Author Topic: High power LED damage reasons  (Read 1401 times)

0 Members and 1 Guest are viewing this topic.

Offline roman.isaikinTopic starter

  • Contributor
  • Posts: 16
  • Country: no
High power LED damage reasons
« on: December 26, 2021, 09:39:55 pm »
Hi!
Burned down a 500$ LED today (YJ-BC-270H-G02-56-G/S) and trying to figure out what I did wrong. It was connected to TTi CPX-400S 60V/20A 420W power supply, so my assumption was that it cannot damage it, because power limit would be hit first. Obviously this is not true.. Power supply was in CC mode, I cranked up the current knob a little bit too fast and short circuit protection triggered. Could it be that some of the die's didn't had time to heat up evenly and caused Vf drop (assuming that LEDs have negative temperature coefficient, can't seem to find that far sure). Or maybe 40cm wire inductance caused an instability in the power supply current controller. Either way would it have helped to set PSU voltage protection on the 40V or not?
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: High power LED damage reasons
« Reply #1 on: December 26, 2021, 10:07:25 pm »
Silly question perhaps, but is this the model with thermal radiator and fan, or is it the one without?
 

Offline rpiloverbd

  • Regular Contributor
  • *
  • Posts: 157
  • Country: bd
Re: High power LED damage reasons
« Reply #2 on: December 26, 2021, 10:16:39 pm »
Hello, Sorry if I am wrong. YJ-BC-270H-G02-56-G/S can withstand 35V- 42V input voltage, right? Was the power supply giving 60V input to it? Without any current limiting resistor?
« Last Edit: December 26, 2021, 10:18:40 pm by rpiloverbd »
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: High power LED damage reasons
« Reply #3 on: December 26, 2021, 10:21:53 pm »
OP states that the power supply was in CC mode. No need for resistors.
 

Offline roman.isaikinTopic starter

  • Contributor
  • Posts: 16
  • Country: no
Re: High power LED damage reasons
« Reply #4 on: December 26, 2021, 10:23:54 pm »
With the radiator of course. Theoretically it can output more than the LED is rated for, I'm just trying to figure out why would that happen in CC mode + 420W power limit of the PSU and whether setting an overvoltage protection or cranking up the current more slowly would have helped.
 

Offline roman.isaikinTopic starter

  • Contributor
  • Posts: 16
  • Country: no
Re: High power LED damage reasons
« Reply #5 on: December 27, 2021, 12:16:39 pm »
Thank you for the answers! I'll post a microscope picture of the LED later here.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9333
  • Country: fi
Re: High power LED damage reasons
« Reply #6 on: December 27, 2021, 01:56:12 pm »
Current limitation might be too slow to react due to output capacitance for example, but that should happen if your CV voltage is set high and you connect the wires to the LED suddenly, with output set on.

If it really did die during you ramping up the CC setpoint, it seems strange.

If unsure about of the behavior of PSU, using series resistor is a good idea. Of course it wastes some power because now you need to output higher voltage than the Vf of the LED, part of which is wasted in the resistor. But the PSU still maintains the CC regulation and you can do whatever measurements you want to do in lab.
« Last Edit: December 27, 2021, 01:59:28 pm by Siwastaja »
 
The following users thanked this post: roman.isaikin

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: High power LED damage reasons
« Reply #7 on: December 27, 2021, 02:10:12 pm »
diode researchers are very specific with power protection features on PSU for these applications.

prototype guys pay alot extra for safety
 
The following users thanked this post: roman.isaikin

Offline m98

  • Frequent Contributor
  • **
  • Posts: 634
  • Country: de
Re: High power LED damage reasons
« Reply #8 on: December 27, 2021, 11:15:17 pm »
That TTi power supply has quite an exotic feature set, so I wouldn't really trust the output for such an application without further characterisation. What I do to test all kinds of LEDs and laser diodes on my bench is to just use an ancient linear power supply, set the current, turn the output voltage to zero, connect the diode and turn the voltage up a bit over the point where the set current is reached.
 
The following users thanked this post: roman.isaikin

Offline roman.isaikinTopic starter

  • Contributor
  • Posts: 16
  • Country: no
Re: High power LED damage reasons
« Reply #9 on: December 28, 2021, 03:14:01 pm »
Thank you for your answers, will be more careful next time.
No picture unfortunately, nothing can be seen on a microscope.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: High power LED damage reasons
« Reply #10 on: December 28, 2021, 07:27:51 pm »
It's an awful failure.
The power supply might be slowww. You have to test it to really know. Although it has a fast CV spec, overload is 500msec and no specs exist for CC response time.
Some power supplies screw up at the transition from CV->CC or CC->CV, either done slowly or lots of overshoot.
Since the power supply can easily overpower the LED, I would have added a ballast resistor which can be (voltage) monitored to see if the PSU is well behaved.

I would wonder if the LED thermal (paste and screw torque) was properly done. Were you getting decent heat transfer? Looking at Vf can give a rough estimate of LED temperature.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf