Crossover definition

I'm not too technically knowledgeable but I'll go out on a limb here a little bit.

I don't think that Butterworth is a crossover. Rather it's a filter used in the crossover. It prevents the frequency response from getting too wavy around the crossover point. Since almost all crossovers use Butterworth filters, they are often called Butterworth crossovers.

Perhaps someone more knowledgeable might be able to add to (or contradict) this.

I don't think that Butterworth is a crossover. Rather it's a filter

A crossover IS a filter. As you know, the xover is there to divide frequencies between drivers through attenuation (dependng upon how strong -- the slope or order -- that attenuation is one speaks one 1st order, 2nd, etc). To expand on the original question would be outside the scope of a post IMO. Suffice to say that each "type" of xover, Bessel, Butterworth, L/R, etc has electrical characteristics other than the order. One place to visit is here. There is alot of info on the net. Cheers

Sorry Markphd,

That would be completely incorrect, anyone else care to try?

All topologies do ABOUT the same trick, but depending up a number of variable a designer will chose one over a another. Hence, there is no BEST way. Usually the decision is based upon the drivers measured characteristics.

The biggest difference between the types of X/O topologies is the knee portion of the cross over.

The intent is to provide a seamless match, but due to the nature of the filter unless one opts for eliptic's a or derivative there-of (aka Joseph Audio, or Infinite Slope's)
there is substantial overlap or "blending" of the two drivers. The slope function of the different types are more similiar than different so what's left is the knew or the amount of accoustic loss or gain at the exact cross over point. For some you want absolutely flat, which is usually where the LR 4th orders are employed, again only an example. Butte Worths, Bessels ect, all have slightly different knee functions and can be used to offset notches or peaks at the X/O point.

One cans use LEAP or many other software programs to get on close, but the key to using simulators is they can only model what their told, they can't select topologies or types.

Thge skill of the designer, the choices in type, the choices in components and drivers is what allows two similiar products to sound remarkably different.

Without getting into graphs and charts, and math, thats about as close to a non technical explination as I can get.

loon