In an ideal world, a loudspeaker should present a resistive load to the power amplifier. However, in the real word, it is far from this.

The majority of moving coil loudspeakers are reactive loads. In electrical terms, what this means is the resistance is not constant but changes with frequency. Reviews frequently emphasize how low the impedance of a speaker as a measure of how demanding the loudspeaker is to the power amplifier.

Impedance by itself, is not the only parameter affecting amplifier. What is often taken for granted is the electrical phase displacement. This can place even more stress on the power amplifier than low ohms.

Let's go back to some basic electrical fundamentals. The best power transfer is when the voltage and current are in phase. This would be a purely resistive load, like a filament bulb. What happens when the load is purely capacitive? There will be a phase shift where the voltage lags the current by 90 degrees. When the load is purely inductive, the opposite happens, ie the voltage leads the current by 90 degrees. Whether it leads or lags, the end result of phase displacement is more power needs to be delivered to the load.

Take Swift-III as an example (Fig 29). This kind of impedance and phase responses are typical in speakers (Fig 30). In the crossover region, the power amplifier is subjected to two different demands.

One is Impedance. The impedance at 2 ohms means the speaker wants more power from the amplifier.

Next is the phase shift. Quite a bit of power is not used by the speaker due to the +/- 45 degrees phase shift. To make up for it, the power amplifier must now deliver more power than normal. Looking at it another way, Phase Shift = Power Inefficiency.

The Swift Ultima (Fig 31) is a near perfect load for power amplifiers. At 500Hz upwards, there are no wild fluctuations in impedance and phase. The impedance gradually lowers to 6 ohms at 20kHz. Phase is contained within a tight -8 degrees window, behaving in a linear fashion to 20kHz. This is crucial as the crossover and the bulk of musical information reside in this bandwidth.

Below 500Hz, the phase gradually dips to -45 degrees at 100Hz. Eventhough the phase angle is -45, the impedance has gone up to 20 ohms. Power amplifiers would not have any problems with this. If it had been -45 degrees at 2 ohms, that would be a serious issue.

Fig 32 is the Distortion plot of the Ultima. THD (blue trace) is lower than 0.3% (-50dB) from 500Hz - 20kHz, consisting predominately 2nd harmonics (red trace).

The frequency response of the Ultima is an improvement over Swift-IIIc. The tweeter response is level instead of tapering down. This is from an additional EQ network.

The Ultima crossover is very similar to Swift-IIIc. There's no change to the network for the ZA14. For the XT25TG30, a 3uF in parallel with a 2.2 ohm resistor at the front end serves to EQ the tapering response.