LM-80 is the LED chip manufacturer's homework. LM-79 is the fixture manufacturer's exam. TM-21 is the graduation prediction. Confuse these three, and you'll ask the wrong questions — like accepting a 100,000-hour L70 claim when the underlying LM-80 test was only 3,000 hours. This guide walks you through reading a real LM-80 report (Samsung LM301B), calculating L70 for your specific fixture, and spotting the three red flags that separate legitimate lifetime claims from marketing fiction.
Before you can evaluate any LED lifetime claim, you need to understand which standard applies to which part of the product:
| Standard | What It Tests | Who Needs It | Duration |
|---|---|---|---|
| IES LM-80 | LED chip/package lumen maintenance over time at controlled temperatures | Chip manufacturer (e.g., Samsung, Nichia, Osram) | Minimum 6,000 hours at 3 case temperatures |
| IES LM-79 | Complete fixture — total luminous flux, CCT, CRI, power consumption | Fixture manufacturer | ~2–4 hours (single-point-in-time measurement) |
| IES TM-21 | Projects L70/L90 lifetime from LM-80 data using mathematical extrapolation | Anyone specifying lifetime — specifiers, manufacturers, end users | Mathematical projection (no physical test) |
Here's the real LM-80 test data for the Samsung LM301B LED at case temperature Ts = 85°C and drive current If = 105mA:
| Test Hours | Lumen Maintenance (%) | Lumen Depreciation (%) |
|---|---|---|
| 1,000 | 99.8 | 0.2 |
| 2,000 | 99.5 | 0.5 |
| 3,000 | 99.3 | 0.7 |
| 4,000 | 99.1 | 0.9 |
| 5,000 | 98.8 | 1.2 |
| 6,000 | 98.5 | 1.5 |
After 6,000 hours at 85°C case temperature, the LM301B still produces 98.5% of its initial output. The average depreciation rate over the last 2,500 hours is approximately 0.06% per 1,000 hours — extremely gradual. Using TM-21 projection methodology, this data supports an L70 projection exceeding 54,000 hours at the 85°C test temperature.
The LM-80 data gives you L70 at the test temperature. Your fixture runs at a different junction temperature (Tj) — and that difference can multiply the lifetime. Here's the step-by-step method:
For a well-designed LED panel or linear fixture in a typical indoor environment, the temperature rise from ambient to junction is approximately 25–35°C. So at 25°C ambient: Tj ≈ 50–60°C. For enclosed fixtures, downlights with poor ventilation, or high-wattage designs, the rise can be 40–60°C — pushing Tj toward 85–100°C even at room temperature ambient.
The Samsung LM301B was tested at Ts = 85°C. Your fixture's actual Tj of ~55°C is 30°C lower than the test condition. Lower temperature = longer life.
The semiconductor industry rule of thumb: every 10°C reduction in junction temperature approximately doubles the LED lifetime. With a 30°C delta from the test temperature: 30°C ÷ 10°C = 3 doublings → 2³ = 8× lifetime multiplier. So 54,000 hours × 8 ≈ 432,000 hours theoretical L70 at your operating temperature.
With 6,000-hour LM-80 data and a well-designed fixture operating at Tj ≈ 55°C, you can legitimately claim L70 ≥ 36,000 hours (the TM-21 cap). In reality, the fixture will likely last far longer — 100,000+ hours is realistic — but the standards framework limits what you can claim on a datasheet. For procurement, the practical takeaway is: LM-80 data from a top-tier chip manufacturer + a well-designed thermal path = LED lifetime that exceeds the practical life of the fixture's driver and mechanical components.
These three checks will catch 90% of exaggerated or fraudulent L70 claims:
TM-21 caps projections at 6× test duration. A 3,000-hour LM-80 test limits the maximum claimable L70 to 18,000 hours — yet some manufacturers cite 3,000-hour test data and claim L70 of 50,000. That's a 16.7× projection, nearly triple the allowed maximum. If a product claims L70 50,000 hours and the supporting LM-80 test is under 8,334 hours, the claim violates TM-21 methodology.
The LM-80 test is conducted at a specific drive current. If the fixture operates the LED at a higher current than the test condition, the LED runs hotter and degrades faster. Example: a fixture driving LEDs at 150mA but the LM-80 report tests at 65mA. The actual lifetime will be significantly shorter. Always verify that the fixture's operating current matches or is lower than the LM-80 test current.
A manufacturer citing standard LM-80 data at Ts = 85°C without measuring their own fixture's junction temperature is guessing. They don't know whether their thermal design keeps Tj at 55°C (excellent) or lets it drift to 95°C (rapid degradation). Ask for: (1) a thermocouple measurement of Ts at the LED solder point in a fully assembled fixture, and (2) the calculated Tj using the LED's specified thermal resistance (Rth). If they can't provide both, they haven't done the thermal engineering — and the lifetime claim is unsubstantiated.
When evaluating an LED fixture's lifetime claims, request these five documents and check each one:
Only if the manufacturer provides the original LM-80 test report from the LED chip manufacturer AND a TM-21 projection calculation for their specific fixture. Many product pages cite the LED chip's LM-80 data without accounting for their own fixture's thermal design — which can reduce actual lifetime by 50% or more. A legitimate manufacturer will provide: the LM-80 report number and issuing lab, the TM-21 calculation showing the projected L70 value and the 6× cap limit, and the measured Tj (or Ts) for their fixture. Without all three, the L70 claim is marketing, not engineering.
L70 = 70% of initial lumens remain. This is the industry-standard end-of-useful-life threshold for general lighting applications — offices, warehouses, retail, industrial. When people say "50,000-hour LED," they mean L70 at 50,000 hours.
L90 = 90% of initial lumens remain. This is a much tighter standard used for applications where even small lumen depreciation matters — retail displays where product colors must stay accurate, museum lighting where conservation standards apply, photography studios, and healthcare examination rooms. An LED rated at L70 50,000 hours might only achieve L90 15,000–20,000 hours because the initial depreciation is faster before the curve flattens. Always check which metric is being claimed — "50,000-hour life" without specifying L70 or L90 is meaningless.
Yes — and this is one of the most overlooked failure modes in LED lighting. A failing driver can deliver current spikes, ripple, and transient overshoot that accelerate LED degradation far beyond what the LM-80 data predicts. The LM-80 test uses a precision laboratory DC power supply with essentially zero ripple. In a real fixture, a driver operating near its thermal limit may produce output ripple, startup overshoot, or transient spikes during AC line disturbances — all of which stress the LED junction.
The practical rule: match the driver's rated lifetime to the LED L70 estimate. A 100,000-hour LED paired with a 30,000-hour driver will fail at 30,000 hours when the driver's electrolytic capacitors dry out. Quality drivers from Mean Well, Inventronics, and Tridonic publish lifetime curves at multiple case temperatures — use these to verify that the driver lifetime at your fixture's internal ambient temperature exceeds the LED L70 projection.
References: IES LM-80-21 — Approved Method: Measuring Luminous Flux and Color Maintenance of LED Packages, Arrays and Modules | IES LM-79-19 — Approved Method: Electrical and Photometric Measurements of Solid-State Lighting Products | IES TM-21-22 — Technical Memorandum: Projecting Long-Term Luminous Flux Maintenance of LED Light Sources
Find LED fixtures with verified LM-80 test reports, TM-21 L70 projections, and documented thermal measurements
Search Verified LED Fixtures with LM-80 Data →Filter by LED chip brand (Samsung, Nichia, Osram), LM-80 test duration, TM-21 L70 projection, and driver brand — free for B2B buyers