The Sound of Surprise (the loudspeaker/stand interface) Page 3

The procedure for the experiment was as follows:

Each speaker was placed on an 18" Celestion Si stand, this a single-pillar steel design with the pillar filled with a mixture of lead shot and sand. The stand was spiked to the wooden floor beneath the rug in Tom Norton's office, which doubles as our lab. The interface between the base of the speaker and the top plate of the stand consisted of the following accessories and materials: three "Audio Selection" steel-pointed brass Tone Cones from German Acoustics pointing upward (footnote 5); three German Acoustics soft isolation feet, which appear to have a Sorbothane insert; three Mod Squad Sorbothane feet; three large Monster Cable Footers, which are made of Navcom; three AudioPrism Iso-Bearings, these spherical, nonreactive balls sitting in plastic cups; a ¼"-thick sheet of Navcom; a ½"-thick sheet of polyurethane foam (sneaked from the packaging for a reviewed component that Tom Norton was returning to its manufacturer); three small pads of Blu-Tack; and none of the above.

In every case other than the Navcom and polyurethane foam sheets, which covered the entire stand top plates, I tried to ensure that each of the three supports were in the same places with respect to the speaker's base—two were sited at the front corners of the top plate, where most of the speaker's mass is concentrated, one at the center rear—so that the cabinet's vibrational behavior would be as close to identical in all cases.

Though none of the isolation feet are specifically sold as being appropriate for use between a loudspeaker and its stand, I thought it would be interesting to see what their effect would be. Blu-Tack is a nonreactive, resistive-damping mastic material made by Bostik and sold in the UK for sticking photographs, pinups, and posters to walls, etc. Though it is distributed to US dealers by Artech Electronics, it appears to be identical to the E.Z.Tak mastic material found in most drugstores. (I bought a lifetime's supply for $1.59 from the Santa Fe Woolworth's.)

In a second series of tests, the Celestion stand was replaced with the Target stand intended to be used with the ProAc (the Target is very massive even with its uprights not filled with sand), and an inexpensive 24" wooden stand with two 1x4 uprights.

The accelerometer used, described in June 1992 (p.206), is fabricated from a 4" by 1" strip of the piezoelectric polymer PVDF (polyvinylidene fluoride). This is securely taped to the surface to be analyzed; the strip is attached to a small plastic block, from which emerges a short pigtail of coaxial cable; this in turn feeds a 20dB-gain, op-amp-based buffer amplifier. Each speaker was driven with a 7.23V RMS, 2kHz-bandwidth, noise-like MLS signal by a Bryston 4B amplifier; the accelerometer was left in the same places on each speaker's top and side panels for all the testing; and DRA Labs' MLSSA system was used to calculate each panel's impulse response, from which the cumulative spectral decay plot was derived. This "waterfall" plot shows the amplitude and frequency of every resonance present in the panel being investigated. Because the drive signal was kept constant and the accelerometer kept in the same place, the results for each speaker are directly comparable for each support.

Table 1 shows the results for the AE1's side panel with the speaker sitting on the Celestion stand. The frequency and amplitude of each resonant mode are noted, the latter being given both in absolute terms and referenced to the level measured with the brass, pointed, upward-pointing Tone Cones, which in theory should impose the minimum damping on the cabinet behavior. Where the frequency of the resonance has been changed, this has been noted in parentheses beneath the dB difference.

Table 1: Acoustic Energy AEl, side-panel resonant modes

InterfaceResonant mode amplitude in dB
Resonant mode Freq in Hz6315634448465982014061648
German Acoustics Cones-9.3-10.9-18.75+2.86-16.25-8.8-18.75-17.1
Celestion Stand0.
Mod Squad Soft Feet-9.0-*-20.2+2.92-17.0-9.4-18.2-19.9
Celestlon Stand+0.3<-21*-1.45+0.06-0.75-0.6+0.55-2.8
German Acoustics Feet-8.8-*-12.3+3.27-16.9-9.3-18.1-18.2
Celestion Stand+0.5<-21*+6.45
Monster Cable Footers-9.2-*-21.9+2.42-16.3-8.7-17.4-18.0
Celestion Stand+0.1<-21*-3.15
AudioPrism Iso-Bearings-9.3-*-12.6+3.63-16.5-9.0-18.4-17.9
Celestion Stand0.0<-21*+6.15
Navcom sheet-9.2-*-17.75+3.82-17.1-11.4-17.4-23.5
Celestion Stand+0.1<-21*+1.0
Blu-Tack pads-8.9-*-29.3-3.5-18.5-8.8-23.0-25.9
Celestion Stand+0.4<-21*-10.55-6.35-2.250.0-4.25-8.8
Celestion Stand-0.6-3.2

* Resonant mode has been suppressed below the waterfall plot's -30dB floor

Footnote 5: With some loudspeakers, Celestion's SL600Si and '700, for example, the use of upturned Tiptoes or metal cones is not recommended, as these will penetrate the speaker's thin aluminum skin.