Why use a Waveguide?

Put very simply, a waveguide is an acoustic amplifier. Fig 1 below illustrates the point succinctly. The Black trace is of the XT25TG30 with the original faceplate. The Red trace is with the faceplate replaced by a Pellegrene 6.5"x4.5" Ellipitical Waveguide. For the same power input, the SPL for the waveguide is higher, approximately 8dB on average.

Another property is the shape of the response. It's much smoother and flat from 2kHz-8kHz before tapering off linearly to 20kHz. The roll-off is steeper (almost 12dB/oct) and more linear. These are very favorable conditions where crossovers are concerned.

But the idea behind using a horn or waveguide is not for the higher SPL only. That can easily be achieved with a higher powered tweeter. By focusing the soundwaves toward the listener instead of allowing them to scatter across the room, a waveguide improves on the speaker's imaging and clarity.

Fig 3 is the distortion of the raw XT25TG30. For this measurement, a 170uF capacitor was inserted for protection. The capacitor serves to reduce the bass frequency entering the tweeter.

High levels of distortion, from 2nd to 5th harmonics, were observed from 250Hz to 350Hz. We should be mindful of this inherent distortion when using this tweeter.

Fig 4 is the XT25TG30 with a 2.5kHz, 12dB/oct filter. The XT25TG30 distortion at 250Hz to 350Hz has completely vanished. THD (Blue trace) registered a low 0.18% (-55dB attenuation).

For this plot, no LCR was installed. It goes to show that at a high enough fc, even a moderate roll-off from a 12dB/oct filter is sufficient to arrest the built-in distortion.

As the fc is lowered, it would be prudent to make further measurements to ensure that the distortion of the XT25TG30 is well suppressed.

We now turn our attention to the 6.5" Silver Flute W17RC. This Silver Flute is the Shielded version, so the T/S parameters may not be exactly the same as the non-shielded version.

Fig 5 is the W17RC with a 18dB/oct filter with fc at 2.5kHz. Two distortion peaks (THD=0.56%) were recorded, one from 2nd harmonics at 1kHz while the other from 3rd harmonics at 1.75kHz.

Fortunately, these distortions are inaudible, at least to me. I can live with some imperfections considering this is a relatively cheap mid-woofer. The postive attributes of the W17RC far outweigh the negatives.

Fig 6 is the Frequency Response of the Canary. The W17RC is with an 18dB/oct filter. The XT25 with the Waveguide is filtered at 12dB/oct.

Notice how closely the acoustic slopes resemble the theoretical roll-off.

Crossing the Waveguide at 2.5kHz does not cause any distortion in the tweeter. The two distortion peaks of the Silver Flutes do not interfere with music playback. Overall, the Canary distortion is acceptable.

The Canary presents an easy load for power amplifiers. Though the Impedance (Fig 8) is not as linear as I would like it to be, it doesn't dipped below 9 ohms. The electrical Phase is within a 20° window.

Power amplifiers like the LM3886TF are ideal for the Canary. The LM3886 can be designed to cater for 8 ohm loads instead of 4 ohms. This will maximize the voltage swing, resulting in higher power output.