>You can get away with
>much smaller wire and a much smaller inductor if you're willing to live with
>higher DCR. But how high is "too high?"

As Paul mentioned, you could always use a circuit simulator to do this sensitivity analysis in a rigorous fashion. But I think we just need some prudent guidelines for real world design.

In crossover design low DC resistance (DCR) inductors are required for two reasons. First, low DCR results in a response closer to the modeled response (since most crossovers are designed assuming "ideal" inductors). Second, and maybe even MORE important, low DCR reduces power dissipation in the inductor. The second point may not seem significant until you start doing long-term power handling tests on your system. After 8 hrs at 800 Watts of continuous shaped pink noise I have seen inductors at temperatures above 200 degrees F and getting so hot they shorted out! Imagine what the DCR of that inductor climbed to before it died. So not only does the crossover response shift with the inductors DCR, it also shifts as the inductor heats up. Even more reason to keep the DCR ever lower.

I usually think of the DCR I can tolerate in terms of the nominal impedance of the associated drivers. I try to keep my inductors DCR less than about 3% of the driver impedance. For 8 Ohm systems this is 0.24 Ohms. But it depends on the location in the circuit.

You should also try to calculate the dissipation of the L under full power conditions as a sanity check. Or better yet, measure the temperature rise of the inductor over a period of several hours at full power. Trouble can come from surprising directions when doing that long term full power test. But if your speakers will be used in public places you need to be the first to know what happens when they are used at full rated power and also what happens when they are ABUSED with excessive power. Will they catch on fire in a crowded nightclub? There IS a safety consideration here, even for home systems that are subject to use at high power. You don't want to go to bed after a long and loud party with inductors smoldering inside your speaker enclosures. What's that odd smell?

For me the bottom line has been to build the prototype and then measure the temperature rise of all components (R's, L's, C's and transducers) after a period of high power. Yes, electrolytic C's can get so hot they EXPLODE in such tests. Power dissipation is also a compelling reason to use only film type capacitors in our crossovers.

Regards,

John