UV LEDs die fast?

[John Bokma]

Out of a batch of 50 UV LEDs two have "died" somehow. With the first one I
thought it was a bad connection, so I re-soldered the LED, and it worked.
When I turned the circuit off for some time, same problem. So again I
thought: "cold" connection, re-heated it, and worked...

But when the circuit got cold again, same problem. I replaced the LED, and
no problems for hours.

Two days ago, however, another LED stopped working, same problem. Heating
it, it works again. This time I was able to "see" (indirectly, UV LEDs can
harm your eyes) what happened: the three LEDs in series started to flicker
for some time, and then all went off. Heating it -> works.

Does anyone have an idea why this is happening? Just 2 bad LEDs out of 50?

See http://johnbokma.com/pet/scorpion/detection-using-uv-leds.html for the
circuit diagram.

Also: I use most of the time 9V to feed this circuit, so I doubt it has
anything to do with overloading the LEDs.

TIA,

 

[Lawrence Oravetz]

John Bokma explained :

Out of a batch of 50 UV LEDs two have "died" somehow. With the first one I
thought it was a bad connection, so I re-soldered the LED, and it worked.
When I turned the circuit off for some time, same problem. So again I
thought: "cold" connection, re-heated it, and worked...

But when the circuit got cold again, same problem. I replaced the LED, and
no problems for hours.

Two days ago, however, another LED stopped working, same problem. Heating
it, it works again. This time I was able to "see" (indirectly, UV LEDs can
harm your eyes) what happened: the three LEDs in series started to flicker
for some time, and then all went off. Heating it -> works.

Does anyone have an idea why this is happening? Just 2 bad LEDs out of 50?

See http://johnbokma.com/pet/scorpion/detection-using-uv-leds.html for the
circuit diagram.

Also: I use most of the time 9V to feed this circuit, so I doubt it has
anything to do with overloading the LEDs.

TIA,

look at the traces leading to the led. Is a pad lifted? or has a crack
in the board developed?. reheating is the key common factor.. I'd look
for external faults before the led.

 

[John Bokma]

John Bokma explained :
look at the traces leading to the led. Is a pad lifted? or has a crack
in the board developed?. reheating is the key common factor.. I'd look
for external faults before the led.

2 more leds have died. I bought a digital meter, and am going to do some
measurement. I got more and more the impression that the shop sold me
garbage. (Especially since I use 9V to run the circuit, not 12).

 

[Don Klipstein]

2 more leds have died. I bought a digital meter, and am going to do some
measurement. I got more and more the impression that the shop sold me
garbage. (Especially since I use 9V to run the circuit, not 12).

Should you determine that it is the LEDs, could you tell us what brand
these LEDs are?

- Don Klipstein ([email protected])

 

[John Bokma]

Should you determine that it is the LEDs, could you tell us what
brand
these LEDs are?

I have no idea about the brand, but I ordered them from LSDiodes.com.
They seem to have a good reputation, but I am amazed that I manged to
kill 4 UV LEDs, especially since I thought they would last for thousands
of hours.

I'll check it out next week.