LED reverse breakdown voltage

[E-design]

One of my friends recently approached me with something puzzling about the reverse breakdown voltage of some LED's.

He bought some cheap bulk 5mm red and green LED's from China and decided to measure the characteristics on his curve tracer.
What he found surprised him. Most data sheets show the reverse breakdown voltage to be in the region of 4-7V.
The red LED's that he tested broke down at approximately 35V and the green at about 124V.

When he informed me, I thought that he was not reading a scale correctly or something. Well, he gave me a visit and brought some of these LED' along.
I tested it on my Tektronix 577 curve tracer and got the same readings. The forward readings were normal around 1.8V for the green and 1.5V for the red.

I then did some search on the Internet and found someone who had a similar experience with an even higher breakdown voltage of 185V.

"The diode started conducting a significant current at 185 V (which, by the way, is surprisingly high as all datasheets I've seen rate such LEDs at maximum 5 V reverse bias)"

https://www.tablix.org/~avian/blog/archives/2009/11/reverse_biased_led/

Anyone here had a similar experience?

 

[Venkadesh_M]

It is not punctured more than the specified reverse voltage but did you tested It is working in forward direction or not??? I have blown many leds with 12V in reverse direction thats why asking.........

 

[E-design]

Yes, they are still working perfectly after the reverse voltage tests. Current was limited at a few hundred uA's during the tests. I was wondering if some types of LED's have these unusual high breakdown voltages. I have seen/tested other LED's before that had much lower breakdown voltages but can't remember ever testing ones with these crazy high reverse voltages.

 

[Venkadesh_M]

as we know each color, power requirement of led depends the materials, may be the certain combination could with stand such a reverse voltage but we cant use it at all because that is not recommended........

 

[E-design]

Here are some pictures taken from the curve tracer display.

 

[erikl]

... a reverse voltage ... we cant use .. at all because that is not recommended........

If the reverse current is sufficiently limited to avoid damage, the reverse-conducting LED is a useful noise diode.[/URL]

quoted from here.

Reverse breakdown voltages depend on the doping of the more lightly doped region of the pn-junction. GaN, InGaN and SiC LEDs usually use lower doped raw substrates than the usual GaAs or GaAsP LEDs.

 

[Venkadesh_M]

quoted from here.

Reverse breakdown voltages depend on the doping of the more lightly doped region of the pn-junction. GaN, InGaN and SiC LEDs usually use lower doped raw substrates than the usual GaAs or GaAsP LEDs.

Hi, I mean it is not recommended to use in such a high voltage(more than specified in datasheet).........

 

[schmitt trigger]

Interesting!
I am going to repeat the experiment myself over the weekend.

However, how much light was shining on the LEDs? I know LED exhibit Photodiode action, which means that current flows when reverse biased and light applied.

 

[Venkadesh_M]

Interesting!
I am going to repeat the experiment myself over the weekend.

:grin: But take the precautions of blowing off at any time......... :grin:

 

[E-design]

Interesting!
I am going to repeat the experiment myself over the weekend.

However, how much light was shining on the LEDs? I know LED exhibit Photodiode action, which means that current flows when reverse biased and light applied.

At first, I just had some indirect ambient light from the bench reaching the LED under test. I was also wondering about the effect of this light, so later I shined a bright light from my LED flashlight into the green LED, and it made very little difference. The only noticeable effect was that the breakdown knee wasn't that sharp (just a slight rounded corner) anymore, but the breakdown still occurred at the same voltage.

 

[erikl]

I know LED exhibit Photodiode action, which means that current flows when reverse biased and light applied.

That's how some automatic wardrobe LED lights work: opening the door creates some photocurrent, which lights the same LED(s). Of course this circuit has to shut down the light after a while in order to test for external light.