LED reverse breakdown voltage

[jaylar41]

Hello,

I'm wondering if anyone has any experience with subjecting an LED to a reverse voltage that is higher than its Vrbd.

The specific question would be this: How long before the LED would see catastrophic failure? Most LED datasheets that I've read specify a Vrbd of 5V. Hypothetically, if the LED was connected in series with a resistor and an 4VAC power source, and there was no reverse protection on the LED, then how long would it be before the LED failed? I specify 4VAC because at its peak, it would be just slightly higher than the Vrbd of 5 V.

Also, what generally happens to the forward voltage of an LED over time. Does it tend to decrease or increase? I have not found much discussion on these topics and was hoping someone had direct experience.

Thanks for any advice.

 

[IRstuff]

As with humans, it's the current and not the voltage that kills. You need to see what reverse current winds up being. Anything higher than about 1 mA peak would be of great concern, both for increasing the diode temperature as well as increasing the potential for a junction short.

TTFN

 

[itsmoked]

If you reverse it beyond V_RBD, it will likely be destroyed or its characteristic's will change to something unacceptable.

If you are going to run AC to an LED you always put a diode in reverse across it(parallel). That way you will never subject it to more than ~1V backwards.

Use any old diode; 1N40xx or a 1N4148.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jaylar41]

Thanks guys.

My question about the forward voltage may have been a little unclear earlier. Let's forget about the AC example for a moment. Under normal operating conditions, does anyone know if Vf on an LED tends to increase, decrease, or stay the same?

 

[IRstuff]

Generally, again, it's not the voltage that matters, but the current. The high current densities in these types of diodes causes migration of the metallization into the diode junction, which effectively behaves like an ever increasing short across the diode.

So the actual diode's turn-on voltage probably does not change that much, but the amount of current drawn will continue to increase, until it's essentially impossible to turn on the diode itself, without running so much current as to destroy it anyway.

TTFN

 

[itsmoked]

There are about 10 chemistries laid out in 8 different geometries so all of this is pretty variable.

Current screws with V_F more than temperature.

Temperature screws with V_F more then age.

And age will increase V_F over a few thousand hours. On the order of 100-200mV.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MacGyverS2000]

You will rarely, if ever, find an LED that lists anything other than 5V for the reverse voltage. Manufacturers no longer take the time to find out the exact voltage as most customers don't care. Some will truly be in the 5V range, but most are higher these days. Grab a handful of the LEDs you expect to use and throw some reverse power at them, see where they turn over. There are plenty of reverse strings of LEDs running off of 120VAC with long life out there (I don't recommend it from an engineering standpoint, but you can't argue with success).


Dan - Owner

http://www.Hi-TecDesigns.com

 

[OperaHouse]

Take any LED and put a 47K rsistor in series with it. Plug it into a wall outlet and post back when it fails. I wouldn't do it in a space shot, but I haven't seen one fail yet.