The driver MISCO sent me to be characterized this month is from its Bold North Audio product line. The first Bold North release was a nearly exact replacement woofer (down to the seamed white cone) for the Yamaha NS10 studio monitor originally popularized by recording engineer legend Bob Clearmountain with his infamous high-frequency tweeter filter fashioned from tissue paper. The 6.5” MS-10W was featured in the February 2019 issue of Voice Coil. However, the Bold North lineup was recently expanded to include a new 29mm dome tweeter, a woven carbon fiber 4” bass/midrange, and two new 6.5” XBL2 motor woofers. Previously, Test Bench featured the new BDT-2901 high-end tweeter, and the equally high-end BWX-6501 6.5” XBL2 woofer/midbass driver. This month, I received another 6.5” XBL2 woofer, the BWX-6502.
The BWX-6502 is built on the same platform with a similar motor structure to the BWX-6501, however, the BWX-6502 has some interesting differences. As such, both of the XBL2 woofer/midbass drivers share the same substantial feature set that begins with a proprietary 12-spoke cast-aluminum frame, comprised of a set of four groups of three closely grouped narrow (about 4mm) tapered spokes, and includes 16 8mm X 4mm vents below the spider (damper) mounting shelf.
Additional cooling for this driver is provided by eight 5mm diameter peripheral vents in the motor return cup, as seen in the cutaway drawing in Figure 1. The cone assembly consists of a curvilinear profile Abaca fiber (paper) cone and a 60mm (2.4”) diameter Abaca fiber dust cap (Photo 1). Compliance is provided by wide (for a 6.5”) high excursion NBR surround, with the remaining compliance supplied by a 3.5” diameter flat cotton spider (damper).
The XBL2 motor design is licensed from inventor Dan Wiggins for both of the new Bold North Audio woofers. Similar in concept to the Mmag patented dual-gap dual-magnet motor technology developed by Enrique Stiles and Patrick Turnmire, the XBL2 motor features a short voice coil traveling between two gaps. While Mmag uses two magnets and two plates, XBL2 creates two gaps generally by using a thick front plate with a CNC’d groove in the middle of the plate to form the dual gap structure. However, the BWX-6502 forms the gap with two milled front plates. This methodology results in a typically shorter voice coil that travels into the second gap as it is traveling out of the first gap, the upshot being a higher Xmax than most conventional motors can provide. Specifically, the stated Xmax for this driver is actually not calculated by the usual formula (Hvc – Hag)/2, but by the Klippel criteria for Bl decreasing to 82% for the Bl value at the magnetic center. This is also the Klippel criteria for determining the contribution to a 10% distortion figure. For the BWX-6502, the excursion number is 10mm.
The BWX-6502’s two milled front plates are fitted with dual inside and outside copper shorting rings (Faraday shields). Driving the cone assembly is a 38.1mm (1.5”) diameter voice coil wound with round copper-clad aluminum wire (CCAW) on a nonconducting Kapton former. Last, the driver’s IEC268-5 power handing is rated at 85W and the voice coil is terminated to standard solderable terminals.
I began testing the BWX-6502 woofer using the LinearX LMS analyzer and LinearX VIBox to create both voltage and admittance (current) curves with the driver clamped to a rigid test fixture in free-air at 0.3V, 1V, 3V, 6V, 10V, 15V, and 20V. The BWX-6502’s free-air multi-voltage impedance protocol pretty much matched the BWX-6501 from last month with the highest useable voltage sweep being 20V. While I did not use the 20V curves, the 20V sweep potentially could have been used as the curve fit was close, so it was a judgment call. Either way, most 6.5” woofers don’t make it past 10V before extreme nonlinearity sets in as the number of voice coil turns in the gap is decreasing as the coil is riding out of the gap.
Following my established protocol, I no longer use a single added mass measurement and instead use the physically measured Mmd data (21.3 grams for the BWX-6502). The collected data, in this case the 12 550 point (0.3V-15V) sine wave sweeps for each Bold North sample were post-processed and the voltage curves divided by the current curves to generate impedance curves. The phase was derived using the LMS calculation method, and along with the accompanying voltage curves, imported to the LEAP 5 Enclosure Shop software software’s Guide Curve library.
This data was then used to calculate parameters using the LEAP 5 LTD transducer model. Because most all manufacturing data is being produced using either a standard transducer model or in many cases the LEAP 4 TSL model, I also generated LEAP 4 TSL model parameters using the 1V free-air that can be compared with the manufacturer’s data. Figure 2 shows the BWX-6502’s 1V free-air impedance plot. Table 1 compares the LEAP 5 LTD and LEAP 4 TSL Thiele-Small parameter (TSP) sets for the two Bold North driver samples along with the Bold North factory data.
LEAP LTD and TSL parameter calculation results for the BWX-6502 appear to not correlate well with the factory published data, however, the Fs/Qt ratios were close. When I programmed in the factory TSP into LEAP, and did a closed-box simulation, the frequency response results were nearly identical and the two simulations overlaid each other, even with the factory data using a more conservative Sd number. Xmax for the driver, published by MISCO, is actually the XBl number calculated in the Klippel LSI file, which if you have been reading Test Bench reports for a while, is the excursion limit as Bl at rest falls to 82% of its value (the criteria for 10% distortion). Since this is not a calculated Xmax using the driver’s motor geometry, but a reflection of Klippel LSI maximum excursion at 10% distortion. I used that number, 10mm, to compute the maximum linear SPL rather than my usual Xmax + 15% criteria.
Next, I proceeded to set up the computer enclosure simulations using the LEAP TSL parameters for Sample 1. I programmed two simulated enclosures into the LEAP 5 software — one Butterworth Qtc=0.7 sealed box with 0.46ft3 air volume with 50% damping material (fiberglass); and a vented Chebychev/Butterworth) alignment with a 0.85ft3 volume tuned to 32Hz with 15% damping material.
Figure 3 displays the box simulation results for the BWX-6502 woofer’s sealed and vented enclosures at 2.83V and at a voltage level that achieves the 10mm XBl limit, since I am treating this as an “Xmax” number. This resulted in a F3 of 52Hz (-6dB=42Hz) with a Qtc=0.7 for the 0.46ft3 closed box and a -3dB for the vented simulation of 40Hz (-6dB=32Hz).
Increasing the voltage input to the simulations until the Klippel LSI XBl limit of the 10mm excursion was reached resulted in 104dB at 32V for the sealed enclosure simulation and 106.5dB with a 66V input level for the larger vented box. Figure 4 shows the 2.83V group delay curves. Figure 5 shows the 32V/36V excursion curves.
Klippel analysis for the BWX-6502, which this month was performed by Warkwyn (the US purveyor of Klippel GmbH products), produced the Klippel data graphs given in Figures 6-9. The Bl(X) curve for the BWX-6502 seen in Figure 6 is rather broad, especially for a 6.5” woofer, and with fairly symmetrical behavior with a degree of “tilt” coil-in to coil-out. The Bl symmetry curve in Figure 7 shows a 1.0mm Bl coil-out (forward) offset once you reach an area of reasonable certainty and progressively decreasing to a trivial 0.6mm coil-out offset at the 10mm excursion point.
Figure 8 and Figure 9 show the Kms(X) and Kms symmetry curves. The Kms stiffness of suspension curve seen in Figure 8 is also reasonably symmetrical and with a degree of rearward (coil-in) offset. The Kms symmetry range curve shown in Figure 9 exhibits 1.5mm coil-in (rearward) offset at a region of high certainty (6mm) increasing to about 1.85mm at the XBL 10mm location.
Displacement limiting numbers calculated by the Klippel analyzer using the full-range woofer criteria for Bl was XBl @82% (Bl dropping to 82% of its maximum value) equal to great than 9.97mm for the prescribed 10% distortion level. For the compliance, XC @75%, Cms minimum was only 9.20mm, which means that for the BWX-6502, the compliance is just slightly more of a limiting factor for getting to the 10% distortion level.
Figure 10 gives the inductance curve Le(X) for this transducer. Motor inductance will typically increase in the rear direction from the zero rest position as the voice coil covers more of pole in a conventional motor, which is not what you see in the graph due to the motor configuration. However, the inductive “swing” from maximum inductance to minimum inductance (from 10mm coil-in to 10mm coil-out) is a very small 0.015mH, which is quite good.
For the remaining test procedures, I mounted the BWX-6502 in a foam-filled enclosure that had a 17” × 8” baffle. Then, I measured the device under test (DUT) using the LoudSoft FINE R+D analyzer and the GRAS 46BE microphone (courtesy of LoudSoft and GRAS Sound & Vibration) both on- and off-axis from 200Hz to 20kHz at 2.0V/0.5m, normalized to 2.83V/1m, using the cosine windowed Fast Fourier Transform (FFT) method. All of these SPL measurements also included a 1/6 octave smoothing, which approximate the 100-point frequency response resolution that I used with LMS for a number of years.
Figure 11 gives the BWX-6502’s on-axis response indicating a fairly smooth rising response with no break-up modes or peaking out to about 3kHz, with a peak in the response at 4.6kHz and 6.2kHz, where the driver begins its low-pass roll-off. Figure 12 displays the on- and off-axis frequency response at 0°, 15°, 30°, and 45°, -3dB at 30° with respect to the on-axis curve occurs at 2.6kHz, so a cross point in that vicinity should be work well to achieve a good power response. Figure 13 gives the normalized version of Figure 12, while Figure 14 displays the CLIO Pocket horizontal polar plot (in 10° increments).
And finally, Figure 15 provides the two-sample SPL comparisons for the BWX 6502, showing a close match (less than 0.9dB) to about 3kHz, with some variation above that, which will of course be in the stop-band of a low-pass filter.
For the remaining series of tests, I employed the Listen, Inc. SoundCheck AudioConnect analyzer and 1/4” SCM microphone (graciously supplied by the folks at Listen, Inc.) to measure distortion and generate time-frequency plots.
For the distortion measurement, I mounted the BWX-6502 rigidly in free-air, and set the SPL to 94dB (my criteria for home audio transducers) at 1m (7.52V), using a SoundCheck pink noise stimulus. Then, I measured the distortion with the Listen microphone placed 10cm from the driver. This produced the distortion curves shown in Figure 16.
Next, I employed the SoundCheck software to get a 2.83V/1m impulse response for this driver and imported the data into Listen’s SoundMap Time/Frequency software. Figure 17 shows the resulting cumulative spectral decay (CSD) waterfall plot. Figure 18 shows the Wigner-Ville plot.
Looking at all the various data I collected for the new Bold North Audio BWX-6502 6.5” midbass driver and considering it’s BLX2 high excursion capability, as well the overall design and build quality, this driver, like the BWX 6501 from last month, is a really well-crafted product for the high-end hi-fi market. For more information about the Bold North Audio line, visit
www.miscospeakers.com/product-lines/bold-north-audio. VC
This article was originally published in Voice Coil, January 2021.
