Why did this LED fail?

I recently purchased a "Jumbo - Super Bright Red LED" from Radio Shack. It says on the package "Rated 2.4V, 20 mA." On the back of the package, it gives the following information:


Absolute maximum ratings: Forward current: 40 mA max. Forward voltage: 5V DC max. Reverse voltage: 6V DC Power dissipation: 100 mW

Electro-optical characteristics: Forward voltage: 2.4 V +/- 0.03 V Peak emmission wavelength: 660 nm Luminous intensity: 5000 mcd (typ. at 20 mA)

note: Short lead is cathode.


I gave it a try today. I used it with my Arduino's 3.3 V output without a resistor (I thought I would be fine because I was well within the absolute maximum ratings). It was never nearly as bright as my Jumbo blue LED, but it seemed fine. Then when I unplugged it, and plugged it back in later, it didn't work. It only gave a very dim glow. I made sure to never get the leads mixed.

Am I misunderstanding something about absolute maximum ratings, or was this just a faulty LED?

I thought I would be fine because I was well within the absolute maximum ratings

How do you know how much current was flowing through the LED when you first tried it in the circuit?

Sounds like you didn't limit the current with a proper series resistor ;)

https://learn.adafruit.com/all-about-leds?view=all

Yes! That should have been obvious to me. I’ll pick up a new LED tomorrow and wire it in a 5V series with a resistor around 130 ohms (from V=IR). Thanks!

Need a resistor of around 150 ohms from 5V to Anode. Cathode to ground (shorter leg), without the resistor, the LED will draw too much current and it will damage pretty quickly.

Are you sure it’s the LED that is gone and not the Arduino pin?

Shpaget: Are you sure it's the LED that is gone and not the Arduino pin?

Good idea to check but I am willing to bet $100 right now that it is the LED. MEGA328Ps have a decent amount of series resistance on their outputs and they are far more robust than the datasheet indicates, at least in my experience making screwups similar to the OP for a short period of time when I started working with electronics.

JoeN: MEGA328Ps have a decent amount of series resistance on their outputs and they are far more robust than the datasheet indicates

Did you miss this?

ChrisCrossCrash: I gave it a try today. I used it with my Arduino's 3.3 V output

That's not an I/O pin. So I'm guessing your $100 is even safer than you realise :)

Well, I missed it for sure.

ChrisCrossCrash: Absolute maximum ratings: Forward current: 40 mA max. Forward voltage: 5V DC max. Reverse voltage: 6V DC Power dissipation: 100 mW

Am I misunderstanding something about absolute maximum ratings, or was this just a faulty LED?

Well yes, because you have to stay within all of them, here the abs max voltage happens to be irrelevant since the current will reach amps and the device rapidly fail before the voltage can get to 5V. The forward voltage is given as 2.4V and current is an exponential function of forward voltage. Note that the power maximum is just saying the same thing, 40mA is the limit, as 40mA at 2.4V = 100mW...

Incidentally the lead length is not the way to determine the polarity of an LED since the convention varies from manufacturer to manufacturer. However all leaded LEDs have a notch or flat on the body on the cathode side, just as all diodes have a band or mark on the cathode end.

UPDATE: I got it working on the 5V pin in series with two resistors in parallel (560 ohm and 220 ohm). None of the pins on my Arduino are dead.

I found the required resistors in the following way:

  • The total circuit votage is 5 V. The voltage drop across the LED should be 2.4 V (from the package info), so the voltage drop across the resistor must be 5 V - 2.4 V = 2.6 V
  • To choose a resistor, I take that voltage along with the current flow(20 mA specified by the LED package), and plug it into the equation R=V/I so that R=(2.6 V)/(0.020 A)=130 ohms
  • Since I don’t have a 130 ohm resistor, I will connect two resistors in parallel to get a closer resistance; in this case, two 220 ohm resistors. I did this using the equation: Rtotal = R1R2/(R1 + R2). Using this formula, these resistors in parallel give 110 ohms of resistance.
  • It’s alive! Now I will check it with a multimeter. The resistance across the two resistors is 109.5 ohms, which is close to the predicted 110 ohms. The resistance across the LED is practically nothing. When the circuit is closed, the voltage across the LED is 2.02 V, and the voltage across the resistors is 2.94 V. The current through the circuit is 26.4 mA.

I’m still a little confused about why the voltage across the resistors is higher than 2.6 V, even though the resistance is slightly lower than it should be (110 ohms instead of 130 ohms). This voltage gets even higher if I increase the resistance by removing one of the resistors, which is what I would expect.

Nothing is ideal in practice, so the fact that you get something sightly higher isn't that surprising. There could be more resistance in wires you may be using etc.

Its good to check with calculations and ohms law etc as you have done, so good job! :)

When it comes to diodes, calculating exact voltage from current (or visa verca) Is very tricky. They have a characteristic curve, something like this image. You are somewhere on the very steep (but not quite vertical) part of the curve.

Actually LEDs tend to have a much sharper curve than this as the bandgap is higher. However the actual curve varies between devices and with temperature, so you never control the voltage, you control the current and let the device find its own Vf

Couple of comments: 1) I've burned out a crapload of LEDs, but I've never seen one fail in any way other than open circuit, no light. If you get dim light out of it, that suggests a problem elsewhere (which seems to be what you found)

2) If an LED is being driven by an arduino pin, but looks almost invisibly dim, you probably forgot to do pinMode(pin,OUTPUT).

3) You can get LEDs much cheaper on ebay than from radioshack. Same is true of other common parts - resistors, caps, common transistors, etc. http://www.ebay.com/sch/i.html?_nkw=LED+assortment&_sacat=12576 The Shack is often 10-100 times more expensive - for desperation only. Like, you get 5 resistors for the price of 100 (shipped from china) on ebay.

4) If you apply the forward voltage recommended in the datasheet directly to the LED, you will almost never get the "expected" current. The voltage is an estimate, to help you size the resistors. You'll be surprised how much in electronics isn't exact.

I've seen a lot of LEDs cooked from overcurrent that shift color and get very dim. Red LEDs that turned dim amber.