Why LED Grow Lights Usually Fail at the Driver, Not the Diode
Most growers expect the LEDs themselves to wear out. In the field, the abrupt failure that ends a fixture's working life is far more often electrical than optical, and it usually traces back to one component inside the power supply.
What actually fails first
A modern horticultural LED package is a remarkably durable device. Its output declines slowly and predictably, and manufacturers describe that decline with lumen-maintenance ratings rather than a death date. The power supply behaves differently. It contains a mixture of semiconductors, magnetics, solder joints and at least one aluminium electrolytic capacitor, and it is the capacitor that dominates the temperature sensitivity of the whole assembly.
A survey of field failures in LED drivers presented to the Nordic Test Forum reached a blunt conclusion: electrolytic capacitors are typically the most sensitive electronic component in reliability terms, and they are the component whose ageing sets the practical life of the driver.[3] That is not a defect of cheap products. It is a property of the chemistry inside the capacitor, and it applies to well-built drivers as well.
The 10 °C rule, and why it governs your fixture's life
An aluminium electrolytic capacitor is an electrochemical device. Its electrolyte slowly escapes or degrades, and the rate of that process follows the Arrhenius relationship: warmer means faster. United Chemi-Con's technical documentation states the rule plainly, noting that a 10 °C rise in temperature will double the chemical reaction rate, and that measured acceleration factors generally fall between 1.7 and 2.3 depending on the sample.[1]
The same source reports an electrolyte weight loss of about 1.9 times for a 10 °C temperature increase, with a corresponding rise in equivalent series resistance. Voltage matters far less than temperature inside the rated range, which is why a driver's thermal design is more informative than its voltage margin.
Because the effect is exponential, the arithmetic is unforgiving. Take a capacitor rated for 5,000 hours at 105 °C and place it in an environment 40 °C cooler. Using the ideal doubling rule, projected life becomes 5,000 × 24 = 80,000 hours. Using the more conservative 1.7 factor measured on real samples, the same calculation gives roughly 42,000 hours. Both numbers describe an idealised case at constant ambient temperature, but they show why a 20 °C difference in driver temperature can separate a five-year fixture from a fifteen-year one.
| Ambient at capacitor | Life at 2.0× per 10 °C | Life at 1.7× per 10 °C |
|---|---|---|
| 95 °C | 10,000 h | 8,500 h |
| 85 °C | 20,000 h | 14,450 h |
| 75 °C | 40,000 h | 24,565 h |
| 65 °C | 80,000 h | 41,760 h |
| 55 °C | 160,000 h | 70,990 h |
Where the heat actually comes from
Three contributions stack up inside a grow light.
- The LED board. Even at 2.8 µmol/J, roughly 35 to 40 percent of the electrical input leaves as heat rather than photons, and that heat lands on the heatsink that often shares an enclosure with the driver.
- The driver itself. Conversion losses of 8 to 12 percent are normal for a well-designed constant-current supply, and those watts are released inside a small metal box.
- The room. Canopy-level temperatures of 26 to 30 °C are comfortable for plants and unremarkable for electronics, but the air directly above a fixture can be 10 to 15 °C warmer in a low-ceiling room without vertical mixing.
Mean Well, a large power-supply manufacturer, frames the same issue from the driver side: the key factors influencing luminaire lifetime are the electrolytic capacitors, the light source and the driver, and all three are directly related to temperature.[2] A fixture that keeps its LEDs cool but buries its driver in a dead-air pocket has solved the smaller half of the problem.
What a "50,000-hour" claim really measures
Lifetime figures on grow light datasheets are usually a projection for the LED package, expressed as the time to 70 or 80 percent of initial output at a controlled case temperature. They are legitimate numbers, but they describe the optical component, not the assembly. Once a driver is added, the service life of the product is the shorter of the two, and in most designs that is the driver.
This is why two fixtures using identical LED bins can carry very different real-world reputations. The difference is rarely the diode. It is the size of the heatsink under the driver, whether the driver sits in the air stream, the quality of the capacitors, and whether the enclosure lets convection do its job.
Six things worth checking before you buy
- Driver location. Exposed on top of the fixture, remote-mounted, or enclosed in a sealed compartment? A remote driver in moving air can run 10 to 15 °C cooler than one bolted into a closed housing.
- Capacitor temperature rating. Ask for the rated temperature and endurance of the primary capacitor (for example 105 °C / 5,000 h) rather than the driver's headline life figure.
- Driver efficiency. A 92 percent driver releases roughly half the heat of an 85 percent driver at the same output, which feeds directly back into capacitor life.
- Dimming and standby losses. Continuous dimming to 50 percent reduces dissipation, but a driver that stays warm at zero output is wasting energy and ageing while doing nothing.
- Ingress rating matched to the room. IP65 is sensible for a sprayed greenhouse bay; in a dry vertical farm it adds sealing and can reduce cooling.
- Serviceability. A driver with a standard connector and accessible mounting can be replaced in minutes, which changes the economics of year six.
Frequently asked questions
Do LED grow lights fail suddenly, or just get dimmer?
Both happen, for different reasons. LED packages lose output gradually and are rated to a lumen-maintenance point such as L70 or L80. A driver failure is usually abrupt: the fixture stops starting, flickers, or trips its protection. In practice the abrupt failure is the one that ends the product's working life.
Does a higher IP rating make a grow light driver last longer?
Ingress protection helps in humid, sprayed or condensation-prone rooms, but it does not cancel heat. A fully sealed housing traps internal heat, which raises capacitor temperature and shortens life. The useful question is whether the driver is rated for the room's humidity while still having a credible thermal path to the outside.
Can a failed driver be replaced instead of the whole fixture?
Often yes, if the replacement matches output current, voltage window, dimming protocol and connector type, and if the fixture allows service access. Constant-current drivers are not interchangeable by wattage alone: a driver set to a different current will either underdrive the LEDs or overheat them.
Does running a grow light at 70 percent dimming extend its life?
It can help, because lower output means less internal heat in both the LEDs and the driver. The gain is modest at small dimming steps and is no substitute for adequate thermal design or for keeping the driver out of a hot dead-air pocket.
About the publisher
XineLam is a LED lighting manufacturer with 17 years of experience and 300+ patents in China and internationally. The company designs and produces horticultural luminaires, drivers and thermal assemblies for greenhouse, indoor farm and research applications. This article is educational; product selection should follow a site-specific thermal and electrical assessment.
References
- United Chemi-Con. Capacitor Life — temperature acceleration factors for aluminium electrolytic capacitors. https://chemi-con.com/capacitor-life/
- MEAN WELL. Does using aluminum electrolytic capacitors in LED drivers shorten the lifetime of LED luminaire? 19 July 2018. https://www.meanwell.com/newsInfo.aspx?c=5&i=609
- Nordic Test Forum. Electronic drivers for LEDs — why are many failing? NTF17 session slides (PDF). https://nordictestforum.org/files/slides17/NTF17_s7.1.pdf