Disclaimer-- The following could contain errors and no expertise is claimed.
Various coax speakers have long been available. There may be a wider selection of coax speakers available today than ever in the past.
The potential advantages of Coax speakers include improved phase coherency and better-controlled directivity. Conventional 2 or 3 way speaker cabinets might have uneven frequency response because of comb-filtering between widely spaced drivers. Sound reaching the ear having variable time-delay according to frequency. Widely spaced individual drivers can also contribute to frequency-dependent irregular speaker directivity.
The distance between woofer and mid/tweeter becomes increasingly significant at close listening distance-- The listening angle between woofer and tweeter becomes more acute, potentially causing more-significant time-delay discrepancies. Therefore, conventional speakers tend to have minimum useful listening distances. Bigger speaker cabinets tend to need bigger listening distances.
Because coax speakers emit all frequencies from a shared horizontal and vertical center, they behave as point sources and possibly work better at short listening distances.
Typical coax speakers do not precisely align the woofer and tweeter in the speaker's front-to-back axial direction. The tweeter may be mounted a small distance in-front of the woofer cone. Alternately the tweeter may be mounted a small distance behind the woofer cone-- Slight axial time-delays.
Crossovers sometimes have compensating electronic delay. Even if we do not compensate for delay, it is possible that an inch or two of axial offset would not significantly affect the sound. For instance, given a coax speaker crossed over at 1 kHz-- A 1 kHz audio wave in air has about a 1.125 foot wavelength. If a coax tweeter happens to be spaced 1 inch nearer to the listener than the woofer cone-- Or spaced 1 inch farther away from the listener than the woofer cone-- This one inch delay would cause about 34 degrees of woofer-vs-tweeter phase difference at a frequency of 1 kHz. Not necessarily an extreme problem. In my opinion serious problems would need larger phase difference.
Continuing the above example, the phase difference would become less significant below 1 kHz because the comparatively quiet tweeter output would not strongly affect the comparatively louder woofer. The phase difference would become less significant above 1 kHz because the comparatively quiet woofer output would not strongly affect the comparatively louder tweeter.
It is not an unusual situation that a woofer's phase shift at the crossover frequency differs from the tweeter's phase shift at the crossover frequency. Slight axial differences could either improve or worsen natural phase differences between the two drivers. In some cases an axial time-delay might somewhat compensate natural woofer/tweeter phase discrepancies. In other cases the axial time-dalay might exacerbate natural phase discrepancies.
Tweeter In Front
Some coax speakers use a conventional tweeter in-front of a conventional woofer, the tweeter often suspended on a bracket spanning the woofer front. Rather than a front-bracket, a few tweeter-in-front designs have suspended the tweeter on a rod extending from the front-center of the woofer magnet assembly. A few tweeter-in-front speakers apparently mount the tweeter flush on-top of the woofer magnet pole piece inside the woofer voice coil.
Apparently most inexpensive coax speakers are tweeter-in-front designs-- Especially common among coax car audio speakers. More-expensive commercial-grade speakers are sometimes tweeter-in-front-- Electovoice, Altec, JBL and others have made tweeter-in-front commercial-grade speakers. I know nothing about tweeter-in-front commercial-grade speakers. For curiosity's sake would enjoy auditioning them. I do not know whether any would be “good enough” for studio monitoring or quality HIFI work. The expensive tweeter-in-front speakers which I have seen in catalogs were targeted at commercial sound and permanent install systems. However, no law requires that industrial-grade speakers must always sound mediocre.
Diffraction might happen as the woofer sound passes the tweeter and its bracket. I do not know though this seems possible.
Tweeter In Back
Most heavy-duty audiophile or pro sound coax speakers have been tweeter-in-back designs. I don't know but some of the shared-magnet models might be built with nearly zero axial offset between tweeter cone and woofer cone. Most tweeter-in-back coax speakers have a hole in the middle of the woofer magnet, a rear compression driver driving thru the hole, with a front horn sprouting from the woofer's center.
Possibly the tweeter delay in a tweeter-in-back design could lead to smoother results. The low pass woofer crossover filter will show some group delay in vicinity of the crossover frequency. This natural crossover group delay could POSSIBLY compensate for some of the tweeter-in-back distance delay. On the other hand, the same woofer crossover group delay could perhaps exacerbate problems with tweeter-in-front, making the combined delay larger, not smaller. Or perhaps the difference would be too small to matter.
Some tweeter-in-back models use a stub horn small enough to hide underneath an acoustically transparent woofer dust cap. Given the “invisible” stub horn, as viewed from the front they look like ordinary woofers. Small horns typically require high supertweeter crossover frequencies. A small horn will insufficiently load a compression driver at low frequencies. Insufficiently loaded compression drivers cannot properly couple to the air and produce sufficient SPL. Poorly loaded drivers show over-excursion-- Slamming the diaphragm against the phase plug-- Making bad sound and eventually destroying the diaphragm.
I do not know whether all stub horn coax speakers are “excellently designed” though many models form the little horn's exit profile to acoustically match the woofer cone voice coil opening-- The woofer cone becomes an extension of the horn profile. The stub horn + woofer behaves like a bigger horn tolerating lower crossover frequencies.
Have read about possible drawbacks but do not know the prevalence or severity in real-world speakers. Because the woofer cone also functions as the outer part of the high frequency horn, then low bass woofer motion might distort highs via doppler modulation of horn frequencies.
Though I do not know if this matters in the real world, it has also been claimed that some stub-horn speakers have too much offset between the horn edge and the woofer's voice coil edge-- Causing excessive diffaction as high frequencies traverse a too-rough gap between the stub horn edge and the woofer cone. Diffraction might happen as audio traverses a sharp surface discontinuity, such as a square speaker cabinet edge, or uneven spots in a horn throat, or a non-smooth horn exit profile. Diffraction can cause patterns of fine-grained frequency response irregularities.
Rather than a small stub horn, the classic Altec 604 coax speakers and many others use a separate larger horn extending from the center of the woofer. I do not know whether a discrete larger horn is innately better or worse than a stub horn approach.
For instance a 12” or 15” stub horn coax speaker which uses the woofer cone to acoustically extend the horn size-- A well-made stub horn design MAY allow as-low or lower crossover frequencies than a discrete larger horn. Even though a discrete large horn might sometimes be big enough to obscure a large fraction of the woofer cone-- If we can still see any part of the woofer behind the large horn, then the stub horn plus woofer cone would be a bigger horn-- ASSUMING that a combination of stub horn plus the woofer cone is capable of doing just as good a job as a smaller discrete horn. I do not know whether any of the stub horn speakers are slick enough to really act like an effective HF horn as big as the woofer cone. Which would be interesting to know.
A discrete horn might be less likely to show doppler distortion which stub horns might have. The discrete horn wouldn't have a stub horn's possible diffraction at the transition boundary between stub horn and woofer cone. On the other hand, possibly there could be more woofer sound diffracting around the outside of a larger discrete horn. Also a discrete horn would need a good-shaped final exit boundary to minimize high frequency diffraction as sound exits the bounds of the discrete horn.
Some “pro quality” discrete horn speakers used foam or other damping on the horn to control diffraction. For instance, the Urie time-align coax studio monitors.
Coax Speaker Candidates
I am frugal. Was reluctant to gamble vast sums on this experiment. It was possible that I might have built speakers using the most-expensive deluxe components available, only to discover that I do not like the sound. It also seemed possible that a less-expensive coax speaker might turn out quite acceptable for my needs. Uncertainties only solvable via experiment.
Expensive coax speakers are designed to sound good driven insanely loud in the pro-audio context-- On-stage floor monitors, on-stage keyboard speakers, PA cabinets, big-studio farfield monitor speakers or theater applications. My nearfield home studio application will never play insanely loud. It seemed possible that a less-expensive, less-powerful coax speaker MIGHT sound fine for low-SPL nearfield monitoring. On the other hand, perhaps a less-expensive, less-powerful coax speaker would sound bad even at modest SPL.
So I decided to first try an affordable coax speaker. If the cheap coax speaker turns out unfit for the task, then a more-expensive coax speaker could be mounted in the same cabinet-- Assuming that the cabinet is well-designed and well-built. Sealed cabinets tend to be more forgiving of such tom-foolery so long as one uses speakers with specs friendly to sealed cabinets. Many modern pro speakers are designed for best performance in vented or horn-loaded cabinets, so one must take care to select speakers with specs compatible to sealed cabinets.
I wanted to crossover the subwoofer at about 80 Hz. A suitable speaker needs to effortlessly operate flat down to at least 80 Hz, but does not need to work fabulous down to 40 Hz or lower. Perhaps many or most good 8 inch or 10 inch coax speakers would do great down to 80 Hz at modest SPL but I did not want to perform this experiment more than once to get it right. A proper 12 inch speaker in a sealed cabinet ought to be a no-brainer to operate flat down to at least 80 Hz and 12 inches is not necessarily too big to work pretty good up to 1 kHz or possibly higher.
On the other hand, a 15 inch speaker in a properly-sized sealed cabinet would be too big for my available speaker locations.
I would probably be delighted with an old Altec 604 but those are expensive and are also 15 inch speakers. The old Altec 604's are expensive used and Great Plains Audio still makes brand-new 604's which are also quite expensive. 604's are pretty easy to find if you don't mind paying the price.
Altec made a 12 inch 601 which would probably be delightful. Used prices may be lower than 604 but not as many 601 seem to get offered for sale. The 601 recommended 3 kHz crossover frequency is higher than I want. Great Plains Audio (founded by ex-Altec staff after Altec closed) makes a new-design 12 inch coax speaker which is probably fabulous. GPA does not post prices. You have to send an email inquiry. Having seen prices of other GPA items, the new-design GPA 12 inch coax is most likely fairly expensive. If you have to ask the price of a luxury item then you probably can't afford it.
Tannoy made many models and sizes of “stub horn” dual-concentric speakers. They are generally highly regarded. Prices tend to be rather high.
Parts Express is a good place to shop modern coax speakers-- Good selection and reasonable prices so far as I know.
US Speaker is apparently another good source, at the time of writing carrying a wider variety of brands/models than Parts Express.
At one time Parts Express carried more FaitalPRO coax speakers than at the time of this writing. They look pretty good but are somewhat pricey though less expensive than Altec or Tannoy. I was unable to find many internet user reviews. Technical data was sparse on some models of interest.
PE carries many B&C coax speakers. When I was searching, B&C seemed similar to FaitalPro in that they tend to have sparse user reviews and spotty technical data. Maybe there is copious technical data and user reviews but if so I could not find it. Most B&C coax models are fairly expensive though B&C products typically enjoy good reputation. B&C sells both stub-horn and discrete horn models.
The Eminence Beta 12CX, Beta 10CX and Beta 8CX are reasonably priced and enjoy plentiful positive user reviews. Eminence provides lots of technical data. Eminence is a good speaker company. The Beta CX models are not “complete” coax speakers. The listed price gets you a woofer with a threaded hole in the back and a stub horn on the front. You have to separately buy a threaded compression driver to screw onto the back in order to have a complete coax speaker. Nowadays there is a wide selection of not-expensive and pretty good compression drivers. The total price can be rather attractive even after buying compression drivers.
Beta CX are stub horn designs. As with a majority of coax speakers, recommended crossover frequency is fairly high, 1.5 kHz or higher. Part of that is because the typical use would be high power and high SPL pro audio. One wouldn't want to recommend a crossover frequency low enough to blow the compression drivers at rock'n'roll stage levels. At lower SPL such as 90 db nearfield monitoring, there is surely more wiggle room in setting the crossover lower without frying the drivers or over-excursion distortion. There are many speaker builder threads about this. I do not know “how low you can safely go” for instance with a Beta 12CX plus a heavy-duty compression driver designed to handle vastly more power than needed for nearfield monitoring. Would be interesting to experiment with it.
08-28-16 Second Thoughts-- I wasted more time web-researching opinion and hearsay about Radian coax speakers. Some folks are enthusiastic and others not so impressed. So I don't know what to think. If I was gonna spend hundreds of bucks on coax drivers, for some reason the Radians seem a “safer bet” to me compared to some of the other equally-expensive brands which also have sparse first-hand qualified user reviews. But if a fella was gonna spend $300 to $1000 per speaker, perhaps some other expensive speaker would be better. It is hard to estimate given such sparse available data. I would be anxious about spending a fortune on speakers only to be severely disappointed, regardless of which brand might happen to be the source of discontent.
I am ignorant but after obsessive reading the Radian 5212B is my best guess for a “best bang for the buck” higher-price coax speaker for low SPL nearfield monitoring-- At the time of writing priced at about $660 for a pair. Efficiency looks good and XMAX is specd at 6 mm which is bigger than many coax speakers, even among expensive coax speakers. I suspect that coax speakers typically have relatively small XMAX because of design trade-offs to make the woofer cone compatible with a horn sprouting from its center. As such, one shouldn't expect loud room-shaking sub-bass out of a typical coax speaker-- Especially when mated to a sealed cabinet. Sealed cabinet low bass response typically being limited by XMAX rather than speaker power handling capability. At low frequencies a sealed cabinet can only get as loud as the XMAX allows. That is one reason I never intended to use coax speakers without the assistance of a subwoofer.
Should the reader be curious, numerous web articles discuss the relative merits of sealed vs vented cabs. I am ignorant but concluded that sealed cabinets offer advantages for best fidelity at modest SPL. On the other hand I would want vented or horn-loaded cabinets for high SPL pro audio use.
All the Radian pro coax speakers recommend crossover frequencies in the ballpark of 1 kHz which is much lower than typical, and perhaps even more remarkable because Radians are stub horn designs.
Radian makes more-expensive and fancier coax speakers. The 5212B is the least expensive 12 inch Radian pro coax speaker. I merely suspect that for 90 dB nearfield monitoring, this “budget model” would possibly perform as satisfactorily as more-expensive Radians which go even louder and cleaner in stage and PA usage.
Some Radian coax speakers use beryllium driver diaphragms rather than the Radian default aluminum diaphragms. Some folks are convinced that beryllium is the best sound available from compression drivers. Supposedly superior accuracy at high frequencies and superior performance at low crossover frequencies. A pair of beryllium Radian 12” coax speakers would set me back at least $1700. Many audio enthusiasts spend more than that for speakers but I would have a hard time letting go of 1700 bucks regardless how great they might sound.
For curiosity I spent awhile researching beryllium and beryllium horn diaphragms. Beryllium is an expensive metal but there are several more-expensive metals in common use. Beryllium dust can be hazardous to workers. A LLNL paper about machining beryllium gives the impression of a difficult to machine metal even neglecting stringent health precautions. Given these factors, the insane price of beryllium driver diaphragms MAY be justified. I don't know. The paper lists two parameters which may account for beryllium's supposed superiority in driver diaphragms-- Young's Modulus and Natural Frequency Factor. So far as I can decipher, it means that beryllium is naturally more self-damping and more rigid than aluminum or titanium, while also being rather strong and light, at least in some applications. It is strong by some mechanical measures but weak in other mechanical measures. Which could allow the diaphragm to deliver purer sound with fewer resonances and such. Regardless, unless I win the lottery it appears too luxurious for me.
I made the provisional decision that if I try cheap coax speakers and they don't work good enough then I might bite the bullet and spend $660 to replace the cheap coax speakers with a pair of Radian 5212B.
Seismic Audio (it isn't polite to laugh out loud)
There are several more tweeter-in-back brands but I eventually ran across a very inexpensive model which made me insanely curious-- Seismic Audio CoAx 12. Seismic Audio sells low-priced pro audio gear. Though some web denizens dispute the pro part of that description. My web searching couldn't find any users who absolutely hate that CoAx 12, and several users were complimentary, considering it “really good for the money”. So I bought a pair for the grand sum of about $210, about $105 per speaker.
Researching, found a comment that the Seismic Audio titanium compression driver looks virtually identical to a Pyle titanium driver with similar specs which sells for about $15 or $20. On the other hand, user reviews of that Pyle driver are generally positive considering it a surprisingly good driver for such a small price.
I wanted a mellow compression driver capable of being pushed to as low possible crossover point. What I could learn about this possibly-Pyle Seismic Audio driver looked discouraging for a crossover below about 1.5 kHz, though possibly the high frequency performance might not be too bad. Maybe that titanium driver could go lower than I expected at 90 dB monitoring levels. Have not tested to find out.
I initially spark-tested the speakers on the work bench using the included SA titanium drivers, with a crossover of 1.5 kHz and amplitudes around 90 up to maybe 100 dB SPL. The basement wood shop acoustics are terrible and the speakers were firing up toward the ceiling sitting in their shipping boxes. Had to lean over the workbench to place ears in-line with the up-firing speakers. A difficult environment for “critical listening”. The SA titanium drivers did not sound bad though bad acoustics could hide numerous sins.
After a few hours of initial testing I replaced the stock SA titanium drivers with Dayton Audio polyimide drivers for further bench-testing. I have not tested the titanium drivers in finished cabinets installed in the office. Am curious to know how the SA titanium drivers might compare to the polyimide drivers as installed in the treated office but am too lazy to tear down, disassemble and swap drivers. Possibly I will never know.
Put another way-- If I were to try this SA coax speaker for its intended purpose-- An inexpensive speaker for modest SPL stage monitor, onstage keyboard amplification, or small PA-- I would first try the stock titanium compression driver with a crossover in the ballpark of 1.5 to 2.5 kHz. A higher crossover frequency would be fine (and safer) for live performance sound levels. For live use I would only replace that stock titanium driver if it happened to sound bad.
Polyimide Compression Driver
Had been reading that polyimide diaphragm drivers tend to be friendlier than aluminum or titanium for low crossover frequencies. Polyimide diaphragm drivers have also gained a reputation of being fairly smooth, not so harsh in the top register. So I snagged a couple of Dayton Audio DT250-P about $100 for the pair, raising the total cost of the pair of coax speakers up to about $310. Not too horribly expensive for a risky curiosity-driven experiment.
Had been reading about horns and horn diffraction and various horn lenses or other horn treatments. I wanted to try a few non-destructive experiments along that line. With a stub-horn design such as the inexpensive Eminence CX speaker-- I would have had to take a razor blade to the dust caps in order to access the little horns behind the dust caps. It didn't seem right to buy new speakers, even inexpensive speakers, and immediately take a razor blade to them for sake of curiosity. That was probably the main reason I decided to try those Seismic Audio speakers-- I can reversibly experiment on the horns without defacing brand-new speakers.
Another thing which caught my eye-- The horns appear identical to pictures of the horns in much-more-expensive coax speakers such as B&C and others. Having now examined the SA speakers close at hand, the horns still look identical to pictures of some more-expensive coax horns. The more-expensive coax speakers which apparently feature the same horn appear to have much-better-quality cast woofer frames. I don't know whether that horn shape for both the cheap and expensive coax speakers happen to come from the same Chinese factory? Or maybe the Seismic Audio speaker features a shameless Chinese clone of a Brazilian or Italian horn? Or maybe something else entirely? In today's global economy it is hard to guess.
Overall the SA 12 inch coax speaker seems well-made considering the price. Both of them seem to work about as expected. Connections and cone appear quite adequately made.
The speakers were excellently packaged for shipping. Inside the cardboard boxes were small open “mini-enclosures” made of half-inch MDF board, with the speaker sitting in a properly dimensioned hole in the little MDF enclosure's baffle board. Seemed very good support for shipping the speaker. After the speakers arrived I was able to use the MDF “mini-enclosures” sitting inside the cardboard boxes for initial spark-testing on the workbench. Of course it wasn't as good as mounting the speakers in a good cabinet-- But the speakers sitting inside the box, suspended in the “mini enclosure” with the box top flaps folded back, was enough of a test enclosure to make an initial rough estimate of how the speakers might sound in a cabinet. I'm not saying that the MDF shipping frame would be useful as any kind of permanent or even semi-permanent cabinet. Just that it looks like a solid safe way to ship speakers and was useful for bench spark-testing.
The woofer cones were fairly stiff right out of the box, though large XMAX was never advertised. After a couple of months light use installed in cabinets, the cones appear to have loosened up. Perhaps it is my imagination but they do seem quite a bit looser than when new. Not “about to fall apart” loose. Now it just feels closer to the compliance I would expect from a decent 12 inch mid-price speaker. I'm skeptical of the phenomenon of “speaker break-in” but in this case perhaps that is what has happened. If so, the speakers did not need massive power over many hours to break them in. They have been played for a few hours per session at about 90 dB SPL for a couple of months but were not played daily.
The Seismic Audio stamped frame probably ought to be somewhat stronger to support the weight of the hefty magnet plus the weight of a compression driver. Maybe this speaker can survive repeated rough road handling-- Tossed out of trucks on road gigs without the magnet weight bending the frame-- Or maybe not. If I were gonna take a speaker on the road then I would prefer a stronger frame. If not a cast frame I would at least want thicker stamped steel. However for permanent installation in the home office, am guessing that the stamped frame is probably strong enough. For all I know it may be plenty strong enough for rough road work. The steel just looks a bit thin to me though I am not nor ever will be an expert.
The horns might appear in pictures to be perhaps 90 degree dispersion both horizontal and vertical. However, the inner parts of the horn are smoothly distorted so that there is a wide axis and a narrow axis. A few B&C coax speakers whose pictures show an apparently identical horn, quote the dispersion as 60 X 40. So if the SA horn really is about the same if not exactly the same, then perhaps the SA horn is also 60 X 40 degrees.
The next installment will finally get around to performance reports of the installed speaker cabinets.
Forward to Coaxial Speakers Build Part 3 - Testing
Back to Coaxial Speakers Build Part 1 - Construction
Back to Muddling With Acoustics
Back to Main Page