I have yet to see this in a LED, only HIDs. Not saying there isn't thermal protection just not seeing it in action yet.
This is the failure mode I am seeing with LED drivers.
LEDs aren't always rainbows and butterflies and I feel like they're just not cut out for severe duty. Despite the common belief that LEDs are unaffeceted by cycle length, today's transistorized LED ballasts are full of tiny chips and surface mount components that are very delicate and there are hundreds of solder joints. There are microchips operating at line voltage level and they're not very forgiving of adverse conditions just like telecom equipment and some drives. The elements routinely expose them to high humidity and place them into dew conditions like water near the pole base rising up as the ground and base gets heated up by the sun.
Inside telecom cabinets you're much more likely to find climate control or equipment with internal heater to stay above dew point. Something not found on transformers or regular ballasts. Some LED ballasts are filled with gap fillers but the clear gel that holds the fluorescent phosphor on LED elements and some ballast sealing compounds are water vapor permeable. Some compounds cause physical damage to delicate solder joints or microelectronics parts because of thermal expansion or roadway vibration stress. Giving dew avoidance heater to LED ballast would quickly undo its alleged operation cost savings. Standard HID ballasts simply bake itself out of condensation during normal operation and lamp doesn't mind condensation as long as there isn't liquid water dripping onto hot lamp.
This one explains how ambient conditions affects drives:
https://www.semikron.com/dl/service...ower-electronics-systems-en-2016-07-15-rev-00
There are a few lower wattage HID fixtures like retail lighting ceramic metal halide and novelty applications like indoor pot growers that use electronic ballasts, but they're not used where high reliability is a requirement and would struggle in the climate seen by roadway lighting.
