LED Calculations

How to Calculate Takt, Burn-In Capacity, and Photometric Workload for LED Fixture Assembly

Step-by-step formulas for the core LED and luminaire assembly calculations, with worked numbers, real units, and where each input comes from.

Start with assembly takt, the pace one finished luminaire must leave the line to meet demand. Takt equals available production time divided by required units. If a shift runs 8 hours with two 15 minute breaks and 10 minutes of startup, net available time is 445 minutes, or 26,700 seconds. For a daily requirement of 900 fixtures across one shift, takt is 26,700 / 900 = 29.7 seconds per unit. The Luminaire Final Assembly Takt calculator uses this exact structure. Pull required units from the firm order backlog, not a forecast, and subtract every planned stoppage from gross shift time before dividing.

Once takt is known, size each station by comparing its cycle time to takt. A station that torques 6 screws at 2.4 seconds each plus a 4 second gasket placement runs 18.4 seconds, comfortably inside a 29.7 second takt, so one operator suffices. A driver wiring station at 41 seconds exceeds takt, so you need ceil(41 / 29.7) = 2 parallel operators to keep the line balanced. Line balance efficiency is the sum of station times divided by (number of stations multiplied by the bottleneck time). Track it to catch stations idling below 80 percent utilization.

Burn-in capacity governs how many drivers or full fixtures you can soak-test in parallel. Capacity in units per day equals (chambers multiplied by slots per chamber multiplied by available hours) divided by burn-in duration hours. With 3 chambers, 48 slots each, running 20 available hours against a 12 hour soak, capacity is (3 x 48 x 20) / 12 = 240 units per day. The Burn-In Test Capacity calculator models this. If your build rate is 900 per day but only 30 percent get burned in per an AQL sampling plan, you need to soak 270 per day, so 240 capacity is short by 30 and forces a second daily rotation or a fourth chamber.

Photometric test workload sizes your integrating sphere and goniophotometer bench. Workload in test hours per day equals units requiring test multiplied by test time per unit, divided by 3,600 if inputs are in seconds. If 6 percent of 900 daily fixtures need a full goniophotometer scan at 25 minutes each, that is 54 units multiplied by 1,500 seconds = 81,000 seconds, or 22.5 test hours. A single goniophotometer running 20 hours per day cannot absorb it, so you either add a bench or cut sampling. The Photometric Test Workload calculator separates sphere quick-checks (often 90 seconds) from full angular scans, which matters because the two differ by a factor of 15 or more.

Driver thermal derating protects lumen maintenance and warranty life. LED drivers carry a rated case temperature, often 75 C at the Tc point, and life halves roughly every 10 C above it per Arrhenius behavior. If measured Tc is 92 C against a 75 C rating, you are 17 C hot, cutting a nominal 50,000 hour rating to about 50,000 / 2^(17/10) = 15,400 hours. To hit a 50,000 hour target you must lower Tc to rating or below, which usually means reducing drive current. Dropping from 700 mA to 550 mA cuts driver dissipation by roughly (550/700)^2 = 0.62, a 38 percent heat reduction.

Convert electrical inputs to light output using efficacy. Fixture efficacy in lumens per watt equals delivered lumens divided by system input watts, where system watts include driver losses. A board producing 4,800 lm at 32 W of LED load, driven by a 90 percent efficient driver, draws 32 / 0.90 = 35.6 W at the wall, giving 4,800 / 35.6 = 135 lm/W system efficacy. Always quote at system level, since board-level efficacy of 150 lm/W collapses once driver and thermal losses are included. This number feeds both energy labels and the retrofit savings math downstream.

SKU variant math determines how many distinct build configurations you actually run. Total variants equal the product of options per attribute: 4 lumen packages multiplied by 3 CCTs multiplied by 3 optics multiplied by 2 finishes equals 72 SKUs. If each SKU needs a unique burn-in recipe and photometric baseline, your test matrix explodes far faster than volume. The SKU Variant Complexity calculator quantifies this. A practical control is attribute pooling: sharing one driver across three lumen packages via programmable output cuts driver SKUs from 4 to 1 and removes 18 of the 72 combinations from your qualification backlog.

Published 2026-07-01.