In the global plush toy trade, there is zero tolerance for metal contamination. If a child squeezes a stuffed animal and an embedded sewing needle tip pierces their hand, the consequences are immediate: emergency surgery, catastrophic brand liability, million-dollar lawsuits, and immediate border import bans.
Despite this risk, many buyers assume needle detection is as simple as “running cartons through a metal scanner at the end of the line.”
In reality, high-sensitivity metal detection requires complex electromagnetic field calibration, strict unbroken needle management on the sewing floor, and careful BOM component engineering to prevent false rejection alarms.
Here is how export-grade plush factories eliminate metal contamination risks.
1. Electromagnetic Detection Physics: How Conveyor Systems Work
Plush toys are passed through an electromagnetic aperture containing a central transmitting coil and two outer receiving coils wired in opposition.
[Dual-Coil Electromagnetic Induction Sensor Array]
[Receiving Coil 1] <--- High Frequency AC Signal
==========================
| APERTURE | <--- 25 m/min Conveyor Belt
========================== Plush moves through center
[Transmitter Coil] <--- Generates Uniform Balanced EM Field
==========================
| APERTURE |
==========================
[Receiving Coil 2] <--- Identical Balanced Opposing Signal
When an electrically neutral object (cotton, polyester, plastic pellets) passes through the aperture, the magnetic field remains perfectly balanced.
When a ferrous particle (such as a broken steel needle tip) enters the field:
- It distorts the magnetic flux lines.
- The voltage balance between the two receiving coils shifts by microvolts.
- The DSP (Digital Signal Processor) triggers an audible alarm, stops the conveyor belt, and reverses the belt to eject the contaminated doll.
Typical Detection Capability by Aperture Height
The figures below are nominal values quoted for production needle detectors, measured on a spherical test piece in a clean aperture. They are a purchasing guide, not a guarantee: real performance changes with the metal, its shape and orientation, and where in the aperture it passes. Confirm capability with a certified test card on your own machine at commissioning and at every calibration check.
| Aperture Height | Nominal Ferrous (Fe) Sensitivity | Nominal Non-Ferrous (Cu / Al) | Nominal SUS 304 Stainless Steel | Primary Inspection Stage |
|---|---|---|---|---|
| 80 mm (Low Profile) | 0.6 mm Fe | 1.0 mm | 1.2 mm | Small keychains, plush accessories |
| 120 mm (Standard) | 0.8 mm Fe | 1.2 mm | 1.5 mm | Standard plush (15cm to 30cm) |
| 180 mm (High Aperture) | 1.0 mm Fe | 1.5 mm | 2.0 mm | Medium/large plush dolls |
| 300 mm (Master Carton) | 2.0 mm Fe | 3.0 mm | 3.5 mm | Bulk master carton pre-container audit |
2. The 9-Point Calibration Grid Protocol
Electromagnetic sensitivity is not uniform throughout the aperture: it is strongest near the perimeter walls and weakest in the geometric center (“the dead zone”).
Because sensitivity is not uniform, a single pass or fail check at one spot proves little: the test piece has to be checked across the whole aperture. Production lines commonly run a 9-point grid check every 2 to 4 hours, at the start of each shift, and after any change to the machine, the product or the belt speed. The exact interval belongs in your own QC plan, not in a fixed rule.
[9-Point Calibration Aperture Matrix]
+-----------------------------------------------+
| [Point 1: Top-Left] [Point 2: Top-Center] [Point 3: Top-Right] |
| |
| [Point 4: Mid-Left] [Point 5: DEAD CENTER] [Point 6: Mid-Right] |
| |
| [Point 7: Low-Left] [Point 8: Low-Center] [Point 9: Low-Right] |
+-----------------------------------------------+
<--- Conveyor Belt Direction (25–30 meters / min) --->
Protocol Steps:
- Calibrated test piece: a certified acrylic test card with an embedded 0.8mm Fe sphere for a 120mm aperture, or the equivalent certified test piece for the aperture and product being run, traceable to a metrology standard.
- Sequential testing: The test block is placed on the belt at points 1 through 9. The sensor must trigger an alarm and belt stop on all 9 positions.
- Log record: The date, time, technician signature and machine sensitivity setting are recorded in a physical QA binder.
- Failure trigger: If any position fails to trigger, every unit produced since the last successful calibration check is impounded and re-screened 100%.
3. The Broken Needle Control Protocol on the Sewing Floor
Prevention on the sewing line is just as critical as detection at packaging.
When a sewing needle strikes a plastic zipper tooth, thick backing washer, or metal rivet, it can shatter into 2, 3, or more pieces. Without strict floor protocols, the operator simply puts in a new needle and resumes sewing—leaving the needle tip lodged inside the plush.
[Flechazo Broken Needle Reconciliation Workflow]
[Needle Breaks at Workstation]
|
v
[Workstation Stops Immediately]
|
v
[Supervisor Brings Magnetic Sweep + Sticky Logsheet]
|
v
[Recover All Needle Fragments & Reconstruct 100% Geometry]
|
+-------+-------+
| |
[Pass: 100% Met] [Fail: Missing Fragment]
| |
v v
[Issue New Needle] [Quarantine Entire Batch & Run Industrial De-needle Sweep]
The 100% Puzzle Rule:
- Every broken needle must be pieced together like a jigsaw puzzle on sticky magnetic paper in the official Broken Needle Log.
- Only when the supervisor confirms that the shank, blade, eye, and tip form a complete 100% needle silhouette is a replacement needle signed out from the locked tool crib.
- If a fragment cannot be found after searching the workstation, the machine bed and the piece in progress, the unfinished plush doll currently on the machine—plus all adjacent pieces in the sewing basket—are immediately quarantined and condemned to destruction.
4. Root Causes of False Rejection Alarms & BOM Solutions
One of the greatest headaches for toy manufacturers is high false positive rates: the machine alarms constantly even though there is no broken needle inside. High false alarm rates cause production bottlenecks and tempt lazy operators to dial down sensor sensitivity.
[False Alarm Frequency by Component Material]
Component Material
Nickel-plated Steel Rivets | ■■■■■■■■■■■■■■■■■■■■■■■■ Highest - frequent false rejections
Metallic Lurex Yarn | ■■■■■■■■■■■ Conductive core interference
Standard Metal Zippers | ■■■■■■■■ Residual magnetism in stops and sliders
Demagnetized H65 Brass | ■ Occasional minor noise
POM Plastic Hardware | ■ Near-zero rejections
+---------------------------------------------------
Relative ranking of alarm frequency - verify against your own BOM and machine
Common Culprits and Engineering Fixes:
- Metallic Embroidery Threads (Lurex):
- Problem: Metallic glitter yarns contain aluminum or polyester coated in conductive trace elements that trigger electromagnetic eddy currents.
- Fix: Specify non-ferrous certified lurex supported by a metal-content declaration from the yarn supplier, or route the embroidered doll parts through a high-frequency Tunnel Demagnetizer before final needle screening.
- Zippers & Snap Buttons:
- Problem: Standard metal zippers use ferrous steel stops and sliders.
- Fix: Mandate H65 lead-free copper brass or POM (polyoxymethylene) resin zippers. If metal pulls are aesthetically necessary, use certified non-ferrous alloy hardware with a supplier metal declaration.
- Residual Static Electricity:
- Problem: High-pile faux furs moving across rubber conveyor belts generate intense electrostatic fields that can cause capacitive sensor spikes.
- Fix: Install ionized air blowers (anti-static bars) at the conveyor entrance to neutralize surface charge prior to aperture entry.
Zero-defect toy safety is not an accident. It is the result of continuous sensor calibration, strict workshop tool discipline, and clean BOM materials specification.