How do inductors help impedence matching?

Newbee question: Can someone explain to me how a coil used for low pass filtering helps impedence matching with the amp? My speaker designer told me that the purpose of the 1.5mH inductor on my woofer crossover is primarily impedence matching, rather than as a filter. When I remove the coil from the crossover circuit, the sound changes very little. I would like to bypass the inductor, unless this might cause other problems. Thanks much.
Presumably there is a capacitor that couples the tweeter to the amplifier for higher frequencies. This lowers impedance at these frequencies. The inductor that couples the woofer to the amplifier increases the impedance of the woofer circuit for higher frequencies. If you remove the inductor the woofer will be paralleled with the tweeter at higher frequencies, so that there will be a lower overall impedance at these frequencies.

The frequency response of the woofer is probably going downhill at the crossover frequency, so you may not hear much of a change when you remove the inductor.

However, manufacturers generally don't use components that are not needed. I would leave it alone, or, if you must tweek, replace it with an air core inductor of the same electrical value.
Eldatford - if I understand properly, a cap can be seen to decrease impedence as freq increases, while a coil can be seen to increase impedence as freqs go lower. The resulting overall impedence curve vs. freq is flatter than it would be without the crossover components. Am I correct?

A related question: other than experimentation, how might I predict effects of changing output taps on my SET amp from 4, 8, and 16 ohms? I do plan to experiment, but it isn't trivial because I have to reconfigure the output taps of the transformer each time. And ohms law doesn't apply, since we are discussing AC, correct?

Thank you for your response - I really appreciate the opportunity to learn.

Skushino...About the impedance...correct.

Think of the transformer as being like an automobile transmission. The taps are like different gears. Suppose you want to get 2 volts out of the amp, and you set the volume so as to get this from the 8 ohm tap. The 4 ohm tap will have 1 volt, and the 16 ohm tap will have 4 volts. So you would need to boost the volume if you connect to the 4 ohm tap, and you could reduce it if you used the 16 ohm tap. Use of the 4 ohm tap might mean that the amp circuit couldn't put out enough voltage, and use of the 16 ohm tap might mean that the amplifier circuit couldn't put out enough current. Sort of like trying to go up a steep hill in high gear. The speaker impedance, modified by the transformer, is the load seen by the amp circuitry. Use of the correct tap will optimize power delivery from the amp circuit.
It depends if the inductor is in parallel or in series with the woofer. If it's in series, then its function is to increase the inductive reactance (why though, it's easier done with a resistor) and if it's in parallel, its there to reduce the reactance - possibly to match with the impedance of the other drivers (because the woofer coil is an inductor).

If you remove the inductor and it changes the sound, that is not unexpected. Whether it's better or worse - who knows- but I would agree that extra components are there for a reason and should not be removed. Crossovers filter because of the inherent impedance mismatching the components provide so changing the relationship will affect the performance of the speaker in one way or another.
OK, continuing the automobile transmission analogy, the load presented by the speaker to the amp would be like the steepness of the hill, correct? The steeper the hill (more difficult speaker load, or lower impedence), the more current needed to go. And a more benign speaker load would be like a gradual incline, requiring less current and benefiting from more voltage, right?

I'm driving very high-efficiency horns ~108dB sensitivity. I was first going to try the 4ohm tap, striving for more current to the 15" woofer. Instead, would it be a better idea to try the 16ohm tap for better power transfer compared to the present 8 ohm tap?

bty, listening without the coil filter is to my ears, much clearer and "transparent".

Scott. you seem to lack the understanding of basic concepts of crossovers as they relate to inductors. I am sure Bruce can give you some good info. I have a 5-way edgar horn system, btw. removing the coil is going to basically remove the low-pass on the woofer. The reason it sounds clearer is the woofer is running full range most likely. inductors affect a lot more than impedance, they affect frequency response. go model the crossover using existing values, in an online crossover program and see what the points and slopes are. Swapping taps on the output of the amp is going to fool with the sound even more. you are better off leaving it at 8 ohm tap most likely. Anything different is going to reflect a different impedance to the output tube through the OPT. Frankly, this sort of discussion is better off at H.E. asylum where there are far more skilled people to reply (than here) incl. Bruce Edgar himself. Basically, leave the Xover alone, you stand to do (more) damage to the sound and the gear by fooling with it and not knowing what you are doing . BTW, lose the Fane. it sounds bad in comparison to the better Fostex like T925A and T900A. Bruce himself is going with Fostex in his stuff now, generally (for those that want the best sonics). Fane is efficient but that's all it is. It's harsh and grating compared to Fostex, I have owned all 3 + EV, ST-350, etc.

An inductor of 1.5mH would correspond to an 850Hz crossover for 8 ohm driver (and 1700 Hz for a 4 ohm driver). It must be in series.

If the speaker is 8 ohms, the 8 ohm tap will maximize power. Try the others if you will not hurt anything. If your ears are happy without the inductor by all means leave it that way. What you have done is to use the 15" driver to a higher frequency, and attenuated the horn tweeter by leaving the woofer in parallel with it. Attenuating a horn tweeter is not a bad idea (IMHO).
Coils (inductances) Decrease Impedance as Frequency gets lower. The formula is Inductive Reactance = 2 times Pi times the Frequency times the Inductace value. As frequencies go lower, so does the inductive reactance. The Impedance is the vector sum of the Inductive reactance and any pure resistance in the circuit, which is not frequency-dependent like the reactance; therefore the Impedence of a coil decreases with decreasing frequency (and increases with higher frequencies). Capacative reactance is the opposite- Lower frequencies have higher reactance and vice versa.

I actually just had dealt with this issue when I was building a small circuit to make an oscilloscope display like the ones found in Marantz tuners. When using a Capacitor and Resistor (for phase shift) the lissajous pattern was kinda blurry. When I changed the capacitor with a coil, the blurriness went away. This is because the capacitor blocked the low freqencies and passed the high ones (more fuzzy display). With the coil the lows were passed and the highs were blocked, so the pattern was less fuzzy.
Thanks for many informative responses.

Gs5556 - the inductor is in series, used as a low-pass filter and "impedence matching" component.

Ed_sawyer - have read many of your posts here and at AA, and found some of them imformative. In this instance, your reply is a statement of the obvious:

"removing the coil is going to basically remove the low-pass on the woofer" (thank you...)

"go model the crossover using existing values, in an online crossover program and see what the points and slopes are" (As Bruce says, the model has little bearing on resulting sound. Besides, I did model the xover about 4 weeks ago)

"you are better off leaving it at 8 ohm tap most likely" (why? have you tried the other taps? I'm interested)

"Bruce Edgar himself" (yes, we talk)

"Basically, leave the Xover alone, you stand to do (more) damage to the sound and the gear by fooling with it and not knowing what you are doing" (as Bruce says, the crossover values he uses are the result of an iterative process, and the Titan and Edgarhorn values are the results of his years of listening. My MF driver is different than his)

I am still forming an impression about the Fane, but in general agree with you. Am very interested in listening to the Fostex and T350. Ed, I would be very grateful if you share your experience with these in your Edgarhorns.

I asked alot of basic questions in different ways because it helps me to learn and gain a deeper understanding, which I try to apply in my system.

Joeylawn36111 - thanks for your explanation

Eldartford - you're priceless. Thank you.