OEM Headlamp Sampling: 7 Items to Check on Your First Golden Sample Before Mass Production

Most brand buyers who contact our OEM headlamp factory already know that the pre-production sample decides whether a program ships on time or quietly dies. The problem is not whether to inspect the sample. The problem is that almost every OEM headlamp golden sample checklist floating around online is a generic inbound-QC list copied from a flashlight category. Headlamps are worn on a human head for hours in rain, sweat, dust, and cold. The acceptance criteria have to reflect that.

Across the last decade of running headlamp programs in our Ningbo facility, I have signed off on hundreds of golden samples and rejected almost as many. This article is the checklist we hand every new brand buyer before they review their first unit, plus the reasoning behind each test so the conversation with the headlamp factory stays technical, not emotional.

1. Beam Pattern and Peak Beam Intensity on an Integrating Sphere

The first item on any honest OEM headlamp golden sample checklist is beam performance, because it is the one specification a buyer cannot eyeball into compliance. A sample that looks bright on a workbench can still fail a real trail walk.

We measure every golden sample on an integrating sphere at 1 meter from the sensor, logging peak candela (cd), peak beam distance (m), and total lumens. We then place the headlamp 2 meters from a lux meter at the hot spot, the mid-ring, and the spill edge, and compare the three values to the agreed optical spec.

• Peak candela tolerance: the measured value must sit within ±8% of the contracted value, because anything wider means a different LED bin or a refocused reflector.
• Spot-to-flood ratio: the central hot spot cannot exceed 60% of total beam energy if the SKU is sold as a flood-dominant model, and cannot fall below 35% if it is sold as a throw model.
• Beam uniformity on a 1 m wall: no dark rings, no concentric hot spots, no yellow rings from a misaligned reflector.

2. Continuous Runtime Test on the Highest Mode

Runtime is the second item, and it is the one that quietly breaks more OEM headlamp programs than any other. The reason is simple: most factories quote runtime at the lowest mode to make the spec sheet look generous, and only a careful buyer asks for the highest-mode figure.

For the golden sample, we run a controlled discharge test: fully charge the cell to its nominal voltage, set the headlamp to the highest steady mode, and log output and current draw every 5 minutes until output drops to 10% of peak. The test room sits at 25°C ±2°C and uses a regulated bench supply if the headlamp accepts direct DC, so we remove cell variability from the picture.

• High-mode runtime tolerance: the measured runtime must sit within ±5% of the contracted value. Anything shorter usually means a weaker cell or a more aggressive drive current in the driver.
• Output curve shape: the curve should be flat for the first 70% of runtime, then ramp down. A sharp step-down at 50% means the thermal protection is too aggressive and the headlamp will feel inconsistent in cold weather.
• Cell voltage at cutoff: should match the protection IC setting within 0.1 V. A higher cutoff starves the cell and wastes capacity; a lower cutoff risks a deep discharge.

If you skip this step, you will discover the runtime shortfall three weeks after mass production has already shipped.

3. IP Rating With a Controlled Water and Dust Test

IP rating is the third item, and it is the most commonly misrepresented specification on a headlamp data sheet. Buyers see IPX4 and assume the headlamp is rainproof; what they actually have is a splash-proof unit that will fail under a heavy downpour or a sweaty trail run.

For a golden sample we run the IP test on a real headlamp, not a marketing claim. The procedure depends on the rating on the spec sheet:

Because headlamps are mounted on the forehead, sweat ingress is just as important as rain. So we also tilt the headlamp 45° during the splash test to mimic the angle at which salt-laden sweat runs across the housing during a trail run.

4. Drop Test From 1.5 m Onto Concrete, Six Faces

A headlamp lives on a climbing helmet, a hard hat, or a sweaty forehead. It will be dropped. The drop test is the fourth item on the checklist because the failure mode is not always obvious — a cracked reflector can survive a drop but ruin the beam pattern in production.

The procedure is simple and unforgiving:

1. Drop the unit six times from 1.5 m onto a concrete floor, once per face (top, bottom, left, right, front, back).
2. Power the headlamp on between each drop and confirm the switch still toggles, the mode memory still works, and the beam still ignites.
3. Open the battery compartment and confirm the latch still locks and the seal still seats.
4. Inspect the LED under a magnifier for cracked phosphor or detached bond wires.

If any face fails, the housing material, wall thickness, or rib layout has to be revised before mass production. A common failure I see on first samples is a battery door that pops open on the third drop — that is a field return waiting to happen.

5. Color Temperature, CRI, and Tint Bin Consistency

Color consistency is the fifth item, and it is the one most buyers forget until they have 500 units in a container and three of them look obviously cooler than the rest. White-balance differences on a headlamp are visible to the human eye far more than on a flashlight, because the headlamp beam lights up the ground right in front of your feet where your eyes are already adapted.

For a golden sample, we measure the correlated color temperature (CCT) and the color rendering index (CRI) on every mode with a Konica Minolta or an AsenseTek spectroradiometer. The acceptance window depends on what the buyer contracted:

• CCT tolerance: ±5% of the contracted CCT value (for example, a 5000K SKU should sit between 4750K and 5250K).
• Tint bin: all measured units must fall inside the same 3-step MacAdam ellipse around the contracted bin. A sample that drifts into the next bin is rejected even if the CCT is technically in spec.
• CRI floor: the measured CRI must hit the contracted value (typically 70, 80, or 90+). For a 90+ SKU, we reject any sample below 88.

Because LED suppliers bin by luminous flux, not by color, two LEDs from the same flux bin can sit on opposite sides of the MacAdam ellipse. So the OEM headlamp factory has to specify the color bin by name (for example, “ANSI 5000K 3-step”) on the bill of materials, not just the flux bin.

6. PCB Soldering and Component Quality Under Magnification

The sixth item is PCB quality, and it is the one that catches the most “looks fine from the outside” samples. We open the headlamp and inspect the LED PCB, the driver PCB, the Type-C port, and the switch under 10× magnification.

• LED solder joints: no cold joints, no voids, no tombstoning. The LED pad should be fully wetted, and the thermal pad should have continuous, even solder coverage.
• Driver IC orientation: every IC must be seated square, with no lifted pins, and the dot indicator on pin 1 must align with the silk screen.
• Type-C port: the port should be flush, the mounting tabs should be soldered on both sides, and the IP gasket should sit evenly around the rim.
• Conformal coating: if the SKU is sold as water-resistant, the PCB should carry a uniform conformal coat. Bare copper on a moisture-exposed board is an early field-failure signal.
• Wire management: no pinched wires at the hinge, no stripped insulation near the battery contacts, no residual solder splashes bridging traces.

7. Cross-Check the Bill of Materials Against the Agreed Specification

The seventh item is BOM verification, and it is the one that protects the buyer most of all. A factory can deliver a beautiful golden sample that quietly substitutes a cheaper LED, a lower-CRI LED, a thinner battery cell, or a less waterproof switch. None of those swaps will show up on the integrating sphere alone.

We tear down the sample with the buyer on a video call, walk through every component, and check it against the agreed BOM:

• LED: bin code on the package matches the contracted bin code.
• Battery cell: manufacturer, model, capacity in mAh, and C-rate all match the BOM line item.
• Driver IC: part number matches; no unmarked or house-numbered substitutes.
• Switch: tactile force and rated cycle count match; the rubber boot durometer sits in the contracted range.
• Strap and buckle: webbing width, hook-and-loop strength, and elastic recovery all match the BOM.
• O-rings and gaskets: durometer, ID, and CS all match; any silicone substitute for a specified fluorocarbon ring is a red flag.

Because the BOM is what survives long after the golden sample is sealed in a cabinet. So the golden sample sheet we file lists every component with its supplier code, and the production-line QC work instruction references those codes line by line.

How Many Sample Units to Ask For, and How to Split Them

The sample count matters as much as the test list, because a single sample cannot answer every question. Most brand buyers ask for 5 to 10 units in the first round. We recommend the following split:

The two dispute samples matter because, six months into mass production, somebody will raise a question about a tolerance, a color, or a switch feel. The dispute samples let both sides reach a verdict from a known reference, instead of arguing about which batch the question refers to.

Documentation That Should Travel With the Golden Sample

The golden sample is meaningless without its paperwork. Before you sign off, confirm the factory delivers the following documents with the unit:

• A signed golden sample sheet listing measured lux at 1 m, peak candela, and runtime on every mode.
• An IP test report dated within 30 days of the sample, with the chamber photos and the post-test teardown photos.
• A drop test report with photos of every face before and after the drop.
• A bill of materials listing every component, supplier code, and contracted specification.
• QC work instructions for the production line, cross-referenced to the BOM line items.
• A traceability sheet that names the LED bin, the driver firmware version, and the battery cell lot used in the sample.

If a factory cannot produce these documents on demand, the golden sample they sent you is a marketing sample, not a reference unit.

What Happens If a Golden Sample Fails?

Failure is not a dead end. It is the cheapest correction point in the entire program. Once mass production tooling is released, every change costs real money and real schedule. At the golden sample stage, a buyer can:

1. Request a revised sample with a documented change list (for example, “replace LED bin, re-tune driver to extend high-mode runtime by 8 minutes”).
2. Approve the original golden sample with a written deviation note (for example, “CRI floor 87 accepted in lieu of 90 for this lot only”), signed by both parties.
3. Reject the project entirely if the failure touches safety, IP, or a regulated specification.

Most of the OEM headlamp programs we run end at option 1: one revised sample round, then sign-off. Programs that skip the golden sample gate usually arrive at option 3 with a container of unsellable units and no leverage left.

Where This Checklist Fits in the OEM Headlamp Sampling Lifecycle

The golden sample sits between two other samples in a typical OEM headlamp project, and confusing the three is the single most expensive mistake a new buyer can make.

• Engineering sample (ES): built from off-the-shelf parts, hand-assembled, often in white housing. Used to validate the basic concept. Failures are expected and free.
• Golden sample (GS): built from production tooling, in the final color, with the final BOM, on the production line. This is the article’s focus. Failures here trigger a second GS round.
• Pre-shipment sample (PSS): pulled from the first 100 units of mass production. Cross-checked against the golden sample. Failures here trigger a production hold, not a sample revision.

Because the GS is the only sample built on production tooling, it is the only sample whose result actually predicts how a 5,000-unit run will behave. Treat it accordingly.

Frequently Asked Questions About OEM Headlamp Golden Sample Approval