While a coincident/coaxial approach has potential, if you want to achieve coherent operation its application is quite limited. Still, when starting with a properly designed coincident driver, loudspeaker construction is simple. Just build the box, cut one driver hole, and mount the driver.
While I have not heard or measured every coincident driver ever manufactured, all that I have heard and measured have fallen short of my standards of acceptability. I can understand this. While coaxial is simple in conception, it is difficult in execution.
One of the major problems is the lack of coherence between the two drivers and the resultant interference. That, however, is a doable design challenge. Two variables must be controlled to achieve high fidelity quality performance from a coaxial driver. Those are the spatial alignment of the drivers for coherence and the boundary effects of the low frequency driver upon the high frequency driver.
The use of neodymium magnets in dome tweeters has allowed designs small enough to fit within the larger driver's voice coil. Also, properly handled, a 6.5 or eight inch driver can be pushed high enough in frequency to allow for transient coherence with the high frequency driver set in front of the top of the LF driver's pole piece.
Now, this is not just idle speculation on my part. I am very close. The attached graph shows the response of a coincident pair. For this test, the pair consists of a 5.25 polypropylene woofer and a neodymium motor, one-inch dome tweeter. They are coherent and they sum. The placement of the tweeter causes no ill effects on the performance of the cone driver. The impact of the tweeter will be even less for a 6.5 or eight inch driver. I have not, however, controlled for the tweeter boundary effects.
The combined response is drawn in black, the tweeter only response is drawn in red. While the pair does sum, the response above 8 kHz is too rough and that roughness is all due to tweeter boundary effects. Control for the tweeter boundary effects and use a tweeter with a little bit better top end extension, and this will be a nice option.
Good designing and good building,
Mark
While I have not heard or measured every coincident driver ever manufactured, all that I have heard and measured have fallen short of my standards of acceptability. I can understand this. While coaxial is simple in conception, it is difficult in execution.
One of the major problems is the lack of coherence between the two drivers and the resultant interference. That, however, is a doable design challenge. Two variables must be controlled to achieve high fidelity quality performance from a coaxial driver. Those are the spatial alignment of the drivers for coherence and the boundary effects of the low frequency driver upon the high frequency driver.
The use of neodymium magnets in dome tweeters has allowed designs small enough to fit within the larger driver's voice coil. Also, properly handled, a 6.5 or eight inch driver can be pushed high enough in frequency to allow for transient coherence with the high frequency driver set in front of the top of the LF driver's pole piece.
Now, this is not just idle speculation on my part. I am very close. The attached graph shows the response of a coincident pair. For this test, the pair consists of a 5.25 polypropylene woofer and a neodymium motor, one-inch dome tweeter. They are coherent and they sum. The placement of the tweeter causes no ill effects on the performance of the cone driver. The impact of the tweeter will be even less for a 6.5 or eight inch driver. I have not, however, controlled for the tweeter boundary effects.
The combined response is drawn in black, the tweeter only response is drawn in red. While the pair does sum, the response above 8 kHz is too rough and that roughness is all due to tweeter boundary effects. Control for the tweeter boundary effects and use a tweeter with a little bit better top end extension, and this will be a nice option.
Good designing and good building,
Mark
Attachments
IMO, a perfectly-designed wide-range coaxial driver is much better than a single driver covering the same range or two drivers covering parts of that range.
Are you exploring waveguides to reduce the "tweeter boundary effects" a la Tannoy? Does that approach have the potential to satisfy your requirements?
As the design progresses I will release more information.
Now, I am a theoretician who applies. As a theoretician, however, my work is equally divided between developing new theories and disproving old theories.
I believe the theory that underpins both wave-guides and transmission lines is fundamentally flawed. I will be controlling the boundary effects with something that may look like a wave guide, but will not be designed by wave guide theory because when built it will not be a wave guide.
I realize that what you call it probably only matters to a theoretician (or a logician), but then I am a theoretician and it matters to me.
I am in agreement that with the proposition that in the comparison of two equally designed two-way systems, one coaxial and the other a spaced two-way, the coaxial will be the superior loudspeaker. The problem to date has been that the design challenges of a spaced two-way have been better addressed than in coaxial designs.
I have been working toward the coaxial with a series of closely spaced two-way designs. One has been published (the PE 5.25 buyout and Vifa tweeter design). Two other designs, equally practical, have not been published. All of those designs use tweeter front loading (which have been called wave guides by others).
While I am just swamped right now, I hope to free up enough time to apply some front loading designs I worked up to see if they will eliminate the tweeter boundary effects. I hope to be able to get to that in a couple of days. Before I release details I want to prototype a 6.5-inch woofer. This will allow significantly more bass than the current 5.25-inch test unit.
Although still very early, the coincident pair test units sound very nice even with the tweeter boundary effects. Very promising.
Good designing and good building,
Mark
Now, I am a theoretician who applies. As a theoretician, however, my work is equally divided between developing new theories and disproving old theories.
I believe the theory that underpins both wave-guides and transmission lines is fundamentally flawed. I will be controlling the boundary effects with something that may look like a wave guide, but will not be designed by wave guide theory because when built it will not be a wave guide.
I realize that what you call it probably only matters to a theoretician (or a logician), but then I am a theoretician and it matters to me.
I am in agreement that with the proposition that in the comparison of two equally designed two-way systems, one coaxial and the other a spaced two-way, the coaxial will be the superior loudspeaker. The problem to date has been that the design challenges of a spaced two-way have been better addressed than in coaxial designs.
I have been working toward the coaxial with a series of closely spaced two-way designs. One has been published (the PE 5.25 buyout and Vifa tweeter design). Two other designs, equally practical, have not been published. All of those designs use tweeter front loading (which have been called wave guides by others).
While I am just swamped right now, I hope to free up enough time to apply some front loading designs I worked up to see if they will eliminate the tweeter boundary effects. I hope to be able to get to that in a couple of days. Before I release details I want to prototype a 6.5-inch woofer. This will allow significantly more bass than the current 5.25-inch test unit.
Although still very early, the coincident pair test units sound very nice even with the tweeter boundary effects. Very promising.
Good designing and good building,
Mark
Interesting how you talk about the non-waveguide approach. I've been thinking a lot about a coaxial driver with the larger diaphragm shaped into an oblate spheroidal waveguide to provide the HF driver with directivity control. Make the radius large enough so that the excursion will be minimized and so the throat can be large enough for the required frequency at which the HF is high-passed. A coincident tweeter might work, although it might be better to piggyback a compression driver that fires through the center pole of the LF motor.
Note: this isn't quite original thinking, so get off my back. 🙂
Note: this isn't quite original thinking, so get off my back. 🙂
IME, the main problem with the "tweeter inside the woofer voice coil" coax design, has been that I've yet to see a tweeter, with enough DYNAMIC OUTPUT CAPACITY, especially at the lower end of its range, to keep up with a 6.5" or 8" woofer.
OTOH, the waveguide-loaded units, have SIGNIFICANLY more bottom-end safe output operating area, from the tweeter. Take, for example, the Tannoy 8" dual concentric from the Dimension series speaker... which I have recently incorporated into a custom design. The optimum crossover point was 1400 Hz, which is the SAME as the FACTORY uses on this driver. And it works incredibly well... NO sense of fatigue in the tweeter output range. No dome tweeter of the size possible to install inside of a woofer voice coil, can do this, not of any commercially-available design I've seen so far from ANY of the major parts manufacturers.
As for integration- it should be possible, using a waveguide, to fairly much REMOVE the influence of the woofer cone from the tweeter output, at any frequency where the normal cancellations would occur (ie, above 5KHz or so )... it's possible to reduce the dispersion in the top two octaves, to where the tweeter just doesn't "see" the woofer cone much at all. Also, I have been working with a tried-and tested idea to BYPASS the falling-output problem of horn- and waveguide-loaded tweeters... the use of a dedicated supertweeter to augment JUST the top half-octave of the audible spectrum, and beyond. When the supertweeter ONLY has to go down to 10KHz, then we can start talking about designs that go up to 30 and 40 KHz and beyond. One example used as a design model for me, is the Tannoy ST100 supertweeter... goes from 12KHZ to 47 KHz +- 3 dB rolloff, and is only 10 dB down at 100 KHz. Anyone got ANY IDEA how MUCH improved the phase and time response of a system is, with bandwidth out to beyond 40KHz? NO more group delay problems above 5KHz when implemented properly!
I am about to, as I mentioned in another thread, start looking at other alternative coax type drivers... to begin to experiment with modifying tweeter throats, try alternative driver designs (including the idea of waveguide-loading a standard dual-chamber tweeter behind the woofer), and using damping to control the output of both the woofer and coax tweeter... it's probably a long and hard journey, but I've already HEARD what's possible, when it's done even REASONABLY well, so it's a journey I approach with anticipation...
Regards,
Gordon.
OTOH, the waveguide-loaded units, have SIGNIFICANLY more bottom-end safe output operating area, from the tweeter. Take, for example, the Tannoy 8" dual concentric from the Dimension series speaker... which I have recently incorporated into a custom design. The optimum crossover point was 1400 Hz, which is the SAME as the FACTORY uses on this driver. And it works incredibly well... NO sense of fatigue in the tweeter output range. No dome tweeter of the size possible to install inside of a woofer voice coil, can do this, not of any commercially-available design I've seen so far from ANY of the major parts manufacturers.
As for integration- it should be possible, using a waveguide, to fairly much REMOVE the influence of the woofer cone from the tweeter output, at any frequency where the normal cancellations would occur (ie, above 5KHz or so )... it's possible to reduce the dispersion in the top two octaves, to where the tweeter just doesn't "see" the woofer cone much at all. Also, I have been working with a tried-and tested idea to BYPASS the falling-output problem of horn- and waveguide-loaded tweeters... the use of a dedicated supertweeter to augment JUST the top half-octave of the audible spectrum, and beyond. When the supertweeter ONLY has to go down to 10KHz, then we can start talking about designs that go up to 30 and 40 KHz and beyond. One example used as a design model for me, is the Tannoy ST100 supertweeter... goes from 12KHZ to 47 KHz +- 3 dB rolloff, and is only 10 dB down at 100 KHz. Anyone got ANY IDEA how MUCH improved the phase and time response of a system is, with bandwidth out to beyond 40KHz? NO more group delay problems above 5KHz when implemented properly!
I am about to, as I mentioned in another thread, start looking at other alternative coax type drivers... to begin to experiment with modifying tweeter throats, try alternative driver designs (including the idea of waveguide-loading a standard dual-chamber tweeter behind the woofer), and using damping to control the output of both the woofer and coax tweeter... it's probably a long and hard journey, but I've already HEARD what's possible, when it's done even REASONABLY well, so it's a journey I approach with anticipation...
Regards,
Gordon.
Coaxials
Hi Mark,
I'm intrested in what you find. I too believe that coaxial is the next step in improving speaker design. Jim Thiel is definately working in this direction. I would love to get my hands on one of his 6.5" coaxials to use in an open baffle design with 8th order or higher XO. The're frightfully expensive though ( I considered his wallmounts). The tweeter design seem's like it could be pushed quite low as he is a firm believer in 1st order crossovers. I also looked at PSB's new 8" coax ceiling speakers ($750 each!) which uses a waveguide for the tweeters to smooth the boundary effects. I dont know if you are familiar with these. I'm considering something along these lines to incorporate into my next hybrid of my Orions. Here's some pics. http://photos.yahoo.com/ph//my_photos
The Thiels would drop right in - inplace of the 6.5" waveguide/tweets. Or I could revise the main panel for a single 8" coaxial. I think you want to limit the excursion of the woofer/mid in a coaxial design (open baffle certainly does not help here). Have you considered something like the H-Vi D8.8 8" woofer with a 4"vc. Would it be possible to mount a 4" waveguide/tweeter within such a large voice coil? The 8" driver should have a high enough break-up where a low crossed dome would blend ok. 4" waveguide should, help the tweeters low end, but...boundary effects??? Your thoughts?
Hi Mark,
I'm intrested in what you find. I too believe that coaxial is the next step in improving speaker design. Jim Thiel is definately working in this direction. I would love to get my hands on one of his 6.5" coaxials to use in an open baffle design with 8th order or higher XO. The're frightfully expensive though ( I considered his wallmounts). The tweeter design seem's like it could be pushed quite low as he is a firm believer in 1st order crossovers. I also looked at PSB's new 8" coax ceiling speakers ($750 each!) which uses a waveguide for the tweeters to smooth the boundary effects. I dont know if you are familiar with these. I'm considering something along these lines to incorporate into my next hybrid of my Orions. Here's some pics. http://photos.yahoo.com/ph//my_photos
The Thiels would drop right in - inplace of the 6.5" waveguide/tweets. Or I could revise the main panel for a single 8" coaxial. I think you want to limit the excursion of the woofer/mid in a coaxial design (open baffle certainly does not help here). Have you considered something like the H-Vi D8.8 8" woofer with a 4"vc. Would it be possible to mount a 4" waveguide/tweeter within such a large voice coil? The 8" driver should have a high enough break-up where a low crossed dome would blend ok. 4" waveguide should, help the tweeters low end, but...boundary effects??? Your thoughts?
Ooops
Sorry, I forgot that just because I can see it on my computer, doesn't mean that you can. Here's the proper link.
http://pg.photos.yahoo.com/ph/ajinfla1966/detail?.dir=7e78&.dnm=b27f.jpg
And
http://pg.photos.yahoo.com/ph/ajinf.../pg.photos.yahoo.com/ph/ajinfla1966/my_photos
Hope these work for you.🙂
Sorry, I forgot that just because I can see it on my computer, doesn't mean that you can. Here's the proper link.
http://pg.photos.yahoo.com/ph/ajinfla1966/detail?.dir=7e78&.dnm=b27f.jpg
And
http://pg.photos.yahoo.com/ph/ajinf.../pg.photos.yahoo.com/ph/ajinfla1966/my_photos
Hope these work for you.🙂
Interesting,
Well, I am not even trying for 40 kHz. Has anyone independently verified -3 at 40 kHz or -10 at 100 kHz?
While I have tested a number of units that made claims of extended response (30 kHz), I have found them to be rolling off very steeply below 20 kHz.
With the front load appliance, the dome tweeter I am testing makes it to 20 kHz, but not much farther.
Anyway, I logged on to post an update on some design work. I have determined the critical diameter (at the dome) for the front load appliance. It is .872 inches. This is applicable only for the driver I am using. Other drivers will require different dimensions.
I have also been working with a load appliance with a 1.5 inch length. This places the high frequency release point proud of the woofer and its surround. I have mocked up just the cone and surround with the tweeter to cone relationship identical as in the actual woofer. Yet when mounted identically in the actual driver it measures differently. There is an unidentified variable in play. Must find that variable.
Lastly for today, I am not trying to go low with the tweeter. Rather, by using Whisper Cone TM drivers, I am going for 10 kHz from the woofer. While the final design may drift a few K around this point, all muscial tones will be reproduced by the woofer.
If you are going for a 2 kHz crossover point then the delay caused by the low pass filter will dictate an alignment point well behind the woofer. This is also a problem with none time aligned spaced two-ways. Makes them sound unnatural.
Good designing and good building,
Mark
P.S. I will not call these things (whatever they are) wave guides.
Well, I am not even trying for 40 kHz. Has anyone independently verified -3 at 40 kHz or -10 at 100 kHz?
While I have tested a number of units that made claims of extended response (30 kHz), I have found them to be rolling off very steeply below 20 kHz.
With the front load appliance, the dome tweeter I am testing makes it to 20 kHz, but not much farther.
Anyway, I logged on to post an update on some design work. I have determined the critical diameter (at the dome) for the front load appliance. It is .872 inches. This is applicable only for the driver I am using. Other drivers will require different dimensions.
I have also been working with a load appliance with a 1.5 inch length. This places the high frequency release point proud of the woofer and its surround. I have mocked up just the cone and surround with the tweeter to cone relationship identical as in the actual woofer. Yet when mounted identically in the actual driver it measures differently. There is an unidentified variable in play. Must find that variable.
Lastly for today, I am not trying to go low with the tweeter. Rather, by using Whisper Cone TM drivers, I am going for 10 kHz from the woofer. While the final design may drift a few K around this point, all muscial tones will be reproduced by the woofer.
If you are going for a 2 kHz crossover point then the delay caused by the low pass filter will dictate an alignment point well behind the woofer. This is also a problem with none time aligned spaced two-ways. Makes them sound unnatural.
Good designing and good building,
Mark
P.S. I will not call these things (whatever they are) wave guides.
Attachments
- Status
- Not open for further replies.