The generalization of expensive speakers using smaller midrange drivers is not at all true, Andy2. Look at Coincident, Dynaudio, JMlabs, Merlin, Wilson, and a whole raft of others using 6" - 8" midrange drivers. Speaker designs are as disparate as the people who purchase them. There are small midranged speakers, large midranged speakers, planars, electrostatics, horns, single driver, etc.
I also think that a 6.5" midrange is the "nominal" size, with a 5.25" coming in second place. 7" - 8" is large, and personally (probably not most audiophiles, though) think a 5.25 is on the small side. You are definitely correct about 3" - 4" being small, and I will add Avalon, Thiel, and Vandersteen to your list who build speakers with this sized midrange.
Like everything in life, there are tradeoffs to small, medium, or large midrange drivers. Larger ones can produce lower frequencies, and can go all the way down to the midbass. They can also play louder in many instances. Smaller ones can play at higher frequencies - into the treble. A 6.5" is a good compromise, as it can go pretty deep into both territories.
Notice that most of the speakers (Totem excepted) that use a smaller midrange driver are at least 3 way speakers. A two way will need a more sizable midrange driver to ensure it goes down far enough into the bass to recreate a satisfying musical experience. A lot of it comes down to the speaker builder's priorities, and his ideas on keeping the crucial crossover region out of a certain area, be it the low or high midrange.
Which is better? The real answer is neither. I hope I have made the point that great speakers can come in a variety of flavors. The ultimate answer, like most things audio, comes down to personal tastes. Close your eyes, if the speaker sounds good, it is good. Regardless of what design choices were made. |
Andy, it'd be easy to write a book about this, and I'm 2-finger typer on a damned French keyboard on vacation in Grenoble, so I gotta keep it short...plus I'm not the expert here.
Driver diameter is chosen to balance many needs. Generally, the larger the driver (and motor), the lower the passband, the narrower the dispersion at higher frequencies.
Simply, it gets hard to mate an 8" driver with a 1" tweeter, as there passbands barely intersevt smoothly, and off-axis response usually suffers from excessive "flare".
It follows that the necessary low crossover for such a design risks overpowering the more fragile tweeter's motor.
Rare is the 8+1 that sounds great (I do remember a Genelec pro monitor that was pretty impressive, but again probably intended as a console on-axis monitor rather than normal room use.), but it's a cheap way to get a lot of bass, generally as a trade-off with smooth midrange.
Smaller mids have smoother upper freq response, so usually will result in better blends with tweeters, sometimes allowing the crossover freq to be pushed WAY up, thereby reducing all sorts of crossover response crap in the critical midband. (B&W crosses the Nautili at 4k, but not to my personal liking; Verity Audio uses such a fine mid that they don't cross it until way up at 5500Hz! The "coherence" is phenomenal, as a result...but I digress.)
Since small mids don't work too efficiently down low a designer has to decide whether to boost the bottom with a vented or very large enclosure, or of course simply design the speaker with limited bass. The former may work very well within its dynamic range, the latter may benefit from boundary support or of course a subwoofer. The obvious rejoinder is to add a third driver for bass response, taking the load off the mid, and thus allowing a smaller mid that blends better with the tweeter.
In practice one sees a lot of 5" and 6.5"ish +1" monitors. GENERALLY the larger ones are punchier, having greater dynamic capability, and a more open sound. Many 5" mids sound pinched to me, with a "cupped" coloration. In my own work with an 8+5+0.75 three way I couldn't quite get the lower mids to sound as natural as with my 6.5+1, for example. Of course many designers DO have tremendously successful designs with 4-5" mids, as the physics postulates that size range as most appropriate for the female voice, for example.
Designing ANY three-way design gets much trickier in practice, despite the easy passband summations of 8-10" + 4-6" + 1"...but this is a big subject.
You noticed an inverse correlation between midrange size and price. This is a personal idiosyncracy, and perhaps the result of comparing a well-honed smaller design using a great midrange driver vs a larger design using a cheaper "mid-bass" driver. I dunno. The popularity of 6-7" + 1" two-ways indicates the success of balancing fine, well-balanced response and cost. Since it's a bit easier to improve upon the critical midband with a smaller driver, a designer will (almost) necessarily give up bass response to achieve this holy grail...sometimes the results command a higher price, even though the driver is smaller. There are too many other factors involved, some of which I hope I shed a bit of light on. A bientot. Ern |
No driver over about 4" is capable of upper midrange frequencies without cone breakup. (Breakup means that different parts of the cone are vibrating differently instead of moving as a rigid piston). Cone breakup is not a complete disaster, but tends to produce irregular frequency response, often compensated (sort of) by components in the crossover network. |
Good discourse in this thread!
My Coincidents are a GREAT speaker, using an 8" Seas P21 Excel, with a 1.25" ScanSpeak Revelator. The midwoofer's phase plug helps extend the driver's upper frequency performance and address the point Ernie brings up regarding the difficulty in implementing this. The 10" Peerless driver in the integral sub makes sure the speakers go all the way down. There is no crossover between the 8" and the 10", although the 10" does have a cap and a coil to roll it off.
The Fried speakers we're building have the new, outstanding 6.5" Vifa. The Monitors are a two way, and the Studios are a three way, which include the 8" version of this driver. The previous version of the Studios used a Peerless 5.25" midrange (and it's 8" counterpart), but I can assure you this is a MUCH better sounding speaker, in a smaller, more attractive cabinet. |
Wait a minute! a "Midrange" driver, by definition lives in a three-way system (at least). The 6 and 7 inch drivers in two-way systems are not midrange drivers, although they do need to pinch-hit for the missing driver. Even the best 7-inch drivers develop ripples in their response starting around 1500 Hz, and tweeter performance is compromised if you cross it over that low.
A midrange driver covers the range from about 600 Hz to 3000 Hz. |
El, 600Hz? That's over an octave above "middle" C on the piano! Most mids that cross that high have coherence problems, especially obvious with the human voice. And asking a "big" woofer (8+) to work cleanly at least an octave past there is pretty risky too....
I'm a fan of running a mid "naked", crossing acoustically at
150-200 Hz. Does rule out very small ones, of course. 5.25-6.5 is the generalized best starting point. Cone breakup issues at the top of the passband are minimiwed by use of outstanding cone design/material and a proper motor. Great midrange drivers are usually not cheap, although it's clear that many of them can be made to work pretty well at least for a portion of the 200-4k band (what I call the mids: an octave above the top of the piano). |
Eldartford, you are correct in the classic definition of a midrange driver - your frequency range excepted. However, you must allow that a lot of great two way speakers, such as the Merlins, have fabulous midrange, despite "missing a driver" as you say. A midrange should not go down to 600 Hz, but much below that, I would say it should do 100 Hz, as the crossover between midrange and woofer is ideally around 200 - 400 Hz, and there will be a lot of output below the crossover frequency as a matter of course.
Your assertion that the best 7" drivers develop ripples starting at 1500 puts up a red flag in this thread that many a 4" driver also shows. In my experience, the 18 cm (7"+) ScanSpeak drivers are fine out to 3000 Hz. I personally would set the crossover point a bit under that, closer to the 2500 Hz area, but being a big believer in first order crossovers, there is going to still be a heck of a lot of output at 3000 Hz and more.
My favorite midrange drivers are 6.5", which we're using in the new Frieds. Again, my Coincident Digital Masters are some of the finest speakers I have come across (which is why I own them - they are a lifetime purchase), and they use an 8" midwoofer, crossed over in the 2250 Hz neighborhood. As they were also intended to be used on their own, without the matching Troubass subwoofer which they use as a stand, they go just about all the way down. After I rebuilt the Seas P21 Excels, they have some of the finest midrange I have come across. |
What the above posters have said is correct, and there is some basic physics that can be used to predict a driver's characteristics. If you're not technically inclined, I apologize in advance. This is as simple as I know how to make it while still presenting the basic math involved.
When the wavelength of sound being reproduced is longer (lower in frequency) than the circumference of the moving part of the driver (the piston), the radiation pattern of a driver in free space is approximately spherical. When mounted to a baffle, the radiation pattern is about 180 degrees. As the frequency rises (wavelength gets smaller), the radiation pattern gets narrower. When the wavelength of the sound is about half of the circumference of the driver, the driver starts to "break up", indicated by ripples in the frequency response. At one third of the circumference of the driver, which just happens to be about the diameter of the driver, the response is down about 7dB at 60 degrees off axis. This represents a practical limit on the upper frequency of the driver. There are techniques that can be used to extend this range, but for the sake of discussion assume this is a conventional driver.
As a practical example, take a Scanspeak 18W/8545, a popular 7 inch midwoofer. You can view its data at http://www.d-s-t.com/scs/index.htm. From its datasheet, the pistonic area is 145 square cm. This translates into a diameter of 5.35 inches, and a circumference of 16.8 inches, or 1.4 feet. The speed of sound is approximately 1130 ft/sec, so 1.4 feet corresponds to a wavelength of 1130/1.4=807Hz. If you look at the 8545's frequency response graph you'll see a bump at around 800Hz, and a downward slope from that point upward. At 1600Hz (half the wavelength of 800Hz), the onset of cone breakup is evident. At 2400Hz, the response at 60 degrees off axis is down about 6 dB, which is very close to the predicted value.
Depending on the slope of the crossover, this driver should probably be limited to about a 2KHz maximum crossover frequency. Higher order crossovers can push the limit, while lower order crossovers should be around 1KHz or so. Since it is a rare tweeter that can be crossed over at 1Khz, a higher order crossover will have to be used if you want to make a decent sounding 2-way. Also, while you'd like to have the crossover above 3KHz to get it out of the most audible range, the breakup modes of the driver prevent you from doing that. It would sound too distorted.
If you had a driver with a 4" piston diameter (a roughly 5.25" driver), the 3x wavelength would be about 3200Hz, which would allow the crossover frequency to be moved upward. However, the bass response would suffer, and would likely require a 3-way system instead of a 2-way.
What does all of this have to do with the sound, you ask? If you only use drivers in their linear range, it is pretty much impossible to cover the 20-20KHz range with only 2 drivers. However, you can certainly do 40-20KHz, which is good enough for most people. Full range sound requires at least 3-way operation. But the more crossovers and drivers you have, the more expensive the speaker, and the harder it is to get the drivers to blend well.
If you'd like to read more about this subject, check out "High Performance Loudspeakers" by Martin Colloms. |
I can smell the can of worms that just opened in this thread : ) My personal thoughts are that 100 Hz is too low for a driver expected to cover the entire upper midrange region. I also agree with Trelja's comments that optimal woofer crossovers are quite low. What to do? Well, if you use a "filler" driver aka a "upper bass /lower midrange" driver to fill in the gap, you've now got more dispersion, phase, impedance, etc.. characteristics to blend together. On top of that, you've just made your three way a four way. More cost and far more complex. As such, it is a step forward or backwards? As far as larger drivers covering the range above 1.5 KHz, there are a LOT of variables here. The size of the voice coil, the contour and slope of the cone used, the type, size and shape of the dust cap, the type and size of the surround, the design of the motor structure ( magnet & pole piece ), etc... All of these things will alter frequency response, dispersion characteristics, transient response, etc... While not the "perfect" driver, take a look at this Eminence Beta 12LT. The bandwidth that they get out of this driver is VERY impressive for its' size to say the least. I'm using these as midranges in PA cabinets as they easily cover that entire pass-band with admirable results. The cost is phenomenally reasonable for what you get too. Maybe not "hi-fi" quality, but it just goes to show what proper engineering can do. Sean > |
The midrange driver range that I suggested (600 to 3000) is based on what I have seen done in three way speakers. I think that Nighthawk agreed with me, and provided all the details.
If you suggest that a woofer is good to 2500Hz, what's wrong with using it to 600 Hz? |
I think the issues that Nighthawk and Sean raise are very cogent. Speaker design and building is a craft of tradeoffs. Do you push a driver a bit out of its optimal zone, or introduce an additional driver with the added complexity of dispersion characteristics and crossover? Pick your poison.
I used to be a stauch 3 way guy. Now, my tastes lean towards the two way, due to its inherent simplicity. Do believe anyone who tells you that it is night and day easier to build a good two way than a good three way ala the crossover. Of course, Bud Fried will tell you that the series crossover makes As Nighthawk pointed out, my Coincidents go down to about 40 something Hz, filled in by a subwoofer on the bottom for that last octave or so.
Eldartford, 600 Hz for a woofer has been fine for many in the history of audio. However, many do try to keep a crossover out of the midrange, and 600 Hz is a frequency that most people can hear loud and clear. My preference is to stay far away from that as a crossover point. The three way Frieds we'll be selling cross the 8" to the 6.5" at 200 Hz, and the 6.5" to the tweeter at 2700 Hz. |
While I agree with the gist of the above, I beg to differ slightly. Since I listen to large orchestral music, I need authority, dynamics -- the ability to excite lots of air.
So, for the ~100->8-10kHz part I would use a wide-range 8-12" (supravox, lowther, goodmans, etc). The trade-off here is beaming...
For the upper range, a super tweat to over 25kHz. The trade-off here is the difficulty to align the acoustic centres of tweet & wide-range (at 8kHz the lambda is very small => the margin of error is high).
For the lower register:
At least one 15" per channel for bass (two for open baffle).
At least one 18/24" per channel for 20-45Hz. The trade off here is cost and cost (cost of drivers, cost of amplification).
Overall, such a construction ideally needs a minimum of three amp channels per side. Passive filter for crossing to tweet; Leave the wide-range free on top, cut it with a soft LPin the bottom. Active or PLL for the rest.
Wishfull thinking, eh? |
I think your opinions are very good, Gregm. I also am particularly enchanted with the sound of the Lowthers, and should have mine mounted in a cabinets I am building within a month or so. The speakers that Audio Note Kondo Japan was running at HE2004 were very much as you described, and they sounded fabulous to my ears.
You may or may not know that Bud Fried was the original importer of Lowther (also true of Quad), and when I spoke to him recently about speaker projects I was working on for fun, I mentioned them. I expected him to tear them up, but he recalled an anecdote from long ago where a dear friend of his bought some very expensive AR speakers, and they both found out that they could just not do justice to the piano, which the Lowthers had always handled with aplomb.
Piano is PARTICULARLY important to Bud, as his wife Jane is a pianist. He still holds the Lowthers in extremely high regard. Believe me, if Bud doesn't rip something, it's a compliment. If he praises it, it's more than a tribute to how good a component actually is. |
Trelja...The system you describe can cross over to the "Midrange" driver at 200 Hz because it is a 6.5 inch ubit, which I would not describe as a midrange driver. I would describe the system as a 8 inch subwoofer that can be crossed over at 200 Hz because it is so small, plus a two-way system. Not a bad idea. I have advocated running the subwoofer up to a higher than usual crossover frequency if it can hack it. Takes the heavy lifting out of the main system.
One reason to keep the woofer/midrange crossover higher than what you suggest is the electrical values of crossover network components necessary for subwoofer-like frequencies. Expensive, and bulky. |
Eldartford, may I direct you to the comment which Andy2 makes in the heading of this thread? The statement is that "expensive" speakers tend to use smaller midrange drivers.
By the way, why don't you feel a 6.5" driver is not a midrange? Anyway, you could cross a 5.25" or 4" at 200 Hz without issue.
While I am one of the biggest audio cheapskates here, I have difficulty in grasping how 150 mF of capacitance and 3.0 mH of inductance (200 Hz crossover) is too expensive for audio. Particularly, expensive speakers. This when for the same money (less than $200 for topflight components - 10 gauge North Creek coils and their better caps), an interconnect would never even be considered a serious cable.
For me, it's a no brainer sticking this kind of money into a crossover. It makes more sense to me than cable, shelves, isolation devices, tweaks, etc. But, then I am a speaker guy...
Now, bulky IS where I will agree with you. I personally have had serious concerns in putting components this large in a Transmission Line, but for the sealed or ported speaker with the kind of size that is required to run this type of crossover, I believe could site them fine. |
Trelja...I read Andy2 comment to be saying that expensive systems have midrange drivers (suitably small) whereas less expensive speakers are 2-way.
I think that a 6.5 inch driver will have some problems in the 1500 to 3000 Hz range. (Check Nighthawk). Good tweeters cannot be crossed over low enough to avoid this.
Agree that a 4-5 inch driver can go well below 200 Hz...the 5 inch woofers (and I use that term loosely) in my MTM Dynaudios put out surprisingly low tones, but not very loud. These drivers, with 3 inch voice coils, (almost as large as the moving cone) do make darned good midranges, and I used the Dynaudios that way in a biamped system for a few years. I now have MG1.6 Maggies doing what the Dynaudios used to do.
$200 is dead on for cost of the 200Hz crossover parts (I know because I bought three sets). That's a fair chunk of change for parts that the typical consumer will never see and doesn't appreciate. |
Just about every good 6.5" midrange out there should lend itself to being crossed over at 2500 - 2700 without issue. Classically, a second order crossover would be used, but there are enough great speakers out there doing this with a first order network that I am more or less convinced. Unless the network is quite steep, there is still going to be output aplenty way beyond 3000 Hz.
Add to this the resonance point of a lot of the better tweeters comes in around 500 - 700 Hz. Although I never do this, why couldn't you cross one of these (ScanSpeak Revelator, 550 Hz Fs, for example) at 1500 Hz? It would be more than twice the resonance frequency. Even with a first order crossover, you are not pushing it all that hard.
Oh, I know! Someone is going to give a fire and brimstone speech and pull out a bunch of charts refuting what I say. Let me save you the trouble with two interesting anecdotes. First, I read a Audio Asylum mishap where an inmate blasted a Dynaudio Esotar with 90 watts - full range, for break in. He programmed it, and messed something up in the delivery which he did not monitor. Even after a long time of this, there was absolutely no tweeter damage. Second, a friend of mine, in an absent minded moment wired his tweeters flat out, and ran them that way for a couple of years. Believe me, his speakers are orders of magnitude more complex than most anything built today, it could've happened to any of us. When he went in to do an upgrade, he noticed this in horror. Again, no damage - except to our ears. Putting things right took so much of the edge off of those speakers, they then sounded as glorious as anything.
You are a most dedicated person in the pursuit of excellent sounding speakers if you are spending $200 on the lower crossover! While they are not seen, anyone who does not appreciate the difference they make is missing out on a lot of slam and openess down below. Again, for the price of an interconnect that no one will take seriously or a set of brass cones to be placed under the speakers, why wouldn't one want to make a commitment to seriously better sound. I realize that the hair on fire types wrap the knuckles of anyone who even dares to view the forbidden crossover with their own eyes... |
