done this also, Madison Knight 10's (16 ohm) in parallel for the open baffle midbass, fostex 166 esr for the mid, variety of tweeters, monopole sub, etc.. It IS a good solution.
MJK said:
Gosh, this embarassing, we're all in agreement! Hello, little trollie, where are you? We need a little fireworks. (Shhh, I didn't mean that, nobody heard it, right?)
Yes, what I'm aiming for is the opposite of the B&O approach - a near-fullrange driver that operates in parallel with bass augmentation, and is "helped out" by a tweeter with a high crossover. If the coax is abandoned, the 12" driver can be run out to its full range, perhaps with a modest series inductor (and Zobel compensation) to offset the rising midrange that is typical of large-diaphragm drivers. Despite what the Polar-Pattern Police say, I can tell you the Bastanis sounds really good (in the MBL class) in terms of dimensionality and imaging with its 12" driver running freely out beyond 5 kHz. As mentioned earlier, I think part of the real secret of dipoles isn't so much the polar pattern, but the complete lack of box coloration.
I am partial concurrence with the dipole tweeter discussion. Rather than have the backwards-facing tweeter assembly sticking out of the front baffle above the woofer, I think a better location for the rear tweeter is below the fullrange driver, maybe even at the base of baffle assembly, pointing up and facing the rear wall and ceiling. The purpose of the rear tweeter is to "light up" the back wall and ceiling reflections, bringing the total room-energy spectrum more in line with the direct-arrival spectrum. It doesn't need to be in phase synchronization with the front tweeter to do this - in fact, there's no way to do this unless you're using a prosound ribbon and remove the back plate (a very possible option).
ScottG said:
Make your own. Although I'm sure Dr Geddes would disagree, there is precious little difference between one of his constant directivity waveguides and a conical horn with a "quadratic" throat detail. I calculated this out and the max deviation between the two profiles is only a few mm.
I was going to build something like this before deciding to buy my Azurahorns. I was also going to do a rolled back edge treatment to reduce diffraction. The edge was going to use a 6 inch PVC pipe split in half lengthwise and attached to the plywood horn sides to form the roll. See the rendering.
The BIG problem with doing a constant directivity horn of any profile is finding a compression driver that doesn't screw up the directivity in the HF. The throat detail on the compression driver itself is usually a small exponential horn which causes beaming in the extreme HF no matter what the horn profile does later. You really need one with a conical throat expansion. As of a few years ago, I don't think anyone made anything suitable.
The next biggest problem is finding a good sounding compression driver. Most don't sound very good, trading away smooth response for impressive -3db HF figures. You're going to need a HP shelving circuit when using a constant directivity horn no matter how great the HF response reads in the specs. Better to get an earlier but controlled rolloff.
Jeff,
that is a point (exit of the CD driver) that has been underlined over and over again at the AA board. Maybe the B&C compression driver Geddes is using has an appropriate profile.
How about designing a ribbon with the magnets/steel front structure (and maybe back, if dipole) and flange(s) profiled as a waveguide to control and maybe match the directivity (horizontally) of the 10" or 12" wide-range mid? Linesource, a member of this forum knows ribbon design very well and might be willing to help. he has already posted here lots of interesting material. When I played with the idea of building a ribbon I had the feeling it was nor over-complicated and not expensive. And maybe, instead of a impedance matching transformer for the ribbon one could use a nice small transimpedance amplifier...
I would vote for the DIY ribbon solution, specially as the project now stands as not going for off the shelf components (i.e. designing and constructing a horn for the compression driver, which would be about as complicated as designing and building the ribbon).
that is a point (exit of the CD driver) that has been underlined over and over again at the AA board. Maybe the B&C compression driver Geddes is using has an appropriate profile.
How about designing a ribbon with the magnets/steel front structure (and maybe back, if dipole) and flange(s) profiled as a waveguide to control and maybe match the directivity (horizontally) of the 10" or 12" wide-range mid? Linesource, a member of this forum knows ribbon design very well and might be willing to help. he has already posted here lots of interesting material. When I played with the idea of building a ribbon I had the feeling it was nor over-complicated and not expensive. And maybe, instead of a impedance matching transformer for the ribbon one could use a nice small transimpedance amplifier...
I would vote for the DIY ribbon solution, specially as the project now stands as not going for off the shelf components (i.e. designing and constructing a horn for the compression driver, which would be about as complicated as designing and building the ribbon).
There are two **very special** 18sound drivers that should be considered now the coax alternative has been left behind.
10NDA610
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=241
and its smaller cousin
6ND410
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=244
their 103 and 102 dB sensitivity, low compression (look at the specs) are very attractive.
Given the large price difference (almost 2.5x) between the two drivers (i think the 6" is about 90EUR, certainly less in the US) I would be tempted to use the smaller one, specially because the crossover there would be as simple as it gets '
'
People seeking extreme dynamics could use a pair per channel.
B&C has also nice 6 and 8" mids.
http://www.bcspeakers.com/index.php...escrizione=2&prodotto=3&id_descrizione_prod=3
http://www.partsexpress.com/pdf/294-650.pdf
http://www.bcspeakers.com/index.php...escrizione=2&prodotto=1&id_descrizione_prod=3
What I wonder is, if the mid is to be crossed to the larger woofers at around 300Hz, if there is really the need to go with a 12" mid that would be more expensive, perhaps more difficult to find, have a too restricted dispersion at the high upper crossover points that have been considered, and suffer more from the peaks/dips mentioned to hide out in the smoothed response charts.
10NDA610
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=241
and its smaller cousin
6ND410
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=244
their 103 and 102 dB sensitivity, low compression (look at the specs) are very attractive.
Given the large price difference (almost 2.5x) between the two drivers (i think the 6" is about 90EUR, certainly less in the US) I would be tempted to use the smaller one, specially because the crossover there would be as simple as it gets '
'People seeking extreme dynamics could use a pair per channel.
B&C has also nice 6 and 8" mids.
http://www.bcspeakers.com/index.php...escrizione=2&prodotto=3&id_descrizione_prod=3
http://www.partsexpress.com/pdf/294-650.pdf
http://www.bcspeakers.com/index.php...escrizione=2&prodotto=1&id_descrizione_prod=3
What I wonder is, if the mid is to be crossed to the larger woofers at around 300Hz, if there is really the need to go with a 12" mid that would be more expensive, perhaps more difficult to find, have a too restricted dispersion at the high upper crossover points that have been considered, and suffer more from the peaks/dips mentioned to hide out in the smoothed response charts.
May I suggest an alternative fullrange driver
Hi guys
This driver is totally left-field. The main problem with it is that it has a very low efficiency of around 83dB/W, so large (preferably digital) amplifiers are the order of the day. I has a frequency range of 40Hz to 20kHz+. I don't have all the details here (they are at home on my home pc), but I can tell you that they were measured flat out to about 25kHz. They have a very wide off-axis dispersion, even in the high frequencies (a reliable commentator reported a 120-degree dispersion up to near as high as he could hear). They also feature a VERY flat impedance and phase response (IIRC, they are 4-ohm at 100Hz, and something like 4.4 Ohm at 20kHz). Add to this a very high power-handling ability- they were run at 100W with a full-range signal for eight hours (!!) and still show negligible compression afterwards. I can vide more detail if you guys are interested (I corresponded regularly with the designer during the design phase). The link to them is here:
DC GOLD AUDIO Lorelei
Enjoy,
Deon
PS. At one stage Alan (the designer) remarked that the efficiency of the Lorelei prototype was measuring at around 95dB/w - 97dB/w, but he decided that he wanted to make them a true full-range driver, and then redesigned the VC for the lower Fs and efficiency. I did say at that time that there would be a large market for a high-efficiency design, even with a very high Fs (I suspect it would have been in the region of 70Hz - 100Hz), but he disagreed. I think that if maybe there are enough guys that are interested in the higher efficiency design, that David (the owner of DC Gold Audio) might bow to pressure and release one.
Hi guys
This driver is totally left-field. The main problem with it is that it has a very low efficiency of around 83dB/W, so large (preferably digital) amplifiers are the order of the day. I has a frequency range of 40Hz to 20kHz+. I don't have all the details here (they are at home on my home pc), but I can tell you that they were measured flat out to about 25kHz. They have a very wide off-axis dispersion, even in the high frequencies (a reliable commentator reported a 120-degree dispersion up to near as high as he could hear). They also feature a VERY flat impedance and phase response (IIRC, they are 4-ohm at 100Hz, and something like 4.4 Ohm at 20kHz). Add to this a very high power-handling ability- they were run at 100W with a full-range signal for eight hours (!!) and still show negligible compression afterwards. I can vide more detail if you guys are interested (I corresponded regularly with the designer during the design phase). The link to them is here:
DC GOLD AUDIO Lorelei
Enjoy,
Deon
PS. At one stage Alan (the designer) remarked that the efficiency of the Lorelei prototype was measuring at around 95dB/w - 97dB/w, but he decided that he wanted to make them a true full-range driver, and then redesigned the VC for the lower Fs and efficiency. I did say at that time that there would be a large market for a high-efficiency design, even with a very high Fs (I suspect it would have been in the region of 70Hz - 100Hz), but he disagreed. I think that if maybe there are enough guys that are interested in the higher efficiency design, that David (the owner of DC Gold Audio) might bow to pressure and release one.
Jeff,
I'm glad you brought that up, because conical sure are a lot easier to build, so I'm more than happy to put plans for an OS waveguide on the back burner.
I tried 5 sided dipole waveguides thinking closer to circular was better, but odd number sides definitely don't work for waveguides. The sound changes as you move horizontally.
Using 4" drivers, I also tried quadratic throat CD dipole waveguides ala Peavy with less than acceptable results. There was much more coloration with the quadratic throat than with no throat at all and starting with a CSA much greater than Sd.
I'm glad you brought that up, because conical sure are a lot easier to build, so I'm more than happy to put plans for an OS waveguide on the back burner.
I tried 5 sided dipole waveguides thinking closer to circular was better, but odd number sides definitely don't work for waveguides. The sound changes as you move horizontally.
Using 4" drivers, I also tried quadratic throat CD dipole waveguides ala Peavy with less than acceptable results. There was much more coloration with the quadratic throat than with no throat at all and starting with a CSA much greater than Sd.
A true midrange of merit
I have implemented a really good three-way design using the PHL 1120 midrange driver. I considered a LOT of different mid-bass and midrange drivers, and the PHL 1120 was my top choice.
I have lived with this driver for over a year, and I am still extremely happy with its 'truthful' redering of the all-important midrange. 95% of the time I put music though this system I am drawn right into the music and I lose the sense of my (real) surroundings. In the remaining 5% of the time, I am just not in-the-mood to listening to music - no fault of the speakers.
I strongly recommend giving this driver a serious listen - it sounds that good!
Best Regards,
Edward[
I have implemented a really good three-way design using the PHL 1120 midrange driver. I considered a LOT of different mid-bass and midrange drivers, and the PHL 1120 was my top choice.
I have lived with this driver for over a year, and I am still extremely happy with its 'truthful' redering of the all-important midrange. 95% of the time I put music though this system I am drawn right into the music and I lose the sense of my (real) surroundings. In the remaining 5% of the time, I am just not in-the-mood to listening to music - no fault of the speakers.
I strongly recommend giving this driver a serious listen - it sounds that good!
Best Regards,
Edward[
I don't think anything has been left behind for sure. The newest idea always seems best until its flaws are discovered.
To choose drivers by their curves as supplied by the manufacturers would be pretty silly. But by discussing it, we can at least get the contender configurations and manufacturers out on the table.
If Lynn listens to the HA coax and it sounds better than an old Tannoy, well it's probably back in the running...
PHL always has to be considered for a project like this if they have models which fit the requirements. In this case that would probably be a 12" instead of an 8". Someone can browse their catalogs for contenders.
Another company that is often mentioned as high quality is McCauley. Not so much for horns but for their cone drivers. I was just looking at a pic of the PBN Montana Master Reference, and it has McCauley bass drivers.
Lynn's first posts were that he wanted to try a 12", after hearing the Bastanis. IF we go to 8" there are a whole new group to look at.
OK back to the eighteen sound:
Since we are looking at 12", then the 12NDA520
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=240
already mentioned, still seems like the most exciting candidate so far. It has the Active Impedence Control feature. Others should speak up and mention if it is BS or not , but it seems to really work:
It has about a 13 ohm impedance bump rather than over a hundred ohms for similar drivers. And for it can go out to 4k quite smoothly. One just needs to correct the rising rate.
OTOH:
the 10NDA610
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=241
has a very smooth curve and is mentioned to have "Extremely High Sound Quality" which the 12" doesn't claim. It also has "Active Impedence Control" like the 12. Finally, it appears to be gorgeous to look at, which doesn't hurt! But it probably doesn't have the bass required on an open baffle to cross at a reasonable point. I don't remember what Lynn stated, but I would assume that he wants to keep the crossover to the helper bass down around 200hz..
Maybe Lynn can help us with a discussion of the Helper Bass concept.
A point was made about the horns that they offer being too small. That has to depend somewhat on the crossover point I would think. The horn profiles seem pretty sophisticated compared to most. Since with this 12" driver the crossover can be up around 4k does that help?
Another topic:
Ribbons come up here occasionally as the backup plan, and I love my small ribbons. But ribbons that reach down really low tend to be long and I think would have problems with vertical dispersion.I have also heard that when they have to move a lot to reproduce lower notes, the distortion goes way up.The RAAL guys claim to have fixed the vertical dispersion with foam wedges applied to the top and bottom of their drivers in some models.
I
To choose drivers by their curves as supplied by the manufacturers would be pretty silly. But by discussing it, we can at least get the contender configurations and manufacturers out on the table.
If Lynn listens to the HA coax and it sounds better than an old Tannoy, well it's probably back in the running...
PHL always has to be considered for a project like this if they have models which fit the requirements. In this case that would probably be a 12" instead of an 8". Someone can browse their catalogs for contenders.
Another company that is often mentioned as high quality is McCauley. Not so much for horns but for their cone drivers. I was just looking at a pic of the PBN Montana Master Reference, and it has McCauley bass drivers.
Lynn's first posts were that he wanted to try a 12", after hearing the Bastanis. IF we go to 8" there are a whole new group to look at.
OK back to the eighteen sound:
Since we are looking at 12", then the 12NDA520
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=240
already mentioned, still seems like the most exciting candidate so far. It has the Active Impedence Control feature. Others should speak up and mention if it is BS or not , but it seems to really work:
It has about a 13 ohm impedance bump rather than over a hundred ohms for similar drivers. And for it can go out to 4k quite smoothly. One just needs to correct the rising rate.
OTOH:
the 10NDA610
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=241
has a very smooth curve and is mentioned to have "Extremely High Sound Quality" which the 12" doesn't claim. It also has "Active Impedence Control" like the 12. Finally, it appears to be gorgeous to look at, which doesn't hurt! But it probably doesn't have the bass required on an open baffle to cross at a reasonable point. I don't remember what Lynn stated, but I would assume that he wants to keep the crossover to the helper bass down around 200hz..
Maybe Lynn can help us with a discussion of the Helper Bass concept.
A point was made about the horns that they offer being too small. That has to depend somewhat on the crossover point I would think. The horn profiles seem pretty sophisticated compared to most. Since with this 12" driver the crossover can be up around 4k does that help?
Another topic:
Ribbons come up here occasionally as the backup plan, and I love my small ribbons. But ribbons that reach down really low tend to be long and I think would have problems with vertical dispersion.I have also heard that when they have to move a lot to reproduce lower notes, the distortion goes way up.The RAAL guys claim to have fixed the vertical dispersion with foam wedges applied to the top and bottom of their drivers in some models.
I
Variac said:
Has a couple pretty ugly resonaces (at just over 8 and 10k) ...
and this one looks worse ... that sawtooth looks like it could cut right thru the baffle.
An externally hosted image should be here but it was not working when we last tested it.
And Lynn was suggesting that the 18sound curves were smoothed?
dave
OK, I did say that the 12 is better than the 10, maybe too subtlely. They are only smooth in their "smooth part" Clearly they don't work in their "not smooth part"
BUT my assumption is that Lynn cutting the thing off at 4k or maybe lower like 2 or 3k right? If you flatten the curve starting at 1K and have a low pass filter at 4 k then those resonances are way down in value aren't they? I never assumed that he is going to run the 12" wide open.... After all that isn't how you use a coax which was what this idea came out of. I see it as a plus that they are way out of the bandwidth. If they still would affect the sound too much then that's a problem and I've learned something.
Just because he is using the wide bandwidth with high and low "helper drivers" doesn't mean he's using an ultra high pass 6dB filter though.
I'm sure he'll straighten us out on his intentions..
BUT my assumption is that Lynn cutting the thing off at 4k or maybe lower like 2 or 3k right? If you flatten the curve starting at 1K and have a low pass filter at 4 k then those resonances are way down in value aren't they? I never assumed that he is going to run the 12" wide open.... After all that isn't how you use a coax which was what this idea came out of. I see it as a plus that they are way out of the bandwidth. If they still would affect the sound too much then that's a problem and I've learned something.
Just because he is using the wide bandwidth with high and low "helper drivers" doesn't mean he's using an ultra high pass 6dB filter though.
I'm sure he'll straighten us out on his intentions..