Sources<\/a><\/li>\n<\/ul>\n<\/nav>\n<\/div>\nDirect answer: what power supply does an LED product need?<\/h2>\n An LED is a current-driven device, so the first decision is constant current or constant voltage: use a constant-voltage driver (usually 12V or 24V) when the LED strip or module carries its own current-setting resistors, and a constant-current driver when the LEDs are bare engines whose current you must set yourself \u2014 commonly 350, 500, 700 or 1050 mA within a stated voltage window.<\/strong> For any team choosing an LED power supply OEM partner, the three specifications that decide whether the product survives in the field are correct driver type and current, dimming compatibility, and thermal and life design<\/strong> \u2014 not the wattage printed on the box.<\/p>\nGet those three wrong and you get flicker at low dim, drivers that die at two years, and nuisance breaker trips. At Paiyi Power we build these from 200 pcs per model, with samples in about a week and first production in 5\u20136 weeks after approval.<\/p>\n
Why LED power is its own discipline<\/h2>\n A charger feeds a circuit that asks for a voltage. An LED asks for a current, and the voltage follows wherever the junction temperature takes it. Four consequences make LED power different from any other adapter category:<\/p>\n
\nThe load is non-linear.<\/strong> A small change in forward voltage produces a large change in current, which is why driving LEDs from a fixed voltage and a resistor is a thermal compromise rather than a design.<\/li>\nForward voltage moves with heat.<\/strong> As the LED warms up, its forward voltage falls. A constant-current driver holds brightness steady; a badly matched constant-voltage design drifts brighter and hotter.<\/li>\nDimming is a compatibility problem, not a feature.<\/strong> The driver, the dimmer (wall or protocol) and the LED load must all agree. Mismatches show up as flicker, buzz or a useless bottom end.<\/li>\nThe driver is expected to outlive the fixture.<\/strong> LED lifetimes are quoted in tens of thousands of hours, so the driver’s electrolytic capacitors \u2014 not the LEDs \u2014 usually set the real service life.<\/li>\n<\/ul>\nThat is why specifying an LED driver on wattage alone is the most expensive mistake in this category.<\/p>\n
Constant voltage vs constant current<\/h2>\n This is the single decision that shapes everything else, and it is decided by how your LED product is built \u2014 not by preference.<\/p>\n
\n\n\nAspect<\/th>\n Constant voltage (CV)<\/th>\n Constant current (CC)<\/th>\n<\/tr>\n<\/thead>\n \n\nTypical output<\/td>\n 12V or 24V DC (48V for long runs)<\/td>\n Fixed current, e.g. 350\/700\/1050 mA, in a voltage window<\/td>\n<\/tr>\n \nUsed for<\/td>\n LED strips and modules with on-board resistors; multiple parallel runs<\/td>\n COB, high-power single LEDs, luminaires with no on-board current limiting<\/td>\n<\/tr>\n \nSetting the light output<\/td>\n Watts per metre x metres; you buy enough supply capacity<\/td>\n You choose the current; brightness is whatever the LED bin gives at that current<\/td>\n<\/tr>\n \nThermal behaviour<\/td>\n Current shifts with forward voltage; needs margin and good strip design<\/td>\n Current held constant regardless of temperature; more predictable<\/td>\n<\/tr>\n \nVoltage drop<\/td>\n A real constraint \u2014 long strips dim toward the end unless fed from both ends<\/td>\n Driver compensates; but the LED string must sit inside the driver’s voltage window<\/td>\n<\/tr>\n \nTypical efficiency<\/td>\n Good, and simpler; losses land in the strip’s resistors<\/td>\n Higher system efficiency; no resistor losses<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\nThe practical rule: if your product has resistors built in, you want CV; if it does not, you want CC.<\/strong> Asking for the wrong one is the most common error we see in incoming specifications.<\/p>\nMatching the driver to the load<\/h2>\n For a constant-current design, four numbers have to line up:<\/p>\n
\nOutput current<\/strong> \u2014 set to the LED engine’s rated current, and the driver must hold it accurately across its input range.<\/li>\nVoltage window<\/strong> \u2014 the driver’s output voltage range must cover the LED string’s forward-voltage spread, including the cold end. A 30\u201342V driver and a 36\u201340V string leave almost no headroom.<\/li>\nVf binning<\/strong> \u2014 LEDs are binned for forward voltage; the spread across bins is what eats your voltage headroom. Ask for the bin range, not just the typical value.<\/li>\nThermal end-stop<\/strong> \u2014 the driver needs over-temperature protection and, for enclosed fixtures, a lower maximum case temperature.<\/li>\n<\/ul>\nFor a constant-voltage design the arithmetic is simpler \u2014 watts per metre times metres \u2014 but two habits matter: never load a CV supply to more than about 80% of its rating (strips are often understated), and feed long runs from both ends so the far end is not visibly dimmer.<\/p>\n
Dimming: TRIAC, 0-10V, DALI and PWM<\/h2>\n Dimming is where most LED projects quietly fail, because it is a system property rather than a driver property. The main options:<\/p>\n
\n\n\nMethod<\/th>\n Where it fits<\/th>\n Watch out for<\/th>\n<\/tr>\n<\/thead>\n \n\nTRIAC \/ phase-cut (mains)<\/td>\n Retrofit residential; uses the existing wall dimmer<\/td>\n Low-load flicker and flashing; trailing-edge dimmers suit LED better than leading-edge<\/td>\n<\/tr>\n \n0-10V \/ 1-10V (analog)<\/td>\n Commercial and industrial; simple, robust, widely specified<\/td>\n Needs an extra control pair; dimming curve differs between brands<\/td>\n<\/tr>\n \nDALI<\/td>\n Addressable commercial lighting, building control<\/td>\n Higher driver cost; commissioning expertise required<\/td>\n<\/tr>\n \nPWM (internal)<\/td>\n Consumer products, RGB, colour tuning<\/td>\n Low-frequency PWM causes visible flicker and camera banding<\/td>\n<\/tr>\n \nWireless \/ smart<\/td>\n Connected and tunable-white products<\/td>\n Standby power and always-on connectivity add to the energy budget<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\nTwo rules cut most dimming support calls: specify the dimmer and the driver together<\/strong>, and test the bottom 10% of the dimming range<\/strong>, not the middle. Problems live at the bottom.<\/p>\nFlicker, and why it is a specification<\/h2>\n Flicker is the failure mode nobody puts in a datasheet until it matters. It becomes obvious when a camera films a lit room, when a dimmer is near its bottom end, or when someone is sensitive to it.<\/p>\n
\nSource:<\/strong> low-frequency PWM dimming is the usual cause; the LED is switched on and off fast enough to look steady to the eye but not to a rolling shutter or to a sensitive viewer.<\/li>\nFix:<\/strong> use high-frequency PWM or constant-current reduction (analog) dimming for camera-facing and premium applications.<\/li>\nSpecify it:<\/strong> ask for a flicker figure and a dimming curve. IEEE 1789 is the usual reference for acceptable flicker levels; add “flicker-free at 25% dim and below” to the specification in plain words.<\/li>\n<\/ul>\nIf a product is designed for video, broadcast or retail photography, flicker belongs in the requirements list, not in the post-mortem.<\/p>\n
Efficiency, power factor and harmonics<\/h2>\n LED drivers are judged on more than efficiency, and in commercial lighting the extra numbers are mandatory rather than optional:<\/p>\n
\nPower factor (PF).<\/strong> For loads above roughly 25W, many markets expect PF of 0.9 or better; poor PF means more current for the same watts and can breach installation limits.<\/li>\nHarmonics.<\/strong> Lighting equipment is subject to the IEC 61000-3-2 Class C limits; this is the standard that separates a real lighting driver from a repurposed adapter.<\/li>\nSystem efficiency and heat.<\/strong> Driver losses become heat inside the fixture, and heat is what shortens the capacitor life that sets the driver’s real lifetime.<\/li>\n<\/ul>\nHigher efficiency is therefore worth more than the electricity it saves \u2014 it buys life.<\/p>\n
Inrush and breaker coordination<\/h2>\n This is the specification that causes the most embarrassing commissioning failures, and it is invisible on a datasheet that only lists input power.<\/p>\n
LED drivers draw a very high inrush current for a few milliseconds as their input capacitors charge. Ten of them on one breaker can trip it at switch-on even though the running load is trivial.<\/p>\n
\nAsk for the inrush figure<\/strong> (peak current and duration) rather than “soft start”.<\/li>\nSpecify inrush limiting<\/strong> (usually an NTC) where many drivers share a circuit.<\/li>\nCoordinate with the breaker type<\/strong> \u2014 a B-curve MCB tolerates far less inrush than a C-curve.<\/li>\n<\/ul>\nFixing inrush after a building is wired is far more expensive than specifying it before.<\/p>\n
Thermal design and driver life<\/h2>\n The driver is usually the shortest-lived component in an LED fixture, and its life is set by temperature:<\/p>\n
\nCapacitor rating.<\/strong> A 105 \u00b0C electrolytic lasts several times longer than an 85 \u00b0C part at the same internal temperature. This is the highest-value specification change available in this category.<\/li>\nCase temperature and derating.<\/strong> Drivers inside sealed luminaires run far hotter than on a bench; require a maximum case temperature and an ambient derating curve.<\/li>\nPotting where it counts.<\/strong> For outdoor and vibration-exposed products, full encapsulation removes internal air and the condensation path that outdoor drivers otherwise suffer from.<\/li>\n<\/ul>\nEvery unit we ship goes through 100% ATE functional testing and a burn-in period before packing, and boards are built by qualified partner SMT houses with incoming inspection at our factory \u2014 assembly, test and aging are in-house.<\/p>\n100% ATE functional test before shipment \u2014 every unit, not a sample.<\/figcaption><\/figure>\nBurn-in on the aging rack \u2014 where weak drivers are found before a project is installed.<\/figcaption><\/figure>\nIndoor, outdoor and IP ratings<\/h2>\n Where the driver lives decides its enclosure, and the requirements are stricter than most buyers expect:<\/p>\n
\nIndoor (IP20)<\/strong> \u2014 inside a ceiling void or a luminaire body; the constraint is heat, not water.<\/li>\nOutdoor (IP65\u2013IP67)<\/strong> \u2014 sealed and preferably potted; the practical enemies are condensation and UV, not rainfall, which is why a fully sealed driver that still breathes through its cable gland eventually fails.<\/li>\nWet and damp locations<\/strong> \u2014 some regions require an explicit wet-location rating for the luminaire and the driver; this is a listing requirement, not a preference.<\/li>\nSurge<\/strong> \u2014 outdoor, street and area lighting often specify IEC 61000-4-5 surge levels far above indoor values, because long pole-mounted runs collect transients.<\/li>\n<\/ul>\nIt is the same discipline we describe for outdoor signage in digital signage power supplies<\/a>.<\/p>\nCertifications: what actually applies<\/h2>\n LED drivers sit under a different safety standard from ordinary adapters, and one very common misconception is worth correcting:<\/p>\n
\nSafety:<\/strong> the IEC\/EN 61347<\/strong> series (61347-1 and 61347-2-13) is the standard for lamp controlgear including LED drivers. In North America, UL 8750 covers LED equipment and UL 1310 covers Class 2 supplies. Some constructions are also assessed under the IEC 62368-1 framework.<\/li>\nEMC and harmonics:<\/strong> IEC 61000-3-2 Class C for lighting, plus the usual emissions standards for the destination market.<\/li>\nDoE Level VI does not apply to LED drivers.<\/strong> Level VI is an external power supply regulation; lighting has its own efficiency and quality programmes \u2014 notably the DesignLights Consortium (DLC) qualified products list and California Title 24 for commercial projects. If a supplier answers “Level VI” to a lighting efficiency question, they are answering a different question.<\/li>\nSELV and Class II:<\/strong> confirm which insulation class your installation expects, because it changes the construction and the cost.<\/li>\n<\/ul>\nWe certify per model and per market rather than claiming one certificate covers everything; certification by country<\/a> explains which schemes transfer and which need local registration.<\/p>\nLED power supply specification table<\/h2>\n
\n\n\nParameter<\/th>\n Typical for LED lighting<\/th>\n Note<\/th>\n<\/tr>\n<\/thead>\n \n\nDriver type<\/td>\n CV 12V\/24V for strips; CC 350\u20131050 mA for engines<\/td>\n Set by whether the LED has on-board resistors<\/td>\n<\/tr>\n \nDimming<\/td>\n TRIAC, 0-10V, DALI or PWM \u2014 matched to the load<\/td>\n Test the bottom 10% of the range<\/td>\n<\/tr>\n \nFlicker<\/td>\n High-frequency PWM or analog; flicker-free below 25% dim<\/td>\n Required for camera-facing applications<\/td>\n<\/tr>\n \nPower factor \/ harmonics<\/td>\n PF 0.9+ above ~25W; IEC 61000-3-2 Class C<\/td>\n Lighting-specific, not adapter-grade<\/td>\n<\/tr>\n \nInrush<\/td>\n Figure stated; NTC limiting where many share a breaker<\/td>\n Causes nuisance MCB trips at switch-on<\/td>\n<\/tr>\n \nThermal<\/td>\n 105 \u00b0C capacitors, max case temperature, ambient derating<\/td>\n Driver life = capacitor life<\/td>\n<\/tr>\n \nEnclosure<\/td>\n IP20 indoor; IP65\u2013IP67 potted outdoor<\/td>\n Condensation and UV, not rain<\/td>\n<\/tr>\n \nSafety standard<\/td>\n IEC\/EN 61347 series; UL 8750 \/ UL 1310 in North America<\/td>\n Not the same as a generic adapter<\/td>\n<\/tr>\n \nReliability<\/td>\n 100% ATE + burn-in; written change control<\/td>\n Fleets outlive product cycles<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\nFinished-goods warehouse \u2014 lighting programmes ship in batches matched to project schedules.<\/figcaption><\/figure>\nFrequently asked questions<\/h2>\nDo I need a constant voltage or a constant current driver?<\/h3>\n It depends on whether your LED product contains its own current-setting resistors.<\/strong> Strips and modules with on-board resistors want constant voltage (12V\/24V); bare COB and high-power engines without resistors want constant current, specified in milliamps.<\/p>\nShould I choose 12V or 24V?<\/h3>\n 24V for almost anything with a long run or meaningful power; 12V for short, low-power runs and legacy compatibility.<\/strong> At 24V the current is halved, so voltage drop along the strip is roughly a quarter of the 12V case and the far end stays evenly lit.<\/p>\nWhich dimming method should I specify?<\/h3>\n Analog 0-10V or DALI for commercial projects, TRIAC for residential retrofit, PWM for colour and consumer products.<\/strong> Whichever you pick, specify the dimmer and driver together and test the bottom 10% of the range \u2014 that is where incompatibilities appear.<\/p>\nHow do I avoid flicker?<\/h3>\n Use high-frequency PWM or constant-current (analog) dimming, and ask for a flicker figure and dimming curve.<\/strong> Standard mains dimming at very low loads is the classic source, and any application involving a camera should treat flicker as a hard requirement.<\/p>\nWhich safety standard applies to LED drivers?<\/h3>\n The IEC\/EN 61347 series (61347-1 and 61347-2-13) for lamp controlgear, with UL 8750 and UL 1310 in North America.<\/strong> It is a different standard from the 62368-1 world of generic adapters, and a supplier who does not know the difference is a risk.<\/p>\nDo LED drivers need DoE Level VI?<\/h3>\n No \u2014 Level VI is an external power supply regulation, not a lighting one.<\/strong> Lighting has its own programmes, notably the DesignLights Consortium qualified products list and California Title 24 for commercial work. If a supplier promises “Level VI” for an LED driver, you are discussing the wrong standard.<\/p>\nHow do I size a driver for an LED strip?<\/h3>\n Watts per metre x metres, then keep the load at or under about 80% of the supply rating.<\/strong> Strips are often understated, and leaving 20% headroom also keeps the driver cooler, which directly extends its life.<\/p>\nWhy do breakers trip when the lights switch on?<\/h3>\n Inrush current.<\/strong> Drivers charge their input capacitors in the first milliseconds and can draw many times their running current. Ask for the inrush figure, specify NTC limiting where several drivers share a circuit, and allow for breaker curve when coordinating.<\/p>\nWhat MOQ and lead time apply for custom LED power supplies?<\/h3>\n From 200 pcs per model at Paiyi Power, with samples in about a week and first production 5\u20136 weeks after sample approval.<\/strong> An LED power supply OEM should also fix the specification before tooling: use our sample evaluation checklist<\/a> and the OEM agreement guide<\/a> for change control.<\/p>\n