Phased array hearing loops: how they work and when to use one

A single perimeter loop is the simplest hearing loop there is, and in plenty of rooms it does the job well. In a wide room, though, the field sags in the middle, and turning the driver up just makes the field near the walls too strong. A two-channel phased array solves that with narrow loop elements instead of one big loop.

Key points

• A perimeter loop's field is strongest near the wire and weakest in the middle. Past a certain width, no driver setting keeps the listening area within ±3 dB.
• A phased array uses two sets of narrow elements, Phase A and Phase B, offset by half an element and driven 90° apart, so one channel is strong where the other is weak.
• Narrow elements make the field fade sooner beyond the array, which helps with overspill, and they usually lose less to steel in the floor than one wide loop.
• Arrays take far more wire, more labor and a two-channel driver, and their field changes quickly with the distance to the listeners. Use one when a perimeter loop can't reach the targets, over heavy steel, or when neighboring loops or privacy matter.
• Each channel is one continuous series circuit. The winding direction through each element matters as much as its dimensions.

Why a perimeter loop sags in the middle of a wide room

The field from a loop wire is strongest close to the wire and falls off with distance. At listening height, a seat near a wall gets a strong push from the wire beside it, while a seat in the middle is far from every side. In a small room the difference is small. In a wide room it grows until it no longer fits inside the 6 dB window between −3 dB and +3 dB around the 400 mA/m target.

More current or more turns won't fix this: they lift the middle and the edges together, so the shape stays the same. What matters most is the width, the shorter dimension of the room, because it sets how far the middle seats are from the nearest wire.

Worked example (simplified). A single-turn loop on the floor, no metal, listening height 1.2 m (seated), leaving out a 0.6 m (2 ft) strip just inside the wire. Figures are the spread from the strongest to the weakest point; real designs will differ.

In the 300-seat sanctuary, set the middle of the room to −3 dB and the strongest seats, about 1.3 m (4 ft) in from the wire near the corners, end up around +4.6 dB. Turn it down so they sit at +3 dB, and the middle drops to about −4.6 dB. In this model the limit for rooms of these proportions is roughly 13 to 14 m (43 to 45 ft) wide, and less for long, narrow rooms. Steel, listening height and a ceiling loop all move it.

How two channels and a phase shift fill the gaps

An array covers the listening area with a row of narrow loop elements, each a few meters wide. Within one channel, neighboring elements are wound in opposite directions. Directly above the boundary between two elements, that channel's vertical field drops to almost nothing, leaving a weak line along the array.

The second channel covers those weak lines. Phase B's elements are shifted by half an element width, so the middle of each Phase B element, where its field is strong, sits over a Phase A weak line, and the other way around. Phase B carries the same audio, shifted by 90°, so the two fields never cancel. A telecoil at any seat picks up both, and their combined strength stays far more even across a wide room than one perimeter loop can manage. The 90° shift has to hold across the speech range, which is why arrays need a two-channel driver made for them.

Narrow elements bring two side benefits:

• Less overspill. Neighboring elements carry opposite polarity, so beyond the edge of the array their fields largely cancel, and the field fades over a distance set by the element width instead of the room width. How far it still reaches is measured on site; see hearing loop overspill.
• Usually less trouble with steel. Narrow loops generally lose less of their field to steel in the floor than one room-sized loop, which is why arrays are the usual choice over rebar and steel deck. The loss is still there, worst at high frequencies, so allow for it and set the correction on each channel; see metal loss in hearing loops.

The cost: far more wire, which raises each channel's resistance and inductance, so the driver needs more voltage for each amp. The field also falls away much more quickly with height. At seated height in a wide room, an array often needs less current per channel than one big perimeter loop, because its wires are close to the listeners. Raise the listening height or move the array further away, though, and the current it needs climbs far more steeply than a perimeter loop's, more so the narrower the elements. Model the array at the real listening height.

Standard, uniformity-optimized and low-spill layouts

Arrays come in a few layout families, each a different trade-off.

• Standard. Equal elements in a regular pattern, the two channels offset by half an element. Easy to set out and check; suits most wide rooms. It can leave a slight ripple along the element boundaries.
• Uniformity-optimized. The same equal elements, with the overlap between channels worked out for the room to smooth that ripple. Placement matters more: a few centimeters off on site can undo the gain.
• Low-spill. Element widths and overlaps worked out together and mirrored left to right, with every conductor kept inside the listening area. For rooms beside other looped rooms, or where the audio is private. It needs at least five elements, placed precisely: changing any one of them on site upsets the balance and can bring back spill or a weak line.

In a sanctuary with fixed pews, elements are usually sized and spaced in whole rows, so the wires run in the gaps between pews. Measure the row spacing center to center at the front, middle and back: in older buildings it can drift by a few inches.

When a perimeter loop is enough

Stay with a perimeter loop when all of these hold:

• The model shows the listening area within ±3 dB at the listening height, with metal loss allowed for.
• No looped room sits beside, above or below, and nothing said in the room is confidential.
• The room is narrow enough that the middle seats are not far from the wire, like the meeting room above. Community Hall is close to the limit.

A figure-8 or sectioned loop sits in between: one channel, sections wound in opposite directions, with the weak lines planned under aisles.

Check routes early: an array needs wire across the listening area, in the floor or the ceiling. The hearing loop design method shows where the layout choice fits in the wider process.

Drawing and installing an array

One series circuit per channel. Each channel runs as one continuous wire path through all of its elements, never as separate loops wired in parallel. The direction the wire travels round each element sets its polarity, so the winding arrows on the drawing matter as much as the dimensions.

Crossings are not junctions. Phase A and Phase B cross many times, and a channel can cross itself. Nothing connects at a crossing. Where copper tapes cross, put a layer of insulating tape between them. Round wire is already insulated, so protect it: leave room at crossings in a saw cut or chase so no wire is pinched or nicked, and mark the crossings on the floor before cutting.

Set out, label and feed. Measure element positions from a fixed datum, such as the front edge of the seating. Keep widths and overlaps exactly as drawn; small shifts show up later as stripes in the field. Label Phase A and Phase B at both ends and at the driver. Each channel gets its own tightly paired feeder.

Test before covering. With the driver disconnected, record continuity, resistance and insulation for each channel, including the insulation between Phase A and Phase B, which finds a damaged crossing. Photograph the routes. Keep a compatible cut-wire alarm on the circuits while the flooring goes down.

Commission both channels together. Follow the driver maker's set-up procedure. Check each channel alone first; its weak lines are expected. Then run both with the phase shift on and measure across the listening area at listening height, including over the element boundaries. The target applies to the combined field, so set the final level with both channels running: setting each channel on its own to the full target can leave the combined field too strong. If stripes remain, check the phase setting, channel levels and as-built winding against the drawing before reversing sections or raising the current. Installed performance is confirmed by commissioning.

Hearing Loop Designer draws arrays the way they are installed: each channel as a series-wound wire path with its polarity shown, Phase A and Phase B on their own sheets, and crossings drawn as gaps so an underpass isn't mistaken for a break. See how the software handles phased arrays.

Common questions

Can I run a phased array from a stereo amplifier?

No. Both channels need the same audio with a 90° phase difference across the speech range, and a loop needs a driver built to deliver current into a loop. With no phase shift, the two fields add in some places and cancel in others. Use a two-channel loop driver made for phased arrays, or a combination of drivers and phase-shift unit that the equipment maker specifies for array use.

Will a phased array keep the signal out of the room next door?

It helps: the field fades much sooner beyond an array than beyond a big perimeter loop. Some field always reaches past it. If the next room matters, agree what is acceptable with the client and measure there at commissioning.

Can a phased array go in the ceiling?

It can, but the elements are then further from the listeners, and a narrow-element field falls away quickly with distance. Expect wider elements than in a floor array and still much more current per channel, so model it at the real mounting and listening heights. The field goes up as well as down, so check the floor above too.