Question 9:
Evolution Acoustics speakers (with the exception of the subwoofers) employ a “Constant Voltage” crossover. Would you explain how such a crossover is constructed, how it differs from an ordinary passive crossover, and what its benefits are?
Response:
First of all we need to clarify that a Constant Voltage Crossover Network is not that mysterious. It is simply a passive network which exhibits constant voltage transfer, and is the only type of engineered design that will result in true time coincidence and phase coherence, otherwise known as first-order. Typically, most manufacturers will try to obtain this result through a common parallel network design, having one part in series with each driver. However, we employ what is known as a series design which is inherently advantageous.
Crossover components and transducers all have various tolerances in their values. These tolerance differences will always affect the total network voltage response in a typical parallel network, making it almost impossible to guarantee a true constant voltage design from speaker to speaker. In a series network, the drivers are connected in series across the amplifier output, and there are no crossover component tolerances in the direct signal path to contend with. In this type of design the sum of the voice coil voltages will always be equal to the driving voltage, thus a true constant voltage design.
The main benefit to this type of design is that there is no energy or driving voltage loss. This results in a much more dynamic presentation, because there is nothing between the amplifier and the drivers. Basically, transients are fully in tact and not suppressed by capacitors, inductors or resistors. So, when we rate a speaker at 93 dB sensitivity that is an extremely conservative estimate. You basically will experience greater dynamic contrast than with traditional parallel network speakers rated at higher sensitivities.
Another benefit to this type of design is extreme purity of signal. Because there are no capacitors or inductors directly in the signal path, there is nothing to color the integrity of the signal. Not even the most expensive capacitor in the world will sound better than a direct wire from the amplifier to the tweeter, and there is no arguing that. It is true that all of the shunt or parallel components can have some influence on the overall resulting sound of the speaker, so that is why we use all top shelf components in our crossover network, from hand made reference grade film and foil capacitors to heavy gauge pure copper flat ribbon air core inductors.
The final advantage of this type of crossover topology is the way it handles driver behavior above and below the crossover point. Without getting too technical, this type of design maintains a wide overlap of typical first order slopes around the crossover point and then sharply falls off after a few octaves in each direction. The result is true phase coherent performance with added protection on the bottom end of the tweeter and filtering of upper frequency break up in the midrange, which aids in providing better off axis response than typical first order parallel network engineering.
We must also point out that all of our crossover engineering is performed through the use of very sophisticated computer-aided design applications. However, we do not rely entirely on computer generated optimizations for final circuitry, as some manufacturers do. We perform real time measurements as well to verify all computer-based suggestions. We do use our ears to also verify results, but unlike some manufactures that design by ear, we always insist that the design must be as perfect as possible from a measurement standpoint, and will always double check through measurements that any changes made by ear are truly valid. We do this to ensure that we are not imparting our own coloration preferences on the design to ensure as natural a presentation as possible.