Why doesn't Zu use crossovers in their speakers?

We think crossovers are tone suckers, especially when done at speaker-level as are 99% of hi-fi and home audio speakers. We believe the speaker-level crossovers rob tone from the system; that designing a full-range driver and loudspeaker system that does not need the heavy-handedness of a crossover is preferred and will sound better. Our full-range 10” drivers are designed around this objective and implemented into loudspeaker systems that do not require them—network components or bandwidth limiting elements are not used on any of our full-range 10” drivers. Cable comes in and runs full bandwidth directly to the voice coil—no inductors, resistors, capacitors, nothing. There is a high-pass network on the super-tweeter. Zu’s full-range drivers do exhibit aggressive high frequency attenuation from around 12k and up (high treble), but this is a function of the intrinsic behavior of the transducer (speaker driver) which cannot be defined as a component within a crossover network.

If a driver is running within its usable bandwidth without any network filtering or resistive influences, its low/high roll off or slope (knee, also referred to it as phase angle—dynamic behavior of the transducer as resistive elements, i.e. mechanical, acoustic and electrical functions of the motor...) should not be referred to as a crossover function. There is of course an acoustic crossing point between our super-tweeter and full-range driver, which can certainly be described as the acoustic crossover point, but the term crossover used in this way is attributive and not a definition of technology. The definition of a crossover as used in the audio lexicon, short for crossover network, must function as an isolated component and cannot require external reactive elements to complete it. A crossover network must perform the following functions: split a single input signal into at least two output signals, one output limiting bandwidth to the lower frequencies and the second output limiting bandwidth to the upper; having a “crossover" point between the two bands which will (hopefully) sum close to zero attenuation. In a three-way crossover network you would have the addition of a bandpass—one input signal, three output signals; low frequency band, mid band, high frequency band, with each crossing point between then summing to roughly zero.

Apologies. We did use an impedance modifying parallel filter on our first model of Definition, a zobel network, and its the only make and model to have used such a filter.