LED Fault Monitoring

[rjenkinsgb]

i think i will be using a Zener for each LED as a shunt

Just remember that the zener power rating must be rather greater than the individual LED power consumption and that each zener will need sufficient airflow / cooling ability to continuously dissipate significant power.

eg. if you pick a zener with a voltage 50% higher than the LED forward voltage, in case of a LED failure the zener will be continuously dissipating 50% more power than the LED did..

For high-power LEDs you may be in to the realm of stud-mount zeners & heatsinks, which will not be cheap.

 

[ChrisP58]

ON Semi has some thyristor based LED protectors similar to the Littlefuse link you posted inpost #13.
https://www.onsemi.com/PowerSolutions/product.do?id=NUD4700
https://www.onsemi.com/PowerSolutions/product.do?id=HBL1025
https://www.onsemi.com/PowerSolutions/product.do?id=HBL5006
Note, I have not used any of these parts, I'm only posting the links for reference.

One thing to consider when deciding between these and a zener as an automatic shunt around an LED is power dissipation. Since the same current has to flow through the shunt as through the LED, the power dissipated by the shunt will be that current times it's forward voltage.
A zener will need to have a forward voltage higher than the LED it's protecting. Therefore it will need a corresponding higher power rating.
Whereas the thyristor, though it will have a higher trigger voltage, once latched, the forward voltage will be lower than the LED, and therefore does not need as high of a power rating.
That may, or may not, be a factor in your design.
If you do choose a thyristor based protector, you need to make sure that your power supply has enough voltage headroom to trigger the thyristor gate when one of the LEDs opens.

As for detecting that there is a failed LED in one of the strings. The lower total voltage in one string compared to the other three should be easy to measure and see a difference.