because it’s used in QSC cinema system you can realistically buy and put in your home, which i talked about here:
i was initially skeptical of these Coaxials and BMS in general but the industry is warming up to them and so am i …
yes it would be better to just have a separate dedicated horn midrange and separate dedicated supertweeter but there simply is no demand in the market for such products … almost all the Supertweeters are Car-Audio units or Low-End units, and JBL has discontinued their entire Supertweeter line decades ago it seems, but kept Celestion ( which is like their bargain basement brand ) Supertweeters and used them in Marquis Dance Club series 4-way speakers … then JBL discontinued the Marquis line too.
As of right now i am not aware of a single proper supertweeter on the market. Beyma makes ones that are decent and i used them in my car but they are nothing more than cheap knock offs of decades old discontinued JBL units, which makes them less than glamorous …
and the only dedicated midrange compression driver in production that i know is BMS but it’s hideously expensive, unnecessarily powerful ( like 140 decibels ), and is marketed for MILITARY applications rather than music.
i was not able to think of any good combination of horn and Supertweeter to complete a system with 4599HE … it’s just an awkward driver.
however, there IS a HUGE demand for coaxial drivers like BMS 4594 that we’re discussing here !
i suppose what makes them popular is they can be used as a drop in replacement for regular compression drivers in existing designs, which is why a product like QSC SC-424 can exist in the first place … nobody would bother designing a system like this if it had no flexibility to use different drivers - the company must be able to adapt the design as driver availability changes ( unless they manufacture their own drivers which only JBL and RCF do ).
so let’s consider how a coaxial HF actually works …
so basically 7 khz crossover between Mid-HF diaphragm and Very High Frequency diaphragm.
also what you can’t see in these pictures is BMS uses a plastic radial phase plug … here are some images of BMS diaphragm and phase plug assembly from their various model drivers:
so keep in mind the cutaway view only shows the metal parts of the driver while there are also radial plastic fins that complete the phase plug. BMS is brilliant when it comes to optimum packaging and using the lowest cost materials and manufacturing methods for highest possible performance.
reality is any idiot can make a heavy, expensive system that is great. BMS are geniuses who can make it compact, lightweight and inexpensive.
the reason BMS uses this approach is it allows them to incorporate high frequency mirrors which allow BMS to rotate the wavefront without losing HF energy. regular waveguides must run fairly straight or HF energy dissipates.
so basically sound is collected by narrow slits, which then have 45 degree reflectors ( twice ) to rotate the wavefront by 90 degrees ( twice ) which turns it around completely ( 180 degrees ). in regular ( non coaxial ) drivers it allows BMS to put the throat inside the magnet saving space and bolt the horn directly to the magnet ( saving weight ). in coaxial designs it allows BMS to use two diaphragms that face each other with 7 khz crossover.
also this is just very efficient to manufacture because the entire phase plug is one plastic piece versus traditional phase plugs that are precision multi-part pieces, yet still can’t match BMS performance.
BMS are the most brilliant drivers i have ever seen, showing that Germans still got it when it comes to mechanical engineering.
yes it is true the problems he showed in his simulation are real - any idiot should be able to understand this would happen without any FEA modeling.
what he doesn’t explain is that the problem with other drivers ( including his own ) are WORSE.
when he explains how his own driver works he switches to a different method of analysis and ends up comparing his driver to old school drivers like 4" titanium dome and proves his driver is superior to the 4" titanium dome ( which is an outdated way of doing things ).
congratulations Dr. Brown ! you defeated a half century old HF design and proved that current state of the art driver ( BMS ) is not without issues ( though they are less than issues in your own driver ).
at least to his credit Dr. Brown admits the BMS design is “popular” … yeah because it’s the best !
so lets call them MHF and VHF sections ( Mid-high Frequency and Very-High Frequency ) with 7 khz crossover between them …
the way sound works is VHF section will have basically NO effect on MHF section. because large waves of MHF won’t be able to go into small cavity of VHF …
on other hand the MHF section will absolutely WRECK the response of VHF section because short waves from VHF will end up going into those long slits of MHF and bouncing around there and creating all kind of mess …
so how is this acceptable ?
simple ! because any other compression driver of similar output capacity is WORSE than that.
yes there is destructive interference in the phase plug of BMS … but in regular 4" compression titanium dome there is such interference in diaphragm breakup itself as well as in the phase plug and they combine in weird ways.
in fact that was the entire point of Dr. Brown’s PHD - he developed a mathematical approach to model this interference and minimize it. so if you really think about it - he has debunked his own BS - but only i am smart enough to understand that.
why did he do this ? because he knows he is smarter than everybody else ( it is VERY obvious when you listen to some of his jokes about people in Prosound industry ) and nobody would understand that he is just using them to show off his math skills and get a nice PHD paper under his belt even though the product resulting from it is useless ! ! !
yes ALL compression drivers ( except JBL D2 ) have destructive interference at very high frequencies in the phase plug … and large dome compression drivers additionally have destructive interference in the diaphragm and the combination of the two is why response above 9 khz in most compression drivers is FUCKED ( Dr. Brown demonstrates how this mess is created in his video ).
but the other issue is MASS BREAK POINT ( explained in JBL paper i once linked ) which means that above 3 khz output of compression drivers begins to fall off … and that is clearly seen also in Celestion Axiperiodic driver … but NOT in the BMS ! ! !
not to get too much into physics BMS driver solves about HALF of the issues of compression driver whereas Celestion Axiperiodic is merely the best executed conventional compression driver ever that shows off what you can do when you know math and have FEA ( finite element analysis ) but it was all just to get Dr. Brown bragging rights - NOT to create a good product !
if you listen to his lecture he spends a lot of time praising a mathematician from the past who developed a lot of the relevant math and it is clear Dr. Brown was just trying to one up that mathematician by developing his own mathematical approaches - he didn’t give A FUCK whether the product made sense or not - he wanted to get his name into history books.
and nobody at Celestion dared to say no to him because they didn’t want the only guy with brains to leave them …
so to recap - regular compression drivers begin to roll off around 3 khz and then turn to shit at about 9 khz …
celestion axiperiodic is largely the same except it’s optimized the living piss out of using fancy modeling and math …
BMX uses a 7 khz crossover where it explicitly accepts there will be a nightmare above 7 khz due to internal phase plug reflections and loss of output but these in the end turn out less severe than issues with regular drivers …
and why do any of these drivers work despite all these issues ?
two reasons
1 - power in music falls off with frequency, which means efficiency at 2 khz is way more important than efficiency at 10 khz …
2 - most people can’t hear above 12 khz or so - only teenagers can hear to 20 khz …
3 - music barely has any content above 10 khz …
so regular HF drivers ( like 4" titanium dome ) are ALMOST good enough already IN THE REAL WORLD despite being TOTALLY FUCKED on paper. that is they roll off above 3 khz and go nuts above 9 khz but actually neither of those are huge issues !
whereas the BMS is a significant improvement over that performance - it doesn’t roll off as much or go as crazy in the HF and extends much further in HF. in fact the BMS outperforms even Beryllium diaphragm dome HF drivers, and BMS diaphragm is regular plastic !
BMS are true masters of not just extreme performance but also at delivering it at low cost and weight which is why they are popular. so popular JBL rebadges them and even Dr. Brown himself admits their popularity.
it’s because to get HF energy you need a few things:
1 - you need to collect sound close to the voice coil where diaphragm is in phase with VC.
2 - the volume of air between diaphragm and phase plug must be very low, because this chamber acts as a capacitor short circuiting HF energy.
the ONLY way to achieve those goals is by a supertweeter design which means very narrow ring diaphragm only about 1/4 inch wide that is clamped between the motor and phase plug so tightly that there is barely any air space in between at all …
this approach is SO effective at producing HF that even when you lose 3/4 of all this energy in the phase plug reflections you still have MORE response remaining than in a regular HF driver …
yes a dedicated supertweeter would work even better ! but nobody makes high end supertweeters anymore. because it makes no financial sense to have a separate horn dedicated to frequencies that are barely present in music and barely anybody can hear !
it must be hard for Audiophiles who pay for silver cables to understand why somebody would try to save $10 on a $10,000 system but that’s how things work in the real world. when something is engineered every penny counts. then when you sell it you ask people to spend thousands on a cable. that’s how business works.
but there is actually no need for this to change. engineers save pennies because we are smart. audiophiles throw away thousands because they are dumb.
BMS is doing it right. why engineer for performance that Audiophiles CLAIM to be able to hear when we know they’re lying. BMS engineers for the real world, not for Audiophile LARPing.
it’s just financially / logistically more efficient to integrate the supertweeter into the HF driver and have it mountable into any standard horn than have two drivers that require a custom matched pair of horns …
that said with QSC SC-424 you do actually have a pair of custom matched horns - one for the BMS coaxial and one for B&C midrange - which is why it is worth the money - this is a very high level of integration and optimization with the phase plug on a 10" midrange and all.
the coaxial BMS design IS without a doubt a compromise, and that is why i originally rejected it myself.
but i changed my mind, because it’s the RIGHT compromise.
when i originally rejected it i was still in denial about getting older, imagining that just because i wore hearing protection most of my life that i must still have hearing of a teenager …
but reality is even if i did most music drops off like a cliff above 10 khz … as do most headphones …
and even speakers that measure flat on-axis usually drop off like a cliff above 10 khz off-axis …
essentially stuff above 10 khz is mostly for bragging rights - and with BMS you will get those bragging rights with response to 22 khz … the response above 7 khz won’t be pretty but it will be better than any other HF driver of that output capacity including beryllium ones !
of course you wouldn’t want to master recordings on this tweeter, but for music ENJOYMENT it is more than flat and extended enough …
the other issue with high end compression drivers is that they need a horn to deliver their performance and the horn can often cost as much or more than the driver itself …
this is where the coaxial design shines - by integrating essentially two drivers in one horn your odds of finding the horn that works increase dramatically versus finding two horns that work with each other …
and if we look at QSC SC-424
then you get THREE frequency bands all on HUGE custom horns that are directivity matched and the entire thing with the drivers is well priced …
so yes, in a perfect world i would have a separate MHF and VHF each with their own custom horn and then also a separate midrange with its own custom horn - IN A DREAM WORLD THAT IS
in the real world you can actually buy that QSC system for under a thousand bucks used.
well it all depends on how deep you want to get into DIY. if you want to design and build your own horns - more power to you !
but ask yourself this - what is better - a perfect system that you will never actually build ?
or a system with carefully optimized and acceptable compromises that you can realistically put together that will still be better than what 99.999% of people have ?
i am a perfectionist at heart but perfectionism is childish and weak.
real engineering is about understanding what compromises to make.
