Testing LEDs for ESD degredation

I searched and can't find the Nichia article. I assume driving the LED at a some very low current and checking for light is how it is done.:shrug:

Testing white 5mm LEDs in series with a 1 mega ohm resistor, some light and some don't, but it may be too aggressive test too low of current (using 9v supply).

Thanks

If you are looking for degradation in performance, I would think that you'd need to measure the light output.

If you are looking for damage to the semiconductor... that could be tougher. My experience with ESD is more general. In most semiconductors, ESD damage to the silicon is only observable with an electron microscope, or maybe a very powerful optical microscope??. It's usually damage to a conductive trace in the silicon, or sometimes looks like a crater. This examination of the silicon is usually performed by the manufacturer, since they are the only ones who know the device well.

In your test, are you trying to measure the leakage current, or the light output when driven at low currents?

Steve K.

I guess a ESD degraded LED could work okay, but may fail suddenly in use or have a shortened life time. Someone here mentioned the nichia test, but I could not find the document.

I did some of my own experimenting by holding a LED by one lead with pliers ,shuffling my feet across the rug and touching the other lead to a light switch screw. There is a tiny, faint snap of discharge and it always affected the white LED in the following manner:

Some LEDs would no longer light at normal operating current. These would also now conduct in reverse bias. Apparently ESD causes a conducting "channel" to form. I found that giving them a quick burst of high current at forward bias, 100ma or even a full 6v (no resister, very quick burst so not to burn it), will "remove" the conducting channel and the LED is restored to normal operation (most of the time). Some lost their ability to light at currents in the hundreds of microamp range.

I'm interested in the nature of this ESD induced conduction "channel" and the Nichia article.

I did some more tests by shaking them around in a non antistatic bag and a few other experiments with low static fields.

So far, I've reached the following conclusions of indium gallium nitride blue LEDs (white with phosphor):

They can be degraded/destroyed by improper handling due to static discharge. Although they may be recovered, the longevity of the degraded LED is unknown. I don't have the equipment to check the luminosity of the restored LED, but it seems full brightness visually.

Since the LEDs have a small reverse leakage current, weaker induced static charges simply conduct and even the potential across the leads causing no harm.

It is the sudden snap of static discharge at the higher currents that damage them, so it is important to use a grounding strap on your wrist before handling them. Using antistatic bags does not seem to be necessary due to the reverse leakage as described.

I am not familiar with particulars of ESD damage on LEDs, but in general ESD damage presents itself in many different ways. The LEDs might still be functional, but can turn off at higher currents, as opposed to lower currents that you are testing at.

Also, you are putting around 9 microamps through the LED, according to the datasheet for it, is that enough to have it turned on for sure? Did you test with higher currents?

By the way, what kind of ESD are you getting on these LEDs that got you worried about damage?

rushnrockt said:

I am not familiar with particulars of ESD damage on LEDs, but in general ESD damage presents itself in many different ways. The LEDs might still be functional, but can turn off at higher currents, as opposed to lower currents that you are testing at.

Also, you are putting around 9 microamps through the LED, according to the datasheet for it, is that enough to have it turned on for sure? Did you test with higher currents?

By the way, what kind of ESD are you getting on these LEDs that got you worried about damage?

Click to expand...

9 ua is the current I happen get using a 1 mega ohm resistor. I choose this resistance for a very minimal current to test with. I don't know the minimum recommended turn on current.

I have the LEDs in non anti static plastic boxes and bags. I wanted to be sure any static build up was not going to damage them. Plus I was just interested in finding out their ESD sensitivity. It seems that they are not as sensitive like some MOSFET transistors are.

If you are trying to determine whether you've been shipped damaged parts, I'm not sure if there is a reliable test. Some damaged parts may work fine initially, but die after a short while.

If you are trying to avoid damaging parts, I do recommend following the usual guidelines. Use a conductive (but not metal) bench mat. Use a conductive wrist strap to bleed off charge. Don't wear synthetic clothes. Wear shoes with conductive soles, such as leather. Be especially careful when the humidy is low. If you generate sparks when you walk across the floor, turn on the humidifier!

As far as the packages that LEDs are shipped in, this is certainly a concern. I've seen Cree LEDs come in plain plastic bubble wrap (i.e. not a conductive package, and not even the pink stuff that doesn't generate ESD). Most electronics houses would refuse to accept parts packaged like this. I do recommend keeping sensitive parts in proper ESD conductive bags or with the leads shunted with a low impedance material (such as poking leaded LEDs in the black conductive foam). No sense damaging expensive parts.

Steve K.

JohnR66 said:

9 ua is the current I happen get using a 1 mega ohm resistor. I choose this resistance for a very minimal current to test with. I don't know the minimum recommended turn on current.

I have the LEDs in non anti static plastic boxes and bags. I wanted to be sure any static build up was not going to damage them. Plus I was just interested in finding out their ESD sensitivity. It seems that they are not as sensitive like some MOSFET transistors are.

Click to expand...

Seems like my other comment was posted right after your response. Do you know anything about these LEDs? I guess asking for a datasheet can be a bit of a stretch here, but they usually provide ESD data for Human Body Model and Machine Model.