Why Two Batches of the Same LED Strip Look Different

When batches meet on the same wall

Two batches of strip light carrying the same color temperature label can look different when installed on the same wall. Continuous runs expose it worst: at the junction of the two batches, the eye picks up a visible line of color.

The fixes that follow are all costly. Reels are re-sorted and rewound. The transition is shifted to a corner or a channel joint where the eye compares less. Acceptance is quietly relaxed, because human vision tolerates more mismatch than a drawing does. The cause is rarely one defective chip. It is unmanaged batch variation.

What the color temperature label actually covers

A nominal CCT such as 3000K is the center of a permitted range, not a single color. Under the ANSI C78.377 framework, each nominal CCT occupies a quadrangle on the chromaticity diagram roughly the size of a 7-step MacAdam ellipse. A color point anywhere inside that quadrangle earns the same nominal label.

MacAdam steps measure visible color difference. Around 3000K, one step spans roughly plus or minus 30K. Past about two steps, around 60K of separation, most observers start to see the difference when two sources sit side by side.

Position inside the range matters as much as the label. Two sources can sit on opposite sides of the black-body line and look almost identical, while another pairing inside the same nominal range can differ about four times more, one leaning green, the other pink. Most commodity LEDs reach the market sorted to a 4 to 7 step tolerance, which is exactly the range where side-by-side differences become visible. This is also why a measured color temperature is a specific number that lands near the nominal point, rarely on a round thousand.

Phosphor deposition, wafers and the bins inside bins

White LEDs are made by converting blue chip light through phosphor. Wafer to wafer, phosphor deposition varies slightly, so no two production runs of chips are identical. Manufacturers measure and sort the output into bins by color.

Even within one chip brand, two production runs can land in different bins. Buying "the same chip" is not the same as buying "the same bin". Bins sit inside bins: within a 3-step ellipse, output can lean yellow toward one edge and blue toward the other. Ordering by step count alone does not pin the color. The bin code does.

A separate, slower mechanism sits on top. Over years of heat, phosphor aging can pull output cooler while yellowing of the encapsulant pulls it warmer. That is aging, not batch variation, and it matters for a different reason: replacements years later.

How our production line holds the color

When a project will be supplied in phases, our production line declares the batch plan before the first run. The LED chips for every phase are reserved from one bin code, whether the order is a batch of floodlights or a continuous run of our IP67 side-bend neon light. The reserve is real inventory allocation, not a verbal promise.

From the first production batch, a physical reference sample is retained with the batch records. Every later batch is compared against this sample under controlled lighting before it ships. Each shipment carries test documentation with bin codes and measured values, but setup values drift from run to run and different test instruments can return different readings for the same lamp. When a measured reading disagrees with the reference sample, the sample decides.

Incoming chips are measured before they reach the line, output is sampled during the run, and finished batches go through an aging run before packing, because color can drift in the first hours of life. When a length has to be replaced years later, after water has worked in through a connection or another failure, the replacement is matched to the retained reference sample, not to the original order's paperwork, so the new piece disappears into the old run. Where two batches must share one visible run, the transition is planned where the eye does not compare: corners, columns, channel joints.

Reserved bins, a reference sample that outranks the paperwork, and transitions placed where the eye does not compare: this is how we keep batches matched across orders, phases and years.

Putting the match into the order

A buyer can reasonably put in writing the MacAdam step for the project (three steps or tighter where runs are visible side by side, two steps where the source itself is directly visible against a white surface), the same bin code across all phases on the purchase order, a retained reference sample, documentation with each shipment, and a powered pre-check of every reel before installation, reels lit side by side on a plain surface. All of it is normal specification practice between professionals.