Thermal Cycling: Why a Driver That Survives Heat Can Still Fail From Change
Thermal cycling is a different stress from steady heat. A driver that runs at a constant temperature for years is not the same as one that heats and cools every day, and the difference is mechanical. Materials expand at different rates, joints are stressed at every transition, and a product that survives a steady soak can fail after a few hundred cycles. The test that reproduces this is IEC 60068-2-14, which covers change of temperature.
The test has two common variants. In the rapid version the specimen moves between two chambers at different temperatures, so the change is fast and the internal gradients are steep. In the slow version the specimen stays in one chamber whose air temperature is ramped, so the change is gradual and the whole product tracks the air. The two produce different stresses, and the choice depends on what the product will actually experience.
The cycle parameters matter. A range from minus 40 to plus 85 degrees, a dwell long enough for the specimen to reach temperature, and a transfer time that is short relative to the dwell, together define the stress. A hundred cycles is a common qualification, and a thousand is used for products that must survive many years of daily cycling. The number and the range belong in the specification, not in a generic claim.
Failure modes are mechanical. Solder joints crack where the thermal expansion of a component differs from the board, and the crack grows with each cycle until the joint opens. Delamination appears at the interface between the copper and the laminate. Capacitor seals degrade as the internal pressure changes, and potting can separate from the case. Each mode has a characteristic location, and the failure analysis reads like a map of the thermal gradients.
Leadless components are more vulnerable in one respect and less in another. A small surface-mount part with a short lead span sees less differential expansion, but the joint has less compliance, so the strain concentrates in the solder. A leaded part absorbs some of the movement in the lead, but the lead itself can fatigue. The design response is to use compliant terminations, to control the solder volume and to avoid placing large components where the gradient is steep.
Potting changes the stress path. A filled compound bonds the components to the case and spreads the strain over a larger area, which can improve cycling life. It also constrains the board, so the expansion has to go somewhere, and a rigid compound can transfer the stress to the joints instead of relieving it. The choice between a flexible silicone and a rigid epoxy is a thermal-mechanical decision, not only an electrical one.
Field conditions decide how much cycling matters. An outdoor driver in a temperate climate cycles between a cold night and a warm day, and the daily range can be thirty degrees. An indoor driver in a conditioned space cycles very little. A cold store cycles on every door opening, and a stage lighting fixture cycles on every cue. The application, not the product class, sets the requirement.
For the PV100, the indoor rating means the thermal cycling stress is modest in normal use, and the plastic housing and the potting strategy are chosen for that environment. A project that installs the driver in an unheated space, a cold store or an outdoor cabinet has moved the product into a different cycling regime, and the qualification should be checked against the actual range rather than assumed from the indoor rating.
ESD in Handling and Installation: The Failure That Passes the Line Test
Related Article