If L.E.D. lights rated at 220 Volts are supplied at 240 Volts will they perhaps flash on and off?

Strange this. But I have a number of led lamps that gradually start doing this flashing on and off after a few hundred hours of use. They have ES caps although marked 230V, and I have a bc cap one with the same problem marked 240V. Very curious, no diagnosis yet!

I had a couple of Philips pseudo-bulbs which failed. The first stage was that the columns of LEDs were dancing round and round.

I have opened up some non-dimmable led lamps that are just a series capacitor and then a bridge rectifier and long  string of perhaps 100 diodes totalling 150V or so. These are very sensitive to supply voltage, as the LED has a very sharp (near exponential) turn-on above a threshold of 1.5 volts or so per chip.

I have also seen some with a simple linear regulator in series with the diodes to give a degree  of constant current, but the heat dissipation in the regulator part  is then highly dependant on the difference between the rectified mains and the voltage across the LED chain. We know that the lifetime of the LEDs and any linear regulator will be critically dependant on temperature, so perhaps it is not such a surprise that over voltage causes problems.

At best the flashing may be an automatic thermal shut down  of a regulator, or more finally  it may be the LED die coming away from the metal part of its body that serves as both a contact and heatsink.  You need surprisingly few tens of milliwatts to raise a cubic mm of semiconductor to a failure temperature, once it is not well connected to the outside world in a thermal sense.

As I said earlier, I do not know what is in the dimmable ones, but when one fails I'll come back and report assuming I can find the thread again.

Mike

I have opened up some non-dimmable led lamps that are just a series capacitor and then a bridge rectifier and long  string of perhaps 100 diodes totalling 150V or so. These are very sensitive to supply voltage, as the LED has a very sharp (near exponential) turn-on above a threshold of 1.5 volts or so per chip.

I have also seen some with a simple linear regulator in series with the diodes to give a degree  of constant current, but the heat dissipation in the regulator part  is then highly dependant on the difference between the rectified mains and the voltage across the LED chain. We know that the lifetime of the LEDs and any linear regulator will be critically dependant on temperature, so perhaps it is not such a surprise that over voltage causes problems.

At best the flashing may be an automatic thermal shut down  of a regulator, or more finally  it may be the LED die coming away from the metal part of its body that serves as both a contact and heatsink.  You need surprisingly few tens of milliwatts to raise a cubic mm of semiconductor to a failure temperature, once it is not well connected to the outside world in a thermal sense.

As I said earlier, I do not know what is in the dimmable ones, but when one fails I'll come back and report assuming I can find the thread again.

Mike

mapj1 said:

At best the flashing may be an automatic thermal shut down  of a regulator

Ooh, I think that I may have found out how to quote.

Mine was not flashing randomly - the columns were flashing in turn, so the result was a rotating beam. Quite apt on the coast!