The plot below is a near-field SINE measurement frequency/SPL. This NHT1259 has an Fs of 24.8 Hz, and was installed in a 2.65 cubic feet box with 50% stuffing. The box was deliberately made smaller than the "recommended" 3.3 cubic feet. The goal of this exercise is to electronically assist this design. As such reduced box size will be compensated by active filtering with electronic correction and amplifier power.
The 100% stuffed box (not shown) improved the low response with approximately 1 dB at 20 Hz. The in-box resonance frequency of the NHT1259 came down another 1.1 Hz. Qtc dropped from 1.048 to 0.970. The non-stuffed box resulted in a Fc of 43.4 Hz, the 50% + 50 % stuffed box came down to 37.5 Hz, a reduction of approximately 16%.
The next nearfield measurement shows the effect of break in on the performance of a driver in a sealed enclosure. Fc not broken in was 37.5 with a Qtc of 0.97. After a day of use in the studio on some laser disco programming, a new measurement was made. Fc dropped down to 35.22 Hz, Qtc settled at 0.86. These are not particular audible changes, definitely measurable. The difficult part of this measurement was to recreate exactly the same set-up for the nearfield bass data acquisition. Red is not broken in, white is after break in.
The graph below shows the frequency response of the Linkwitz pole shifter. The box parameters were Fc of 35.22 Hz with a Qtc of 0.86. The "desired parameters" were 25 Hz with a Qtc of 0.707. Modeling was done in Pspice. A nice feature of the Linkwitz "bass" correction is that it takes only a single op-amp to accomplish this curve. Observe that there is a dip in the curve at the same location of the peak in the near-field measurement of the above graph. The maximum gain was kept at 6 dB @ 20 Hz by choosing the "new" Fc at 25 Hz.
After including the Linkwitz correction unit in a prototype Clearview CXR-22, the near-field bass response of the NHT1259 in a 2.65 Cft is shown in the red plot below. The low pas filter was set at 140 Hz, the Q of the 20 Hz high-pass filter was set at approximately 0.55. The white plot shows the response with the Linkwitz correction engaged. The F3 frequency of the red plot is 37.5 Hz, the F3 frequency of the system with Linkwitz correction incorporated is 27 Hz, an improvement of over half an octave. This still could be improved by precise matching of the capacitors in the filter without compromising linear X-max of the driver. The area from 40-70 Hz could be flattened a little more
There is still a Q control to play with on the Clearview CXR-22. This allows to tailor the system Q and influence the F3 frequency. The plot below shows the response of the crossover for 5 settings of the Q control with the Linkwitz correction in the circuit.
All measurements in this project were performed with the SINE instrument of Laud. Frequency range was 10-500 Hz, no smoothing applied. These were all nearfield measurements at 1/4 inch from the dustcap.
The parameters for this project were chosen so that for a-smaller-than-recommended box with Linkwitz correction the performance was similar to a significantly larger sealed box. Maximum excursion in the 2.65 Cft box can still be reached without violating thermal voicecoil conditions.
On March 1, 1998 I had the opportunity to measure the above driver in a large gym. (Thanks John.) The plot below shows two curves of the NHT1259 in the 2.65 Cft sealed enclosure and an RD75 on a dipole baffle. These drivers were crossed over with a CXR 22.
The red trace has Linkwitz correction engaged, the white trace has the Linkwitz correction switched off. Noticeable difference. The measurement was made at a practical listening distance of 3 meter with the microphone at 36" high. Measurement method: 16 k MLS with 16 k FFT, 1/12th octave smoothing applied. Notice also the smooth transition at the crossover point. The dips at approximately 500 and 1000 Hz are dipole cancellation dips.
Next graph shows 5 measurements on the NHT1259 in a 2.65 Cft sealed enclosure. These measurements were taken without crossover. Observe the first major dip in each trace. This dips moves to a higher frequency as the distance between driver and microphone increases. The difference in travel distance between the driver and the bounced mirror image becomes indeed smaller with increasing distance, hence the shift in cancellation frequency. This example shows also that the floorbounce from the subwoofer is not so much a problem as the effects will be dealt with in the crossover. The driver's center in this particular case was 9" above the floor. Moving the driver higher in the enclosure would lower the frequency where the dip occurs.
Observe also the yellow plot at 3 meter and compare with the white trace of the previous plot. Both show the NHT1259 in the sealed enclossure at the same distance. The white trace in the previous plot however has the crossover point between 100 and 200 Hz. It is not possible to distinguish the crossover pont by comparing both curves, showing that the CXR 22 does an excellent job at integrating a sealed NHT1259 and a RD75 in a dipole application.
Copyright (c) 1997, by Rudi A. Blondia, ALL RIGHTS RESERVED. Last update: March 2, 1998.