Page 1 of 14 1234511 ... LastLast
Results 1 to 10 of 132

Thread: The Augmented Wideband Approach to Car-Fi: What, Why, and How

  1. #1
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    The Augmented Wideband Approach to Car-Fi: What, Why, and How

    This thread is a discussion of a different approach to a car audio front stage, the “augmented widebander” (AW) system.

    The first message will describe the augmented widebander approach, comparing and contrasting it to the conventional midwoofer-tweeter (MT) approach.

    The second message will discuss the relative advantages and disadvantages of the augmented widebander (AW) approach.

    The third message will discuss specific wideband drivers

    The fourth message will discuss specific midbass drivers.

    The fifth message will discuss a riff on the AW approach, the augmented small coax//coincident/concentric, and some coaxial, coincident, and Dual Concentric drivers.

    The sixth message will very briefly touch on more involved means for achieving the same basic goals, waveguides and arrays.

    Questions, concerns, comments, airborne rotten vegetables are all welcome!

    Comments about specific drivers used in an augmented widebander approach also welcome, and may be added to the second or third messages (with proper attribution).

  2. #2
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    I. What is the “Augmented Wideband” Approach?

    By far the most common configurations for mains (aka “front” or “satellite” speakers) one sees in car (and home!) audio is an “MT”: midwoofer+tweeter. The “MT” may be one of the following three configurations:
    1) separate drivers, i.e. a midwoofer in each front door and a tweeter in each sail-panel*;
    2) coaxial drivers, with each tweeter mounted on the midwoofer via either a pole in the middle or
    3) less commonly, coincident (tweeter dome at the base of the woofer cone) or Dual Concentric (tweeter mounted behind the woofer, but firing through a “waveguide” that terminates at the base of the woofer cone. In such a setup, the midwoofer is generally “crossed over” to the tweeter at a frequency between between 1.5 and 5 kHz, depending on the size and displacement of the respective midwoofers and tweeters.

    Like an MT, an augmented wideband or augmented widebander (“AW”) design uses two or more drivers for each main channel of the front stage. However, the drivers themselves are quite different. The upper driver (“widebander”) in an AW system is larger than a conventional tweeter, generally between 2” and 4” in diameter, but like the tweeter in a well-conceived MT system placed somewhere high in the car: upper doors, dashboard, sail panels, A-pillars, etc. The widebander is designed to play lower than a tweeter, down to between 200 and 800 Hz depending on the system design. Augmenting this widebander is a midbass driver, which plays up to between 200 and 800 Hz (depending on system design). This midbass driver may be placed in the front doors, in kickpanels, or anywhere one would put the midwoofer in a conventional MT-based front stage. So, the basic differences are, lower crossover point, and beefier drivers top and bottom.

    *I do not discuss separate MT arrangements with both drivers low in the doors or in the kickpanels for a reason. Such setups combine the ills of power response discontinuities in the midrange with the low soundstage and dulled highs endemic to low-mounted coaxes. Especially ones with the tweeters firing at one’s ankles, because high frequencies cannot pass through human legs.
    Last edited by DS-21; 08-23-2010 at 10:20 PM.

  3. #3
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    II. Relative Merits of the Augmented Wideband Approach

    Conventional MT mains have several major flaws that are basically impossible to remedy within the framework of an MT system in a car. First, the midwoofer and tweeter have very different radiation patterns at the crossover frequency. The midwoofer is “beaming,” or reducing its off-axis output, while the tweeter is essentially radiating omnidirectionally. So the spectral balance through the midrange changes perceptibly when you move your head. Second, in a car with the drivers mounted far from each other, for many - but not all! - people the music sounds audibly divided in space, with the lower frequency sounds (tympanis, kickdrum, bass guitar, tubas/euphoniums, etc.) coming from down low and the higher frequency sounds (cymbals, etc.) coming from atop the dashboard. And a low-mounted coax simply brings all the sounds down low.

    The AW approach addresses both of the major flaws in the conventional MT approach. Because the midbass driver is crosses to the wideband driver at a much lower frequency, the midbass driver is still playing within its omnidirectional range, so there is no discontinuity in the midrange power response due to mismatched driver directivities. So when you move your head the sound does not change nearly as much. Likewise, because the frequency ranges containing imaging cues all come from the widebander, the soundstage is coherent up front rather than bifurcated front-dash and front-floor as with conventional MT’s. The midrange is where the music lives, and an AW system just plain does a better job there than an MT.

    An AW system design can also offer other compelling advantages over a conventional MT: higher midbass/bass performance and lower cost. Because the midwoofer no longer needs to play high into the midrange, one may employ a driver optimized for midbass or bass generally. Therefore, depending on system design one can choose to maximize midbass impact/SPL, extend the bass up front much deeper, or some combination thereof. Also, at present at least a top-tier AW system can be considerably cheaper than a conventional MT for two reasons. First, the widebanders are often quite cheap despite the high level of design and technology in them, because they are built in huge quantities and often employed in iPod boom boxes, televisions, conference call units, and so on. So a top-tier widebander is often quite a bit cheaper than a top-tier tweeter. Second, good dedicated midbasses are often cheaper than woofers designed to extend higher in frequency. Want an example? A very good AW driver complement can be had for $150-$160 shipped today (2x Aura Whisper, 2x Peerless SLS8, from madisound.com).

    To be sure, there are also four disadvantages to the AW approach compared to an MT. First, an MT typically runs out of gas first in the low end of its range, while an AW typically runs out of gas first at the bottom of the widebander’s passband. Generally, flaws in the midrange are more audible than flaws down low. So an MT will often have higher perceived volume limits before the onset of distortion, and may have higher actual volume limits as well. Second, the dedicated tweeter of an MT may have better performance in the top two octaves than a widebander does. Third, AW systems are not constant directivity; like any piston their response window narrows as frequency rises. And since their diaphragm surfaces are generally larger than those of dedicated tweeters, they start narrowing their response at a lower frequency. So arguably widebanders require more care in aiming. Fourth, AW systems often use drivers of lower efficiency than MT’s. So they can require more power to reach the same volume, and may be more susceptible to dynamic compression.

    There are also three things that people may consider flaws with an AW that do not matter terribly much in actual use. First, many of the drivers suitable for AW systems are nominally 8Ω, rather than 4Ω as is typical for car-fi marketed product. In practice, that means your amps just trade off maybe between 1.5dB and 2.5dB of maximum output for cooler running and possibly higher efficiency. Since in many systems proper cooling is a bigger issue than maximizing output, the higher impedance may in fact be a benefit. Second, AW systems typically have degraded performance in the top octave (10-20 kHz) compared to MT systems. However, music rarely has much perceptually important signal in the top octave, and even when it does no reasonable and sensible person would prefer degraded midrange and pristine top octave to pristine midrange and degraded top octave! The third “flaw” in an augmented wideband system is that it pretty much requires active crossovers, and really shines only with sophisticated “room correction” processing. While one could do a passive crossover between a widebander and midbass, often the parts requires are large, and the drivers may be very different in efficiency. But that’s a red herring, because prefabricated passive crossovers such as those that come with component sets are worthless hacks, anyway. To properly design a crossover, the designer must know how the drivers will be placed relative to each other, the baffle surface shape, and so on. So conventional MT’s will sound better with well-implemented active crossovers (slope + EQ) as well. Likewise, good modern room correction DSP’s employing spatially-averaged measurements such as JBL’s MS-8 and Audyssey’s MultEQ XT (found in some Alpine boxes) so markedly improve any car system with so little user effort that IMO they are at the very worst a huge boon for everyone regardless of skill level.
    Last edited by DS-21; 01-23-2011 at 09:08 AM.

  4. #4
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    III. Wideband Drivers

    NOTE: this section only currently covers drivers with which I have personal experience, so do not consider this list exhaustive by any means. Drivers not in current production labeled as such. I hope that it will be expanded as others add their experiences:

    Aura NS3-193-8A
    Spec sheet: http://www.aurasound.com/public/pdf/NS3-193-8A.pdf
    Comments: Wonderful driver with a long throw and very pleasing sound. Gets shockingly low with low distortion for an 3” driver. Cheap. Has Aura’s superb “neo-radial transducer” underhung motor. Only real drawback with this driver is the cheap frame, which can buzz. For best results, damp frame and use a flexible gasket under the basket rim.
    NOTE: Aura’s naming convention is NS[driver size in inches]-[voicecoil diameter in mm]-[nominal impedance][revision number]

    Aura NSW2-326-8A “Whisper”
    Spec sheet: http://www.aurasound.com/public/pdf/nsw2-326-8a.pdf
    Comments: By far the best choice of the 2” drivers for a system design requiring the widebander to go below 500Hz, because it has an extremely long throw and an underhung motor to keep to keep distortion in check. Also, the voicecoil is huge in diameter for such a driver, and wider than that of many midbasses, at 32.6mm in diameter. Siegfried Linkwitz, of the Linkwitz-Riley crossover, uses the Whisper as the upper driver in his Pluto system. Also used in multiples in expensive arrays by, inter alia, McIntosh and M-Design. Somebody recently commented to me, “They sort of remind me of a Seas Excel and the Jordan. I think I could live with that.” (Another person recently compared them to a JBL computer system, which may be apt as I believe some of those systems had Aura drivers in them.) When driven to the edge of its limits, it distorts gracefully; best way I can put it is that it sounds like it’s driven by a warm tube amp when asked to get really loud. Fits in the upper door panel of a 1999-2004 “NB” Mazda Miata without bulging the door panel. The solder pads are a pain to use, so if it’ll fit get the version with tabs instead. A highly recommended mod is to snip the bug shield and stuff the space under the diaphragm with a polyfil plug. That really cleans up the midrange performance.

    Fostex FE103
    Spec sheet: http://www.fostexinternational.com/d...f/fe103rev.pdf
    Comments:Similarly cheap frame as the Aura NS3. More efficient, but otherwise not many advantages compared to newer drivers. Overloads easily.

    Jordan Module/JX53
    Spec sheet: http://www.ejjordan.co.uk/drivers/jx53.html
    Comments:Expensive, fragile - literally the only drivers I have ever blown are Modules and JX53’s! - and with a somewhat annoying flange. Before they go up in smoke, they sound a lot like the Aura Whispers. Replaced by the JRX6, but I do not know whether that driver is more suitable for car audio. My guess is no.

    Kef Ci50
    Spec sheet: http://www.kef.com/resources/Current..._manual_en.pdf
    Comments:Very nice-sounding little driver. Easy to pop out of the in-wall/in-ceiling bezel, though the driver rim does not have screw holes or other mounting provisions. Better HF extension than the Aura Whisper, and clean mids, too. About the same depth as the Peerless 830970. Expensive. Provisionally not recommended, because mine delaminated (surround game unglued from cone) in an open roadster in Georgia summer heat/humidity. Perhaps others in other regions will have better luck with them, though.

    Peerless 830970
    Spec sheet: http://www.tymphany.com/datasheet/printview.php?id=357
    Comments: Nominally 4Ω for those who care. A bit deeper than the Whisper; doesn’t fit in a 1999-2004 “NB” Miata without bulging the doorpanel a bit. Has an neo motor with a copper Faraday ring. Works quite well above ~700Hz for those who prefer a cooler sound than the Whispers. Excellent HF for a 2” driver, perhaps because the aluminum dustcap acts as a treble radiator (similar to the Thiel driver is the CS2.4).

    TangBand w3-1364SA
    Spec sheet: http://www.parts-express.com/pdf/264-844.pdf
    Comments: 3" bamboo cone, neo magnet driver with good sensitivity (86dB/w/m) for a widebander. OSN reports that due to its limited xmax (0.5mm) it works best crossed above 400Hz, 4th order. Also, that the directivity is extremely narrow up top, so they work best on-axis if running without tweeters.

    TangBand W4-1757SB
    Spec sheet: http://www.tb-speaker.com/detail/1230_04/w4-1757sb.htm
    Comments: Interesting-looking 4" driver. Relatively shallow (under 2" deep) and flat aluminum sandwich cone. Bassfromspace reports, "it's super detailed and would be great for off-axis installs."

    Vifa 10BGS, aka MG10MD
    Spec sheet: http://www.madisound.com/catalog/PDF/mg10md09-08e.pdf
    Comments: Probably the most popular driver ever in the German DIY scene. Excellent if you can fit it. Really superior midrange performance. Will give better lower mids than the other options here, just because it’s bigger. Likewise, aiming is more of an issue with this driver than with smaller ones. Vifa has newer drivers that are technically more interesting (neo magnet, spindly frames) that may also be better performers.

    Vifa TG9
    Spec sheet: https://www.madisound.com/store/manuals/TG9FD-10-04.pdf
    Comments: May no longer be in production, but Madisound appears to still have stock. I’ve never used it but I’m including it because manufacturer and third-party measurements suggest it would be an excellent choice for an AW system.
    Last edited by DS-21; 09-02-2010 at 03:10 PM.

  5. #5
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    IV. Midbasses

    NOTE: Drivers are "8-inch" basket diameter unless otherwise noted. As others chime in with experience I’ll add other suitable, regardless of frame diameter or original "intended application" ("car," "home," "marine," "pro," etc: a good driver is a good driver, period). Do not consider this list exhaustive by any means. Many drivers not mentioned can work very well.

    Aura NS6-255-8
    Spec sheet: http://www.aurasound.com/public/pdf/ns6-255-8a.pdf
    Comments: True 6" (not 6.5" or 7") OEM driver by Aura, sold through Madisound, though they're currently reporting them as out of stock. Extremely reasonable cost. Fits in some Honda doors stock speaker locations where standard 6.5"-7" drivers are too wide. Features Aura's underhung NRT motor for high linearity and mounting flexibility, a stamped basket, paper cone, and foam surround. Notable for its relatively high efficiency (91dB/w/m) and flat frequency response to above 2kHz. May need some treatment, if you live in a humid area. RickVS reports this driver as a "solid all around performer," and writes, "if your constraints revolve around mounting space and cost, I'd say it's the go-to driver." Chad reports that, as long as you cross them above 80Hz they are very low in distortion, sound effortless, take "GOBS of power," and have no mechanical noise. However, as one would expect for a 6.5" driver, they exhibit power compression at high levels. Chad also reported, "It is an OEM driver and really not suited for car use. I scotchguarded the foam surround and paper cone. The front gasket is simply laminated paper and wants to lift quite easily, I'd imagine a car door will compound this problem. What I did was dilute white glue in water and brush it onto the gasket on the sides so it can soak in, THIS made it VERY sturdy. "

    B&C 8NDL51
    Spec sheet: http://www.bcspeakers.com/PDF/PRD/8NDL51.pdf
    Comments: If you have processing that can address its limitations, it’s the highest performing 8” driver I’ve seen. Great impact, but does not extend low without significant EQ. So it’s a poor choice for a subless system, or a system where you can’t extend the bottom end of the midbass electronically. But if you can do those things, I love this driver. Also fairly shallow, with a small-diameter motor structure, so it may be easier to mount than other drivers.

    CSS Trio8
    Spec sheet: http://www.creativesound.ca/pdf/TRIO8.pdf
    Comments: Good choice for a subless system. It can run a bit higher than the Peerless SLS8, but besides that subjectively they don’t sound appreciably different. Needs more EQ down low than the SLS8 does. Also HEAVY, due in large part to the thick XBL^2 top-plate and AlCu shorting rings. Very nice cast frame that narrows quickly.

    Dayton RS225-4
    Spec sheet: http://www.parts-express.com/pdf/295-376s.pdf
    Comments: Never personally used, but heard enough to recommend. Well-priced, solid performance up to 800Hz or so. Good efficiency, too. Could probably use some low-end EQ in most setups.

    Dayton RSS210HF
    Spec sheet: http://www.parts-express.com/pdf/295-456s.pdf
    Comments: Never used, but heard enough to recommend. Another good choice for subless use due to its low throw and solid motor/suspension. Needs power do the low sensitivity. If you want something that “looks better” than an SLS8, or need more HF extension, and don’t mind paying $100 per woofer, here’s your driver.

    JBL 116H-1 (T-80)
    Spec sheet: N/A
    Comments: Old JBL studio monitor driver, not in regular production but seems to pop up a bit on Lansing Heritage and eBay with some regularity. Same bolt circle as many of the Bose inverted-motor 8" woofers, such as the one found in the 1998-2004 Mazda Miata. An awful lot like the B&C 8NDL51 (2” coil, paper cone, well-implemented Faraday rings, beefy cast frame) but with less overall throw. Much less efficient, but needs a lot less EQ because it has a weaker (higher-Q) motor that works well in door installations. Really likes power; I would recommend at least 75W/8Ω. Also, the durability of its foam surround may be a question mark. Wouldn’t use it subless, but with a subwoofer this midbass really shines. Good up to about 1.6kHz, so use up to about 800Hz if you’re as conservative as I am. Or higher, with very steep crossover slopes.
    NOTE: there are several JBL 116's. The 116A has an AlNiCo magnet and a lower-Q, and has a shorter stroke. The 116H is basically the 116A with a JBL's Symmetrical Field Gap ferroceramic motor. The 116H-2 seems to be a more modern 116H. The 116H-3 is hard to find information about, but may be an updated version of the 116H-1.

    Peerless 830667 (“SLS8”)
    Spec sheet: http://www.tymphany.com/datasheet/printview.php?id=36
    Comments: Extraordinary price-performance ratio. Probably the best driver of those listed, if you’re not running a subwoofer and are using a widebander that can play down to 300 Hz or lower, you can spend more but won’t appreciably gain performance. Motor performs very similarly to Peerless’s venerable XLS line, with Faraday rings and such, though the suspension seems not quite as good. If you hate stamped frame drivers then avoid the SLS8, though, because even though it’s a great frame with under-spider venting and so on. They also do surprisingly well on relatively low power, such as a high-quality amp that is rated 75W/4Ω and probably does 45W/8Ω per channel.

    Peerless 850519 (“CSC-X 8”)
    Spec sheet: http://www.tymphany.com/files/products/pdf/850519.pdf
    Comments: Doesn’t look like much, with the stamped frame, poly cone, and ugly logo dustcap, but this woofer is a great all-around performer. Can be used to 1kHz or higher if need be, and has a nice high Q for good bass performance in doors. High efficiency for the bass extension. Does lack the immediacy and “snap” of the JBL 116H-1, B&C 8NDL51, and Seas Lotus L0011.

    Seas Lotus L0011 (CW21EX-001)
    Spec sheet: http://www.madisound.com/pdf/seas/L0011.html
    Comments: Discontinued car-fi variant of the old Seas Excel W21EX. There is a newer version, I believe. Expensive, but extremely crisp and transparent driver if you use it under 800Hz.
    Last edited by DS-21; 08-25-2010 at 11:26 PM.

  6. #6
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    V. Variation on a theme: the Augmented Coax/Coincident/Concentric (“AC”) Approach

    Some people don’t want to give up the highs of a separate tweeter, but still want coherence. A small coaxial, coincident, or concentric driver can provide that. The problem is that most small car-fi coaxes are awful-sounding. And many home-focused ones now are hard to use in cars. Here are some such drivers.

    Coaxial: JBL GTO220
    Spec sheet: http://www.crutchfield.com/p_109GTO2...20.html?tp=107
    Comments: The only small coax I’ve heard that wasn’t cringe-worthy was the long-discontinued JBL GTO220. It has a 2.5” midrange with a small tweeter mounted on its grille. Gets high, and needs EQ for sure. Still, it works well, and if you must have sizzling highs it’s a good way to get them.

    Coincident: KEF Ci80QR/KHT1001.2 sat
    Spec sheet: http://www.kef.com/resources/Current...manual_en.pdf; http://www.kef.com/resources/Current...0manual_en.pdf
    Comments: Unconventional but, if it’ll fit, likely near-ideal choice. The driver sounds great, and the woofer cone constrains the tweeter’s directivity. Good down to 200Hz, but use a fairly steep slope. Driver has a conventional basket. It’s a little deep, but has a compact neo magnet. I’ve not used the Ci80QR, but it should be substantially similar.
    NOTE: the KEF KIT home-theater-in-a-box system has a driver that’s much harder to use, because the driver basket is part of the front baffle.

    Dual Concentric: Tannoy Arena sats
    Spec sheet: http://83.138.162.162/products/204/Arena%20Manual.pdf
    Comments: Man, the Arenas look tempting for car-fi. For their size they flat out sound phenomenal. Take them apart and a big problem reveals itself. The “basket flange” is in fact the entire front baffle. Damn. So unless you can fit the whole front of the egg, never mind. The CMS401 “ceiling speaker” may be more useful for car-fi, but I’ve never seen one.

  7. #7
    Founding Member
    Join Date
    Aug 2010
    Posts
    274

    VI. On Waveguides and Arrays, Too Briefly

    I would be remiss if I didn’t discuss two methods that can yield potentially higher performance than an AW and otherwise maintains all of the AW’s advantages: waveguides and arrays.

    Waveguides, briefly, are simply shapes in front of a driver designed to control its directivity. Waveguides have seriously compelling advantages in every facet of audio. Frankly, any home speaker that does not control the directivity of its tweeter with some type of waveguide is a defective design. However, to be effective down below 1kHz they need to be quite large. And the typical “HLCD’s” used in cars, the wide/short under-dash horns, just don’t work well. The shape doesn’t maintain pattern control well, they honk, and they image low. So waveguides aren’t that practical for car use.

    Arrays - multiple identical drivers summed per channel, often in a line - offer many of the advantages of waveguides, but can actually fit in some cars. For example, the A-pillars can hold a line of drivers. The primary disadvantage of arrays is complexity. Also, it’s hard to do a stealth array.

    For those reasons, even though waveguides and arrays offer theoretically and (in the hands of someone who really knows what s/he’s doing) actual advantages, they are simply not practical for most people in most cars. The AW system then becomes the best practical way to get good sound in a car.

  8. #8
    Founding Member
    Join Date
    Aug 2010
    Posts
    324
    Excellent thread! I will give it a read in the morning again and pose some questions as I have been thinking of going this route.

  9. #9
    Founding Member capnxtreme's Avatar
    Join Date
    Aug 2010
    Location
    San Diego
    Posts
    104
    Thanks for the write-up, Jay. I had gotten a pretty good idea of your approach to "AW" from your posts elsewhere, but it's cool to hear more of your thoughts and ideas on the subject.

    Anyone else out there running a similar setup?

  10. #10
    Founding Member
    Join Date
    Aug 2010
    Location
    Louisiana
    Posts
    68
    I'm running a dedicated midbass driver, Mach5 Audio MLI-65, with a Tang Band W4-1757SB in the doors of my 2006 Mustang GT. I chose the Tang Band because of its off-axis response plot and the Mach5 Audio driver for its price for performance factor.

    Currently, I have the Mach5 driver in the factory ~.5 cubic foot sealed location playing from 80 Hz to 315 Hz with the Tang band handling everything from 315 Hz on up with a 24 dB/octave slope. I want to experiment with running the Mach5 driver up to 800 Hz, but I really like the way the setup sounds as it is. I've tried running the Mach5 lower, but then I reach the law of diminishing returns as the factory enclosure tends to resonate badly causing another form of grief.

    Also, GREAT post DS-21!

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •