Once again, i repeat that there is NO reason that a speaker using a first order crossover can't play loud. It is all up to the design. Just because the market is not flooded with such designs does not mean it isn't possible.
The easiest way to get around this would be to use a 4 or 5 way design with drivers that are wideband, relatively efficient and handle good power to start off with. Using this approach, you can still maintain time coherency ( if taking into account baffle step ) and minimize the strain put on any one driver. Since most designers using a first order crossover try to keep things simple, going to a four or five way design is kind of the opposite of what they are trying to achieve. Hence the lack of popularity of such an approach.
Keep in mind that in a first order crossover, more signal is fed into the actual drivers of the speaker due to the reduced amount of loss and "increased leakage" associated with such a shallow slope. Signal is "eaten up" by the increased amount of components necessary to build higher order crossover networks and the associated "notch filters", impedance compensation, etc.. that typically go with such designs. As such, first orders should play louder with less input and continue to do so until the components within the speaker system reach a point of "saturation" or "dynamic compression". It is only once the speaker has reached the point of saturation that the higher order crossover will be able to play louder, but it will take more power to do so. I am speaking of an apples to apples situation i.e. a speaker using identical drivers with both a simple and a "fancy" crossover network.
If you doubt this, try using speakers that have high level "fancy" crossovers with a GOB of parts in them. Speakers such as Dynaudio's, Nestorovic's, etc... need as much power as you can feed them for the very reasons mentioned above. The losses in the crossover are infinitely high compared to speakers like the Coincident's, which don't use anything but the bare minimum in crossover parts to get the job done.
I would love to hear contrasting points of view on this subject if one thinks that i am wrong. Sean
>
The easiest way to get around this would be to use a 4 or 5 way design with drivers that are wideband, relatively efficient and handle good power to start off with. Using this approach, you can still maintain time coherency ( if taking into account baffle step ) and minimize the strain put on any one driver. Since most designers using a first order crossover try to keep things simple, going to a four or five way design is kind of the opposite of what they are trying to achieve. Hence the lack of popularity of such an approach.
Keep in mind that in a first order crossover, more signal is fed into the actual drivers of the speaker due to the reduced amount of loss and "increased leakage" associated with such a shallow slope. Signal is "eaten up" by the increased amount of components necessary to build higher order crossover networks and the associated "notch filters", impedance compensation, etc.. that typically go with such designs. As such, first orders should play louder with less input and continue to do so until the components within the speaker system reach a point of "saturation" or "dynamic compression". It is only once the speaker has reached the point of saturation that the higher order crossover will be able to play louder, but it will take more power to do so. I am speaking of an apples to apples situation i.e. a speaker using identical drivers with both a simple and a "fancy" crossover network.
If you doubt this, try using speakers that have high level "fancy" crossovers with a GOB of parts in them. Speakers such as Dynaudio's, Nestorovic's, etc... need as much power as you can feed them for the very reasons mentioned above. The losses in the crossover are infinitely high compared to speakers like the Coincident's, which don't use anything but the bare minimum in crossover parts to get the job done.
I would love to hear contrasting points of view on this subject if one thinks that i am wrong. Sean
>