Helmholtz Resonant Absorber - page 2

3.1 Construction.

The combination of absorbent performance and bandwidth makes a tube the ideal basis for a resonator cavity. Even modest side-walls will ensure absolute rigidity, and completion of the enclosure requires just two end-caps. The material for the cavity is left to the end-user, but products such as concrete pile forming cardboard tube are ideal, as is underground construction-grade PVC tube - as used in this project (fig2).

The only requirement of the tube is that it be sized according to design requirements. Bearing in mind that the required volume comes from internal measurements - side-wall and end-cap thickness need to be taken into consideration, as does the displacement of the port and any internal bracing. **nb: if tube dimensions appear compromised, remember that adjustments can be made to the port to achieve the desired resonant frequency .

It is critical that the finished enclosure be sealed properly, hence the need for accuracy with end-cap sizing. Time spent measuring the diameter of the tube, and in creating templates (fig3) is time well spent. Initial measurements are best done on paper, and transferred to stiff card or hardboard. Material for the end-caps is again left to the end-user, but MDF is a practical and workable material. Marked using the template, cut out and fine-sanded for a good friction fit, this initial end-cap can be used as a template for a bearing-guided router blade (fig4). Even a modest router in this situation will save considerable time and give a repeatable good quality finish.

Fig3.

Fig4.

There are many types of piping that will substitute as a port for the resonator. For resonators designed to be used at very low frequencies a minimum diameter of 2" should be considered - and be of sufficient side-wall thickness that it does not readily collapse. This project uses 68mm (2.7") PVC downspout piping, which fills the criteria, and is extremely cost effective - sufficiently so that cutting several lengths for tuning purposes (fig5) adds little to the overall project cost.

Fig5.

The cut-out for the port in the end-cap follows pretty much the same requirements for sizing the end-cap - making sure it has a good friction fit. The end-caps can now be sealed into place, using a strong PVA-derived adhesive, or an epoxy resin. At this time the port should remain free, for initial tuning purposes. When tuning is completed, and resonant frequency selected, the chosen port can be sealed into place using the same adhesive as the end-caps. Choice and level of finish is left to the end-user (fig6).

Fig6.

3.2 Tuning.

For a device so function specific, tuning should be irrelevant. Indeed, as far as mapping the room, finding problematic frequencies, and designing the resonator to suit, it is. As ever, though, the devil lies in the detail, and there are performance aspects that are difficult to gauge prior to objective assessment. Two primary examples are placement, and resonator bandwidth (through the design addition of acoustic wadding).

Placement ideally goes hand in hand with final tuning of the resonant frequency. It is often quoted that you should use the room corners for placement. Yet, whilst the intersection of room boundaries at corners guarantees modal action, it is often too simplistic an option. Given the effect that room shape, furnishings, other acoustical treatments, and the audio system have on response, measurement is the only accurate indicator of placement. Using an SPL meter and a series of single frequency test tones, one can begin to gauge the effect of the resonator allied to the response at varying points in the listening room

Despite being presented here as separate issues, the tuning process is a whole - each element having an effect on the other. If there is a central, predominant issue, then it is the bandwidth of the resonator and the level of frictional loss introduced. From the outset the aim of this project was to remove as much of the target mode as possible, i.e. elicit the greatest loss (absorption) from the resonator. Under SPL meter and test tone assessment this was easily realised, yet subsequent music listening tests proved unsatisfactory.

Initially, the resonator quite audibly added to the problem it was designed to remove - by increasing resonance, allowing the target mode to sustain, or ring. The problem with the mode, then, was not primarily of level, but of resonance. Widening the operational bandwidth of the resonator, with a modest amount acoustic wadding inserted into the cavity, reduced its relative absorptive effect, but had a marked impact on the problematical resonance. With hindsight, and the use of acoustical software to measure spectral decay (waterfall plot), the design requirements may well have been different - same end, different means.

As a sidebar, and a precursor to the project measurements, it should be pointed out that the resonator, in fact the room itself, has a response that is entirely dictated by the excitation of resonant frequencies. That a room can support certain modal frequencies does not mean that they are activated by the audio system. The flip side, then, is to consider that modified excitation from the audio system (a change of loudspeakers, or a significant change in output level), or from a home cinema system (which often feature high levels of very low bass), will affect the modal response of the room, and the resultant effect of the Helmholtz resonator (fig7) on it.

Fig7.

4. Measurements.

The project resulted in two identical Helmholtz resonant absorbers, now tuned (after the design addition of acoustic wadding to increase bandwidth) to about 60Hz. The following results are the final sets taken, and those now being used in the target listening room. The first set being those from the primary (hifi) loudspeakers, and the second set from the home cinema subwoofer - the differences in response are marked, and highlight the effect that frequency response and level of low bass (particularly) have on room response.

Bear in mind that the results for the primary hifi system, the one around which this project was based, shows smooth, but quite subtle losses - the primary benefit of decreased resonance is not shown. As with many, if not all passive acoustic treatments, it may be necessary to use additional units to produce the desired response.

4.1 Measurement notes.

Each run of measurements has an initial set and a control set. As the measurements were for comparison purposes, not specific SPL levels, care was taken in the provision of a control set to reduce spurious readings or read errors. The results were entered into an Excel spreadsheet and an average taken between the two sets of measurements. The averaged results are those used in the following graphs.

All measurements were taken using an analogue Radioshack SPL meter, using single frequency test tones.