This section is dedicated to the analysis, testing and modeling of high performance woofers for subwoofer use. A significant number of drivers are currently available to the DIY through different suppliers.
Goal is to test the available large excursion drivers, verify the T/S parameters, model at least one alignment and verify the correlation through measurements. Distortion measurements in function of applied power are scheduled to further qualify the drivers under large signal conditions.
T/S parameters are measured using LAUD 3.12. The applied voltage to the driver under test is 1 Vrms. The added mass method is used. (120.0 gram) Dual voice coil drivers are parametrized with single voice coil driven, both coils in series and both coils in parallel. Re is measured with a recently calibrated Fluke 87, accurate to 0.02 ohms.
The T/S test amplifier has a frequency response from 3 Hz to 100 kHz, -3dB or 5 Hz to 40 kHz, -1dB and is stable in any load.
The acoustic measurements are performed using LAUD 3.12. Microphone used is an ACO Pacific, model 7012, response flat from DC to15 kHz, 0.5 dB up at 20 kHz, 1 dB up at 30 kHz. This condenser microphone will handle 164 dB SPL with less than 3% harmonic distortion. The latter was considered a must for large signal testing.
A CinePro 1K2 in bridged mode is used for large signal testing of the transducers as it is capable of delivering over 1 kWatt to the load.
Distortion testing is brutal on drivers and on the researcher ;-). Hardly any driver manufacturer cares to publish distortion figures on their drivers for obvious reasons. Measurements should actually be done in an anechoic environment. By lack of an anechoic room, I opted for a 10" (25 cm) nearfield measurement. I found this to be a decent compromise between room reflections and SPL handling of the microphone.
All distortion measurements are done using swept sinewaves. Each driver has power/distortion specs measured with two different sine stimuli. One series of measurements is performed using 25 cycles (long burst) and the other series takes 5 cycle (short burst) sinewaves. The long burst tends to give a better distortion "average" but shows clearly the effect of compression as the voice coil has ample time to heat up with each burst. The short burst sine shows lower distortion figures but may be less accurate.
Per stimuli type there will be a graph showing "total" harmonic distortion THD+N and another graph showing only second harmonic distortion for comparison. The total harmonic distortion measurement is actually the summation of the ten first harmonics of the fundamental.
Distortion will be measured over the 10Hz to 1 kHz frequency range.
Modeling is performed in LEAP 4.6. This modeling program is powerful and accurate. All modeling is performed for the same size box, sealed or vented applications. This results in different system Q factors but will give a clear indicator which drivers can be fitted in small boxes. Smaller box size is in a lot of applications just a SOAF must. Smaller box sizes are prime candidates for electronically assisted alignments. In these cases box size is traded in for amplifier power.
Copyright (c) 1997-99, by Rudi A. Blondia, ALL RIGHTS RESERVED. Last update: May 10, 1999.