Cold Start: What Happens Inside a Driver Below Zero
Cold start is a different test from normal operation, and drivers that work perfectly in a warm laboratory can fail it. At sub-zero temperatures the electrolytic capacitors behave differently: the electrolyte becomes more viscous, the equivalent series resistance rises, and the capacitance available for smoothing falls. The result is a larger start-up current, a slower rise of the bus voltage and a control circuit that may not reach its operating threshold before the timing runs out.
The control IC is the critical piece. Most switching controllers have an undervoltage lockout, and they need a minimum voltage on their supply pin before they start switching. At low temperature the bias winding and the auxiliary supply come up more slowly, and the startup resistor or the internal high-voltage current source must supply enough current long enough for the controller to take over. A design with too little startup current works at 25 degrees and fails at minus 20.
The LED side has its own temperature behavior. An LED forward voltage rises at low temperature, because the semiconductor bandgap widens as the junction cools. In a constant-voltage system with a resistor, a higher forward voltage means a lower current, so the fixture is dimmer when cold and brightens as it warms. The effect is a few percent for a well-designed module and larger for a cheap one, and it is one reason a specification may ask for a warm-up period before photometric measurement.
The inrush current is also worse when cold, in one respect. The capacitor ESR is higher, which slightly limits the peak, but the control loop may restart several times before stabilizing, and each restart brings another pulse. Combined with a cold electrolyte, the stress on the input rectifier and the fuse is higher than at room temperature, which is why cold-start testing cycles power repeatedly at the low temperature limit.
Standards define the test conditions rather than the behavior. IEC 61347-1 sets the ambient range for the product, and a driver rated for operation down to minus 20 or minus 40 degrees is expected to start and operate at that temperature. The manufacturer tests at the limit, often in a climatic chamber with a defined soak time, and the certificate or the datasheet records the result. A product with no low-temperature rating is not the right choice for a cold store or an unheated outdoor cabinet.
In the field, cold-start problems look like a fixture that needs several attempts on a winter morning, or a light that comes on dim and brightens over half an hour. Both are worth investigating before replacing the driver, because the cause may be the LED module rather than the supply. For the PV100, the operating range should be checked against the site minimum temperature, and where the site goes below the rating, the driver should be installed in a heated enclosure rather than pushed outside its specification.
Some designs include a heater or a self-warming circuit for extreme cold, and some use capacitors with a wider temperature range. These choices are invisible on the datasheet unless the operating range is stated, which is why the temperature rating is the first thing to check when the site is a cold store, a freezer room or an unheated cabinet in a cold climate.
Test data at the low limit is more convincing than an extended range printed on a label. Ask for the cold-start test report, including the soak time and the number of restart cycles, when the application depends on starting reliably at the lowest expected temperature rather than merely operating there once warm.
Operating at Altitude: Why Thin Air Changes Thermal and Insulation Margins
Output Ripple: Where It Comes From and How It Reaches the LEDs
Related Article