Sub for the King – Special Designed Subwoofer for an Unique Space

Campbell’s Corner – Dick Campbell

Dick Campbell designed an subwoofer for a rather unique space that needed to meet the final specification. Read more about it.

With salami, of course. Recently, I measured a custom subwoofer in my studio using CLIO [1] with a sine sweep at moderate power. The phone rang almost immediately – it was my neighbor in the next condo wondering if I heard that awful sound? What did I think it was? House-shaking explanation accepted!

If you read Sound and Communications, you may have noticed the article entitled “All Hail the ‘King’” [2]. Clayton Acoustics Group [3] was the consulting firm hired to do the new audio system in Atlanta’s Cathedral of Christ the King. I am mentioned in the article as provider of a special subwoofer and I would like to discuss some details of this design.

The upper cutoff frequency was originally specified to be 125Hz. Otherwise it was ‘do your best’ from there to design a proper sub, but 40Hz to 125Hz was the final specification. The only rigid thing about it was that it had to fit under a rear transept pew, avoid kneeling feet, and it had to produce a specified output with acceptable distortion.

The available space allowed a box with interior dimensions of 107″ long, 12″ high and 8″ deep. I selected the venerable Swan 305 [4] 12″ driver and placed one driver at each end of the cabinet angled 450 to accommodate the mounting diameter and magnet structure. Dan Clayton created a mechanical drawing (see Fig. 1) and subsequently approved the carpenter’s sketch.

Top view of subwoofer cabinet

I used AkAbak [5] to predict the performance. This software is remarkable in that it allows connection of myriad acoustic devices in a net to create a complete model. Such details as the diameter and depth of the dust cap and the driver mounting angle can be included.

I chose to model the cabinet in three sections – the joints between the end sections and center section are where the two ports are located, facing forward (what a thrill!). I try to keep the port peak velocity below 0.1 Mach (higher can generate annoying chuffing sound) and, luckily, the box depth allowed sufficient space for two three-inch diameter ports with adequate clearance between the inside of the pipe and the box wall. The inside ends of the ports were flared slightly . By placing an AkAbak physical boundary at the port locations, I can also examine the box internal pressure there.

The length of the cabinet suggests undesirable wavelength problems. Both ends are driving points so there is no organ-pipe resonance (no passive termination) but at a full wave we can anticipate problems at 143Hz representing the acoustic length of 107″ less 1/2 the sloping end length (15″). This can be seen in Fig. 2, showing the ominous but only 4-5dB bump at this frequency, affecting all curves.

Subwoofer measurements

On the electrical side, a filter may be calculated and placed in the AkAbak script, either polynomial, or synthesized active or R-L-C component-based. Why a filter? The driver excursion below port resonance rapidly goes through the roof and reaches its specified limits in less than two octaves. Another reason is the out-of-band output of the loudspeaker at high frequencies will have to be attenuated.

Notice in Fig. 2 that the AkAbak prediction is within +/- 2dB of the measured sum from 25Hz to 175Hz. I would estimate the useful output as from 35Hz to 120Hz.

Examining the driver excursion below port resonance, Fig. 3, reveals that a second-order band-pass filter from 35Hz to 120Hz will prevent over-excursion and attenuate out-of-band output

Graph showing band-pass filter of the subwoofer

The driver excursion is now under control and this filter is to be inserted in the signal chain to the subwoofer power amplifier. After filtering, the sub drivers reach 9mm peak excursion at 300w input generating 111dB SPL in free-field at 2m. What they do in reality when mounted on the floor under the pew is anyone’s guess. I expect its presence is heard.

Dan Clayton reports: “In rehearsal and testing the sub could absolutely overwhelm the two Intellivox DS-1608 large-format, high-output line-array loudspeakers. We could get it WAY too loud and it still sounded pretty good. [the church] is a very bass-friendly space: simple geometry, massive masonry construction, very little effective LF absorption. The sub normally runs at a very modest output level…”

So if you are in Atlanta, visit the ‘King’ and kneel in the left transept rear pew with a leg on each side of seat center support. Remove your shoes, and wait for the bass to hit the right note. I’m sure that your neighbors will be alarmed at your expression! Then you have to go find your socks. dc


[1] Browse for data on the CLIO system.

[2] Sound and Communications, April, 2008, pages 18-24.

[3] Browse to see Dan Clayton’s work.

[4] The Swan 305 12″ woofer is out of production. The Peerless 830669 seems a good replacement in case of damage (change both, of course).

[5] Browse AkAbak is free for non-commercial use.

Many thanks to Communications and Entertainment, Atlanta, for doing such a great job in reassembling the drivers, preparing and testing the sub properly, and performing a splendid installation. dc

Photos of the building process and the final placement of the sub.