LED deck lights burned out

[James L]

I have a limited info scenario on my hands, and the info has trickled in slowly

About 2 months ago I went to a service call, bad gfci on a deck. There's a landscape lighting transformer plugged in that is powering 10 small LED lights on the posts.

There's also a 120v plug which is activated by the built in timer on the transformer.

The homeowner had a string of Christmas lights plugged in there, strung along the railing.

At the time, the concern was that the Christmas lights weren't working. With the new gfci, the transformer powered up and the string of lights came on.

But last Friday, I was called out because 7 of the deck lights are burned out.

I checked the voltage at 2 of the fixtures, where the wires are soldered in, and it showed 12.7 volts AC.

I thought it strange to have AC voltage for 12v LEDs....maybe, maybe not? The lights don't appear to have any labeling on them.

I was under the impression that these lights stopped working one at a time, and was working under that assumption. I was explaining that it could be cheap components in the light, accumulating heat inside the sealed fixtures, etc

Then the homeowner threw me a curve, and said they all worked fine until I changed the gfci.

They didn't like my reply to that, though... They might have worked fine until the gfci went bad, but how could they work just fine when the transformer wasn't getting power from the receptacle?

So my question(s)

Is there a scenario where a voltage spike could damage the LEDs and gfci, but not damage the transformer?

If an LED is designed for DC voltage, yet is powered by AC voltage, does the voltage range need to be narrower?

I'm supposed to speak with the lighting rep who sold them the fixtures and transformer, but thought I'd get some input here first.

 

[gar]

150112-2149 EST

If the LEDs are simply an LED with a series current limiter, then they should not be fed AC. An LED is basically a diode. This means it conducts with a fairly low voltage drop in the forward direction, and conducts very little in the reverse direction until a breakdown voltage is reached.

The series current limiting device (possibly a resistor) limits the current in the forward direction and at that limiting current the voltage across the LED is determined by the LED.

In the reverse direction the current drawn (leakage current) is so small that the current limiter does not limit the reverse voltage on the LED. If the peak AC voltage is greater than the breakdown voltage of the LED, the the LED gets destroyed.

.

 

[James L]

150112-2149 EST

If the LEDs are simply an LED with a series current limiter, then they should not be fed AC. An LED is basically a diode. This means it conducts with a fairly low voltage drop in the forward direction, and conducts very little in the reverse direction until a breakdown voltage is reached....


... If the peak AC voltage is greater than the breakdown voltage of the LED, the the LED gets destroyed.....

That's what I was thinking, and trying to put that together with the bad gfci.

The 12.7 volts seems fairly nominal, that there almost had to be some sort of voltage spike which would account for the gfci and LEDs being damaged at the same time.

 

[gar]

150112-2414 EST

I thought I might find a low PIV breakdown and I tried two grossly different LEDs and with a 100 k current limiting resistor and 28 V from an an adjustable power supply neither one broke down. So for reasonably new LEDs I don't have a measured result that would be a problem witha a 12 V AC source. But that does not mean that some production LEDs might not have low reverse breakdown. This is not a high voltage controlled parameter.

For the most part the following discussion is useful http://www.electro-tech-online.com/threads/reverse-voltage-on-led.130914/ .

Damage to the GFCI and the LEDs is probahbly a good correlation. A reversed bias diode in parallel with each LED would probably have prevented LED failure.

.

 

[James L]

gar,
Thanks for the insight, and the link.

 

[Electric-Light]

Many "electronic transformers" for low voltage lighting output high frequency AC. It's a lot more similar to your laptop power supply without an output rectifier than a "transformer".

If the GFCI was also fried at the same time, I would suspect excess input voltage due to neutral issues.

 

[James L]

Many "electronic transformers" for low voltage lighting output high frequency AC. It's a lot more similar to your laptop power supply without an output rectifier than a "transformer".

If the GFCI was also fried at the same time, I would suspect excess input voltage due to neutral issues.

There's a thread in the general forum where the transformer output was 20k hz, and a Fluke DVM wouldn't read any voltage because of the high frequency. I also use a Fluke, but it read the voltage.

I suspected excessive input voltage, but was curious how an electronic transformer would withstand a voltage spike that fried the electronics in the gfci and LEDs

I still haven't been in contact with the lighting rep, as there I a GC as a go between first. I appreciate the input. Want to make sure the parts were spec'd to work together

I appreciate the input