Originally Posted by
davef
There is a LOT more to woofer performance than simply power handling and bass extension. A few of the main factors I look for are (not necessarily in this order) - smooth on and off-axis frequency response, excellent transient response, and very low cone resonance at higher frequencies.
Bass extension is controlled more by the design of the speaker itself than by the woofer. The reason you are impressed with the bass extension of the Salk Silk is because of the low efficiency of this speaker (83dB). If we dropped the efficiency of the S2 down to equal that of the Salk Silk, we would effectively have the same - actually slightly deeper bass extension.
The Scan 15W does not have deeper bass extension than the woofer we use in the S2. Please familiarize yourself with Hoffman’s Iron Law of Speaker Building.
When I was referring to the complexity of the crossover, I was not referring to the order of the filter slopes. I was referring to additional crossover networks to compensate for issues with the woofer or a tweeter. For example, a parallel notch filter - which is needed to compensate for the cone breakup on the Scan 15W uses 3 components (a cap, inductor and a resistor). While it is a competent bandaid for the issue with this woofer - it creates other issues. Notch filters, impedance compensation networks, complex L-pads, Zobel networks - I prefer to avoid using these whenever possible by focusing on getting the woofer and tweeter right for our specific application. This is precisely the reason we customize nearly all and even fully design some our own transducers. We could simply use the Scan 15w and call it a day (that is actually really simple) - but it is not how we do things here. I like the motor on the Illuminator but prefer the cone of the Revelator -- I suspect the Scan Ellipticor woofers might be a good compromise.
1st order filters create more wave form interference as more of the woofer and tweeter response bleeds into each other, worsening comb filtering. This gets worse the further off-axis you are and for a woofer, with low order filters - you run into issues with beaming and for a tweeter, run into issues with power handling. The only advantage 1st order filters offer is there is no phase shift.
With the tricked out Curv woofer that I intend to use in this Diamond Sierra, the goals were:
even faster transient response
less cone breakup
Better cooling (higher output before dynamic compression)
significant reduction in inductance
Bass response will be determined by how much overall efficiency of the speaker I am willing to sacrifice, but I am not a fan of using too much padding on a tweeter...