Hi all! I’ve read the speaker design books through 1 time and am starting to work my way through some longhand speaker math. My first project is going to be a subwoofer, and as I’m still teaching myself through the equations with the goal right now of just getting the theory down, I’m already stumbling on a question:
In sealed boxes, how does one go about getting a good strong low extension?
My first run through with a Dayton ultimax 12” driver gave me a larger than recommended box and a Fb of between 34Hz and 29, depending on playing with Qtc(did .707 and .669 just to experiment.)
I don’t necessarily mind a large box, and I recognize this might not be the right driver for the goal... What should I look for in the specs that will help me achieve around -3dB at 20Hz? Is it just as simple as finding a low Fs?
Many thanks!
Glad to see you are seriously getting into speaker design.
First there is an easy way to spot if a driver should be loaded sealed or ported. Take Fs and divide it by Qes. If the ratio is less then 50 then that driver needs to be in a sealed enclosure. If the ratio is over 100 then the driver leads some acoustic loading. Between 50 and 100 the driver may work either way.
Now the important point is that a high Q system is going to have a prominent peaked resonance and be a nasty thumper. Some pop enthusiasts like that. I don't.
As Q is lowered there is loss of sensitivity and efficiency, Fs tends to rise and roll off transitions from fourth to second order. This is easiest seen in pipes like the TLs I design. A closed pipe with no damping has high output at the fundamental resonance. It has high output and is efficient. However it is resonant and has two peaks of impedance. It rolls off below resonance fourth order. Now if you sacrifice some output and efficiency by adding damping until you get one impedance peak, then you get second order roll off and the reproduction is essentially non resonant and low Q. Now system Q what ever you do will always be a little higher than the Qts of the driver. For low Q accurate reproduction you need to start with lower Q drivers. A ported reflex enclosure can not be damped in the fashion above or the box resonance is totally killed.
Now in a sealed enclosure F3 will always be substantially higher than the F3 of the driver. Also the lower driver Qts the higher F3 will be above driver Fs.
Now we get to Hoffman's Irons man's law. This says that low frequency output decreases with enclosure size.
Now the only reason to design drivers for sealed enclosures and build them is to reduce enclosure size. The penalty is massive inefficiency.
So if you have the space you don't build bass systems using sealed boxes. The downsides are highly significant.
The requirements for a good sealed sub, is a massive driver. It needs high power handling and very high excursion limits. Rear cone radiation is lost, and the system can only couple to the room from the front of the driver, which is very inefficient. There is no acoustic coupling like a Helmholtz resonator, a pipe, or a horn. In the latter three forms of loading the cone excursion is controlled by pressure changes behind the cone which are transformed to air displacement at the openings (ports or mouth).
So in a sealed alignment lots of power will be required, powering an expensive driver. Eq will be required at 12 db per octave starting just above F3. It is not possible to design or build a sealed sub without this equalization. This has to be part of the design.
If you have the space, in my view it is folly to do a sealed design. A design that can convert pressure behind the cone to air displacement is far preferable. If you have the space, horns and pipes are the best solution as reproduction can be made non resonant, something a reflex ported design can not quite match even under optimal circumstances.
Finally a word about where system Q should fall. As Swerd has pointed out it is accepted wisdom that 0.7 is the aim. However I don't think that is correct, and I'm not alone in that. The problem is that it is not easy to design and develop truly low Q (0.5 or lower) where reproduction is truly aperiodic and therefore essentially non resonant. We have become accustomed to the warmth of slightly resonant reproduction. However if you listen critically to live acoustic instruments, you can tell that your speakers are not really accurate. Really accurate bass reproduction demands a lower Q than 0.7. At first that lack of that artificial warmth takes people aback, but once that becomes you reference you are never happy again with higher Q bass resonances.
You have asked an important question that goes to the heart of getting really accurate and satisfying bass reproduction. Moving coil loudspeaker based systems seldom achieve this and very, very few commercial designs.
