Metal Detector Verification Tests (3 mm / 4 mm Standards)

This topic is part of the SG Systems Global foreign-material control & inspection technology glossary.

Updated November 2025 • Metal detection, 3 mm / 4 mm test standards, ferrous / non-ferrous / stainless, CCP verification, GFSI & retailer codes, MES integration • Operations, FSQA, Engineering, Regulatory, Brand Protection

Metal detector verification tests (3 mm / 4 mm standards) are the routine checks that prove a metal detector is actually capable of finding the metal sizes you say it can—typically 3 mm and 4 mm “spheres” of ferrous, non-ferrous and stainless steel. They are not optional “nice to haves”; they are the monitored proof that a critical foreign-material control step is working, every shift, every product, every line. Verification testing means running calibrated test pieces through the detector in defined positions, at actual line speeds, and confirming that the system reliably rejects them and alarms. If you skip these tests, fudge them, or treat them as paperwork exercises, you may as well not have a detector at all.

“If your 3 mm and 4 mm test pieces don’t reliably trigger the metal detector, it will not magically detect a random 3 mm blade fragment at full speed. The test is the truth, not the brochure.”

1) Why 3 mm and 4 mm Matter

Most food, meat, poultry and bakery codes of practice talk in millimetres. The usual sizes (for retail products) are:

• 3 mm spheres – typical challenge for ferrous and non-ferrous metals in many chilled/fresh products.
• 4 mm spheres – typical for stainless steel in more difficult products or where expectations are less aggressive.

The exact sizes depend on product, aperture size, retailer and risk assessment, but 3 mm and 4 mm standards are the common language. Customers will ask “What is your 3 mm / 4 mm capability?” Auditors will ask “Show me your 3 mm / 4 mm verification records.” Legal teams will ask “What could we reasonably detect?” Your answers are only as good as your verification tests. The numbers on the test pieces define the minimum size you are effectively promising to the world.

2) What Are 3 mm / 4 mm Standards?

“Standards” in this context are calibrated test pieces supplied by recognized vendors. Typically you’ll see:

• Ferrous 3 mm and 4 mm spherical inserts (usually iron or low-carbon steel).
• Non-ferrous 3 mm and 4 mm spheres (often brass or similar).
• Stainless steel 3 mm and 4 mm spheres (often 316 or 304 grade).

These are encapsulated in plastic cards, rods or balls that can be run through the detector safely. They are not toys; they are calibrated measurement devices. Their size, material and certification traceability must be documented and maintained. If your test pieces are chipped, worn, unknown-grade or “borrowed” from another site with no paperwork, your verification is already compromised before you even run a test.

3) Detection Fundamentals: Product Effect & Aperture Size

Before you define 3 mm / 4 mm test regimes, you need to understand what your metal detector is fighting against:

• Product effect – high salt, moisture, temperature or bulk density makes the product itself “look metallic” to the detector. Fresh bread, saline meats and hot products all generate signals that reduce sensitivity.
• Aperture size – the larger the aperture, the harder it is to detect small metals. A 3 mm standard may be realistic on a small snack aperture but not on a huge bulk bag line.
• Orientation and speed – in conveyor systems, the orientation of a metal fragment and belt speed affect signal strength. Test pieces must be run in a way that represents worst-case orientation at full speed.

3 mm / 4 mm standards are not “one size fits all.” A line with high product effect or a very large aperture might reasonably target 4 mm stainless and 3 mm ferrous; a low product effect line with smaller aperture might run 2 mm / 3 mm standards. Validation and risk assessment drive those decisions; your verification tests must align with them, not with what someone thought sounded impressive in a spec document years ago.

4) Product-Specific Test Standards

Most sites standardise on a small set of detection goals by product family:

• Raw / chilled meat, poultry, seafood – 3 mm ferrous, 4 mm non-ferrous, 4 mm stainless (sometimes 3 mm stainless in high-end programs).
• Baked goods – 2–3 mm ferrous and non-ferrous, 3–4 mm stainless, depending on aperture and product effect.
• Dry snacks / powders – lower product effect allows 1.5–2 mm targets in some cases.
• Bulk / catering packs – may accept larger targets, but still often in the 3–5 mm range.

These targets should be documented in the HACCP plan, foreign-material program and customer specs. Metal detector verification tests then use the corresponding 3 mm / 4 mm standards for that product family. Using the wrong standard (e.g., a 6 mm ferrous ball on a line that claims 3 mm detection) is a silent downgrade—and auditors will pick up on that mismatch very quickly if they understand the technology at all.

5) Test Positions: Where to Place 3 mm / 4 mm Standards

Detection performance varies across the aperture. Good practice is to test:

• Leading edge of the product pack or flow.
• Center of the pack / belt.
• Trailing edge of the pack / flow.
• Left and right sides of the belt (for wide conveyors).
• Top and bottom (for tall packs or bulk flows, where practical).

For individually wrapped packs, typical verification is three passes per metal type: front, middle, back within the pack, placed randomly across the belt width. For bulk flows, test cards are buried at varying depths and positions in a product matrix. Your SOP should define the exact positions per line; your verification records should prove they were followed. Testing only in the “sweet spot” center of the aperture gives you a false sense of security; most missed contaminants live in the edge and corner cases you never challenged.

6) Frequency: How Often to Run 3 mm / 4 mm Tests

Verification frequency is part of risk assessment and HACCP design. Typical patterns in GFSI / retailer environments:

• At start of each shift – before running product.
• At product changeovers – whenever set-up or product effect changes significantly.
• After any downtime – especially mechanical/electrical works, cleaning, software changes or storms/power dips.
• At end of shift or lot – to confirm performance at the end of production and bound any potential deviation window.

Some retailers also specify maximum time intervals (e.g., every 2–4 hours). FSQA may require more frequent tests on high-risk lines or where complaint data justifies it. The heavier the reliance on metal detection as a CCP, the tighter the verification schedule should be. “We test once a day because it interrupts production” is not a position you want to explain after a metal-injury complaint reaches a regulator or major customer.

7) Procedure: How to Perform a Verification Test Correctly

A competent metal detector verification test for 3 mm / 4 mm standards looks like this:

• Confirm the detector is in normal production mode (correct product recipe, sensitivity, belt speed, reject timing).
• Prepare test samples: the 3 mm / 4 mm ferrous, non-ferrous and stainless pieces in their certified carriers.
• Place each test sample in or under product (where required by spec) at defined positions, or run it “naked” if testing the unit without product is acceptable and documented.
• Run each test piece through the detector at full speed, at least once per required position and metal type.
• Confirm that the detector alarms and the reject system removes the test product from the line, into a secure reject receptacle.
• Record each test in MES/eBR or controlled logs: date/time, line, product, test piece type/size, positions, results, operator, verifier.

Any failure (missed detection, wrong reject operation, inconsistent alarm) should be treated as a serious event: stop use, quarantine product since the last successful check, investigate, correct and recheck. “We re-ran the test until it passed and moved on” is a confession, not a corrective action. Regulators and serious auditors know the difference.

8) Integration with HACCP: CCP or OPRP?

Metal detection sits somewhere on the HACCP / risk-management spectrum:

• Many plants classify it as a HACCP CCP for foreign materials in RTE or high-risk products.
• Others treat it as an OPRP (operational prerequisite) under a broader foreign-material control program, especially when upstream controls are strong and products are further processed.

Either way, 3 mm / 4 mm verification tests are your monitoring / verification procedures. HACCP documentation should include:

• Hazard description (metal fragments above size X).
• Critical limit or performance spec (e.g., “Detector must reliably detect 3 mm ferrous and 4 mm stainless test pieces in defined positions”).
• Verification plan (per-shift tests, test types, results criteria).
• Corrective action plan (what happens when a test fails, including product hold and disposition).

If metal detection is labelled a CCP but verification is sporadic, loosely documented or run with the wrong test sizes, your HACCP plan is not credible. Auditors will quickly spot the gap between your paperwork and reality; in serious cases, regulators will treat this as a failure to control the hazard you claim to control.

9) Recording Results: From Clipboards to MES

Verification tests mean nothing if you can’t prove they happened. Minimal documentation should include:

• Date and time of each test.
• Line and detector ID (many plants have multiple detectors per line).
• Product and metal types tested (3 mm ferrous, 3 mm non-ferrous, 4 mm stainless, etc.).
• Positions tested (center / leading / trailing / left / right, or defined codes).
• Result for each test (“P” for pass, “F” for fail) and any corrective action taken.
• Operator and supervisor/QA signatures or electronic IDs.

Modern plants embed this in MES / eBR, often tying verification tasks to line start-up and changeover workflows. The system won’t allow a batch to begin until verification tasks are completed; failed tests trigger automatic product holds and alerts. This removes the temptation to “catch up” verification logs later, which is a polite term for falsification. Clipboards stuck to the side of the detector can work at small scale, but they are fragile, easy to fake and hard to query across sites; digital evidence is far harder to argue with in an audit or investigation.

10) False Rejects, Product Effect & Tuning

Verification tests also tell you about stability and false rejects. If a detector starts to struggle with 4 mm stainless at normal settings, or false rejects spike, that may indicate:

• Product effect significant shift (temperature change, formulation, salt content).
• Hardware drift (coil issues, electronic noise, mechanical vibration).
• Poor product presentation (overlapping products, high belt loading, unstable packs).
• Incorrect product recipe selection on the HMI.

Engineering and FSQA should use verification data to adjust and re-tune: selecting the right product mode, re-balancing coils, adjusting belt loading and reviewing handling upstream. The target is a detector that consistently passes 3 mm / 4 mm tests with a stable, low false-reject rate. If you tune away false rejects without re-verifying detection of the target sizes, you’re trading safety for convenience—often silently, until the complaint data catches up with you.

11) Multiple Detectors & Program Standardisation

Larger plants and groups will have multiple detectors: raw intake, in-process, post-pack, case-level. Metal detector verification tests need consistent design across this landscape:

• Standard test kits per site or group, with controlled 3 mm / 4 mm pieces and common documentation.
• Line-specific test plans – each detector has its own frequencies, positions and target sizes, driven by risk assessment and spec requirements.
• Corporate governance – central FSQA/Engineering define minimum standards; sites can add tests but not subtract.
• Cross-site trending – comparing verification performance and failure rates across plants to identify systemic issues and best practices.

This stops each facility from inventing its own metal detection “religion” and lets you speak coherently to global customers: “Across all sites, we validate and verify 3 mm ferrous and 4 mm stainless on finished RTE sausages in these positions at these frequencies.” That’s the kind of answer retailer technical teams like to hear—and the kind of discipline that makes group-level metal-complaint KPIs manageable rather than a constant firefight.

12) CAPA When Verification Fails

Verification failures are not paperwork glitches; they are telling you that the detector is not achieving the performance you rely on. A credible CAPA process should treat them as serious events:

• Immediate stop / hold – block product from passing through the affected detector until evaluation is complete.
• Define exposure window – from last successful test to failure, identify all lots that passed through the detector.
• Quarantine product in the exposure window; perform risk-based evaluation (including potential re-screening, visual checks or destruction, depending on severity and customer/regulator expectation).
• Root cause analysis – hardware fault, setup error, environmental change, product change, test error.
• Corrective action – fix root cause; adjust setup; retrain; revalidate if necessary.
• Verification of effectiveness – increased frequency of tests, complaint trend monitoring, engineering review.

“We just recalibrated and moved on” is not an acceptable CAPA story when a 3 mm / 4 mm standard failed. You need a clear trail showing that you identified potentially affected product and took rational, documented decisions. That trail becomes critical if a subsequent complaint arises for that timeframe and lawyers or regulators start pulling your records apart line by line.

13) Common Failure Modes & Red Flags

Patterns that scream “metal detector verification is a sham” to auditors:

• Logs that show perfect passes for months with no test failures, on old, high-challenge settings—statistically unlikely in the real world.
• Single test piece used (e.g., only 4 mm ferrous) on a line that claims 3 mm stainless capability.
• Operators running tests with the detector in “test mode” or at lower sensitivity than production mode.
• Test pieces run without product when specs require them to be embedded in product for realistic testing.
• Incomplete logs, missing times, copied entries or signatures written for entire days at once.
• Detectors that can be bypassed or left in “monitor only” while documentation pretends they are full CCPs.

These are not minor technicalities; they indicate a cultural problem. Fixing it means aligning incentives (operators not punished for truthful failures), simplifying test procedures, automating logging and making it clear that management cares more about real control than about tidy but fictional paperwork. If senior leadership only ever asks “did you pass all metal tests last month?” and never “what did we learn from failures?”, you know which story staff will write for them.

14) FAQ

Q1. Why do we test 3 mm and 4 mm and not just one size?
Because different metals are harder to detect. Ferrous is usually easiest, then non-ferrous, then stainless is the worst case. Many specs define 3 mm ferrous and 4 mm stainless as realistic targets on a given line. Testing both ensures you are covering the easiest and the hardest common contaminants for that setup.

Q2. Do test pieces have to be run inside product?
It depends on your risk assessment and customer/spec requirements. Testing “naked” pieces shows the best-case detection ability; embedding them in or under product shows performance under real conditions, including product effect. Many retailer codes require product-embedded tests, at least at start-up and changeovers. If you test only in air but claim performance in product, you are likely overstating reality.

Q3. Can we standardise on 4 mm only to reduce hassle?
You can, but you are lowering your control level. If customer or internal specs demand 3 mm ferrous or 3 mm stainless detection and you only verify at 4 mm, you are out of compliance with your own program. Any change in test sizes should go through formal risk assessment, spec review and HACCP reassessment—not quietly changed on the floor to make life easier.

Q4. How long do verification records need to be kept?
At least as long as your product shelf life plus any regulatory or customer-specified retention period—often several years. Since verification records may be critical in defending against injury claims or regulatory challenges, many companies align retention with recall traceability requirements and legal holds, not just minimum local rules.

Q5. Do we need separate 3 mm / 4 mm standards for every line?
You need adequate test coverage for every metal detector. Many groups standardise on a common set of test pieces per site (shared kits), but each detector has its own defined test frequencies and positions. What you cannot do is test one detector and assume identical performance on others without their own verification data.

Q6. Are metal detectors enough to control all foreign materials?
No. Metal detectors address metallic hazards within certain size and orientation limits. They do not see bone, many plastics, rubber, wood or low-density stones. A robust foreign-material program also includes upstream equipment control, visual inspection, X-ray systems where justified, good housekeeping, supplier controls and complaint trend analysis. Metal detector verification is one important piece, not the entire puzzle.

Related Reading
• Foreign-Material & Inspection: X-Ray Bone Fragment Detection Validation | Mass Balance | Mock Recall Performance
• HACCP & FSQA: HACCP | Deviation / Nonconformance (NC) | CAPA
• Traceability & Systems: GS1-128 Lot Transfer Scanning | MES | End-to-End Lot Genealogy