Machinery Wiki

How a Metal Detector Works

By Global Machinery Hub Editorial Team · Published August 30, 2026 · Updated November 10, 2026 · 6 min read

The core idea of this page: a metal detector creates an electromagnetic field and watches for the disturbance a conductive object causes in it. The disturbance depends on the object's size, its conductivity and its magnetic permeability — which is why a large piece of stainless steel can be harder to find than a small piece of iron, and why wet or salty product can hide metal entirely.

The three coil system

Transmit and receive

A search head has a transmit coil that radiates a field and two receive coils arranged so that, in the absence of metal, their signals cancel. Any conductive object unbalances them. The electronics measure that imbalance and decide whether it exceeds the reject threshold.

Balance and phase

The "phase" of the imbalance tells the detector something about the object. Product effect and metal have different phase angles, which is how a modern detector can suppress the product's signal while keeping sensitivity to metal — the basis of product-effect compensation and multi-frequency detection.

What limits detection

Object size versus aperture

Aperture is the opening the product passes through. A small aperture concentrates the field, so the same detector is more sensitive on small packs than on large ones. This is why the specified sensitivity must always be quoted with the aperture size.

Product effect

Wet, salty, hot or conductive products generate their own signal, which must be filtered out. The stronger the product effect, the more sensitivity is lost. Warm product straight from a cooker is the classic case: it detects well when cold and poorly when hot.

Orientation

A long, thin contaminant is detected best when it crosses the field in a particular orientation. A wire lying parallel to the belt can pass undetected, which is why specifications often state a worst-case orientation.

Packaging material

Metallised film and foil lids disturb the field so strongly that detection becomes impractical. Aluminium trays and foil-lidded packs are where metal detection stops being an option. The same applies to packs with metallised labels or a metal closure: the closure, not the product, becomes the dominant signal. For those packs, ferrous-in-foil detectors or X-ray are the realistic alternatives. Ferrous-in-foil units find iron and some magnetic stainless but not aluminium or non-ferrous metal, so they are a partial answer rather than a substitute: the hazard analysis has to accept that only ferrous contamination is covered, and that limitation has to be written into the specification rather than discovered during an audit. Where the specification requires non-ferrous detection in foil packs, X-ray is the only technology that meets it.

Types of system

Balanced coil systems are the standard conveyor units. Ferrous-in-foil detectors use a strong permanent magnet and can find ferrous contamination in foil packs, at reduced sensitivity. Gravity fall detectors inspect free-flowing product as it falls through a pipe, used for powders and granules.

Multi-frequency and product-effect compensation

Single-frequency detectors struggle when the product's signal overlaps the metal's. Multi-frequency machines measure at two or more frequencies and combine the results, which lets them separate product from metal more effectively. The practical benefit is that a wet or salty product can be inspected at a sensitivity that a single-frequency machine could not reach — but the improvement is product-specific and must be proven with challenge packs, not assumed from a brochure.

Aperture and sensitivity

Sensitivity is quoted as the smallest detectable sphere of a given material at the centre of a given aperture. Larger apertures reduce sensitivity, so a detector on a wide conveyor for large cases is inherently less sensitive than the same electronics on a narrow pipe. When comparing machines, compare at the same aperture and the same product, and ask for the stainless figure separately from the ferrous figure.

Testing and challenge packs

Challenge packs should include ferrous, non-ferrous and stainless samples at the specified size, and they should be passed in the worst-case orientation — typically with a long thin sample lying parallel to the belt. Testing only with a sphere passed through the centre overstates performance. The test frequency belongs in the HACCP plan, and the results belong in the batch record.

Failure modes

Vibration and nearby moving metal cause false rejects; a loose conveyor belt or a metal support too close to the head changes the field; product-effect settings left over from a different recipe suppress real signals; and a reject device that is not interlocked lets a contaminated pack continue down the line. None of these show up as an alarm — only challenge testing finds them.

Installation environment

The search head needs a stable, vibration-free conveyor, no moving metal within the exclusion zone, and no nearby motors or inverters radiating interference. A metal frame leg welded too close to the head, or a stainless conveyor a few centimetres from the aperture, will create false rejects that no amount of tuning resolves. Installation is part of the instrument's performance.

Verification checklist

Confirm: sensitivity specified with aperture, product and orientation; challenge packs including ferrous, non-ferrous and stainless in the worst-case orientation; tests run at a defined frequency and logged; reject confirmation interlocked; product-effect settings recorded per recipe; and no metal structures or moving metal within the manufacturer's exclusion zone.

Source: Metal detector — Wikipedia

"A metal detector is an instrument that detects the nearby presence of metal."

Frequently asked questions

Why does sensitivity drop when the product is warm?

Warm product has a stronger product effect, which the detector must filter out, leaving less sensitivity for metal. Test at production temperature, not on cold samples.

Can a metal detector find stainless steel?

Yes, but stainless is less conductive and less magnetic than iron, so the detectable size is larger. Specifications should state the stainless size separately.

Why do I get false rejects with no metal present?

Vibration, nearby metal, a loose belt or a product-effect setting that no longer matches the recipe. Fix the environment before adjusting sensitivity.

Related reading: metal detection vs X-ray, checkweighing vs metal detection, product inspection systems, and the inspection systems category.