Gentlemen,
When an amplifier drives a loudpeaker with a crossover the back emf from the woofer will "hit" the inductor of the crossover low pass filter first which will isolate this from the amplifier. When driving a loudspeaker directly I would recommend the same technique that is used when amplifiers drive motors. In that application a series choke is used between the amplifier and motor. For driving a loudspeaker directly I would recommend the same, especially for woofers.
I don’t mean to throw a monkey wrench into things here, but I have a couple of issues with this suggestion. I ‘m not trying to refute Dan’s logic but if you look a little deeper, there are several considerations regarding this approach and the question of efficacy regarding any real benefits it may provide.
To start, the “inductor in series with the motor” technique has a two-fold benefit that is specific to that application. First, the inductor will limit start-up current draw from the motor. This helps reduce the resultant heating of both the internal windings of the motor and any electrical connections/wiring back to the power source that results from the large voltage drops across them. This extends motor life and reduces line voltage drops, i.e., voltage dips and spikes on the power line that crashes computers, etc.
Because typical AC motors are inductively coupled devices, their impedance is a function of inductive coupling of the windings and ferrous components internal to the motor. At start-up, that impedance consists of little more than the “leakage inductance” of the windings until such time as the magnetic fields within the ferrous materials have been established and the resultant impedance is reflected back onto the power line. In other words, it takes a finite amount of time for the impedance to rise to a “safer” level. During this short interval the motor “looks” almost like a dead short-circuit which is dominated by the DC resistance of the windings. Depending on the horsepower rating of the motor, this can be a very low resistance, which means huge amounts of current draw. Remember: One Horsepower is equal to about 750 watts of power so an even relatively small motor can draw a large amount of current, even when it’s running up to speed - much more so at start-up.
Again, as the field in the motor becomes established, it produces a back emf that is reflected onto the power line as a voltage spike. The series inductor helps then to reduce the magnitude of this spike as per Dan’s comments. During either event the inductor can be of significant help in moderating these current/voltage swings.
A speaker is a motor as well and as Dan points out, will exhibit similar properties to a traditional motor. The differences reside within the details. First of all, the current draw will never exceed the amplifier's output voltage divided by the DC resistance of the voice-coil (whatever form it may take). Typically in audio, most speakers exhibit DC values greater than 3-ohms. For a DC value of 3-ohms, a 17.32 Volt output will result in a power output of 100 watts, with a current draw of 5.77 amperes. Although this is not an inconsequential value, it is relatively quite small compared to a traditional motor and well within the limits of any modern amplifier with an equivalent power rating. Let’s put it this way, if it weren’t, manufacturers wouldn’t be selling many.
Also, the total current draw with AC signals is a function of the rise-time of the waveform. Faster rise-times are directly related to higher frequencies. Faster rise-times equal greater current draw into capacitive loads as well. Above a speaker's resonant frequency it represents a combined resitive/capacite load, up to the point were voice-coil inductance (Le) begins to dominate the total impedance. An added amount of inductance of the correct value has the potential to completely cancel this capacitive effect though. That’s the supposed benefit of the series inductor. The problem is, an inductor that large would severly interfere with the speakers frequency response.
The total average current is the integral of the RMS current over the period of the waveform, but instantaneous current draw is our primary concern here so we’re back to the impedance/rise-time function. The only way to raise the impedance without increasing power dissipation is to add inductance via the series inductor. The problem is with regards to a loudspeaker, in order to raise the impedance sufficiently in the frequency range over which the woofer operates, the inductor must be rather large. This will have a direct effect on the frequency response of the system and alter the chosen transfer function of the crossover network. In fact, that’s how it works. If it doesn’t affect the frequencies within the pass-band of the driver, it won’t have any appreciable effect on limiting current either. If it doesn’t appreciably attenuate current into the speaker from the amplifier, it will not limit Back EMF either – it’s a two-way street.
So…the upshot is that an inductor large enough to be of any value will necessarily alter the frequency response of the speaker. If there were higher frequencies above the operating range of the woofer being delivered to it by the amplifier, then a series inductor might be of some value. But if that were the case then there would not be any up-stream filtering to begin with. Either that, or there’s a problem with the amplifier oscillating or generating high frequency noise. If that’s true, well…you need to fix/get a better amplifier. Otherwise, all active subwoofers employ an active crossover and thereby attenuate higher frequencies by the very nature of the crossover function - so, no benefit there.
Back EMF voltages at higher frequencies, such that they are capable of stressing the feedback circuit of an amplifier, have to be "excited" in the driver by an equivalently (or nearly so) high frequency signal from the amplifier in order to come about in the first place. The amplifier's active crossover prevents such frequencies from ever reaching the driver, so Back EMF at higher frequencies will virtually never be generated to begin with. If it’s a passive sub, then the inductor is already present so no problem there either.
In a full range speaker such as the Magnapan, a series inductor may limit current/Back EMF, but it will knock the high frequencies off too. Unless you like speakers with virtually no high-end frequency response, I can’t imagine doing that.
So as I see it, there’s little if any benefit to using a series inductor for limiting Back EMF in a loudspeaker. In lieu of such, your only viable alternative is to find an amplifier that is of such quality design as to not induce errors of its own as a result of a faulty error correction circuit. Either that, or select a low/zero feedback design and forget about correcting for Back EMF in the first place.
Think about it, if loudspeaker manufacturers could simply throw an inductor in series with their speaker to make it an easier load on the amplifier – without altering the speaker’s frequency response – then we’d all be doing it! I know I would. Unfortunately, there’s just no “free lunch’ here.
There is one little twist to all of this. One could develope a "conjugate match" Zobel network that would be wired in parallel with the speaker. This network would consist of a single coil, capacitor and resistor with values such to cancel the impedance rise at the speaker's box resonance frequency. This would have the tendancy to damp any Back EMF seen by the amplifier generated around this frequency range. Then again, there are two potential drawbacks to this. First, depending on the output impedance of the amplifier, it could severely over-dampen the bass response - we're talking dead here - almost no bass. Second, it could be a waste of money as the components would have to be quite large in value and therefore expensive. If the selected amplifier has even reasonable low frequency performance, it should handle any Back EMF at those frequencies quite easily - so what's the point?
OK, back to working in the yard. I can’t milk it any longer. I hope the above is taken in the spirit is intended. All I hoped to accomplish was to examine this issue a little more deeply. I certainly hope I didn’t offend the other posters here and I certainly don’t mean to come across as a “know it all.” If someone here sees were I’m off base, please feel free to let me know. I hope to never be above learning and being corrected in my logic is the first prerequisite in order to learn more. Hope this is of some help.
-Bob