- Written by Chris Groppi
- Published on 02 October 2008
On the Bench
I used a Smith and Larson Woofer Tester to measure the Druid's impedance as a function of frequency. The results are shown in two ways in the plots: the traditional amplitude and phase plots, and the smith chart plot I wrote about several months ago. Impedance stays at 8 ohms or higher for the entire audio band, with large phase angle deviations seen only at low frequencies. These magnitude and phase angle excursions are due to the port tuning of the loudspeaker enclosure. Phase angle rises gently in the treble. Looking at these results on the Smith chart, normalized to 16 ohms, the impedance trace stays nicely near the center of the plot, more or less within my arbitrary +/- 50% impedance circle. Most of frequency space is well within the circle. The endpoint near the center is 20 kHz, and the marker cross is at 1 kHz. Most of the data slightly outside the circle is at low frequencies, and occupies a small range of frequencies. This makes the Druid a relatively easy loudspeaker to drive, with high average impedance throughout the audio band, and no weird reactive impedance deviations.
I also measured frequency response for the first time with any of my reviews. I used a calibrated Earthworks M30BX microphone with a Roland Eridol UA-101 24 bit 96 kHz USB soundcard and SpectraPlus FFT analysis software. John Johnson, the editor of Secrets, provided a calibration file for SpectraPlus. Absolute calibration was done with a Radio Shack SPL meter. The microphone was mounted on a boom stand on axis with the tweeter, and placed about 1 meter away (after calibration).
The results of playing a 0 dB full-scale white noise signal are shown here. I would completely ignore all frequencies below 200 Hz in detail. All the structure you see, save the roll off at 40 Hz, is due to my room. My room is terrible in the bass. I had to control huge resonances with bass absorbers, which have made the corresponding suckouts worse. The general trend of a gentle roll off below 100 Hz towards 40 Hz, with a more aggressive roll-off after that is the main point. Midrange and treble performance is rather flat, with a bit of a dip at around 12 kHz, at the crossover frequency between the wideband driver and supertweeter. As I can't have the microphone on axis witht the wideband and supertweeter simultaneously, I would guess that this is also a measurement effect. Again, most of the small scale structure is due to room interaction. If I had an anechoic chamber, the response would be smoothed out quite a bit, but I don't.
As this is my first try at loudspeaker measurement, these results may be replaced with better, and possibly additional measurements in the next week or two.