NerdKits - electronics education for a digital generation

You are not logged in. [log in]

NEW: Learning electronics? Ask your questions on the new Electronics Questions & Answers site hosted by CircuitLab.

Support Forum » Why don't the LEDs burn out?

April 29, 2012
by Pew446
Pew446's Avatar

I have a question I've been quite confused about for a while now, if the MCU outputs 5v, and the LED takes 2.2v max, how can the LED Array kit flash all those LEDs without a single one burning out?

I've been working on a small version of the led array I'm designing, here are a few pictures if you want to see it:

But what I was wondering is how the Nerdkits LED Array doesn't burn out every LED when it's turned on, because the MCU pins output 5v High, and the LEDs are rated at 2-2.2v

Thanks!

April 29, 2012
by Pew446
Pew446's Avatar

And yes I've searched this question, I couldn't find anything :(

April 29, 2012
by Ralphxyz
Ralphxyz's Avatar

Isn't the 2-2.2v rating the voltage drop?

Play around with a multi meter to see what readings you get.

Current is more important for burning things out, more than voltage.

The MCU will output 5 volts at 45ma I believe, you should read the specsheet.

Also the Nerdkit method of lighting an array led does not leave the led on for very long.

Hope someone that actually knows what they are talking about jumps in here :-)

Ralph

April 29, 2012
by mongo
mongo's Avatar

The outputs are only capable of so much current. If you could measure the actual voltages at the LED's, you might just find they are a lot less than 5V

April 29, 2012
by Pew446
Pew446's Avatar

So let me see if I'm understanding this correctly:

Voltage isn't what burns out the LEDs as much as it is Amperage, and by pulsing the pins the amperage is dropped from 45mA to a lower amperage? I hope I'm getting this right.

Bare with me! ;)

April 29, 2012
by Ralphxyz
Ralphxyz's Avatar

Yeah basically, you could hit the led with 12 volts and with a correctly sized limiting (current) resistor light the led without problem

or hit it with with pwm so that it is "lit" for micro seconds. The pwm counts on the viewer's POV (persistance of vision) which removes (reduces) the flickering.

Ralph

April 29, 2012
by Pew446
Pew446's Avatar

Cool! Thanks everyone :)

May 01, 2012
by Pew446
Pew446's Avatar

So if the current is lowered by pulsing, is there an equation to find out how much the current (or amperage,) is lowered based on how often the pin is pulsed?

If every LED is on for 1/10th of the time on the LED array, what would it be for 1/4 of the time?

Thanks for any help :)

May 02, 2012
by Ralphxyz
Ralphxyz's Avatar

I do not believe the current is literally "lowered" it just is not constantly turned on.

Ralph

May 02, 2012
by Pew446
Pew446's Avatar

I must have missed what you said about the pwm in your last post. My new chip came today so I will be testing out a program tonight.

May 11, 2012
by JimFrederickson
JimFrederickson's Avatar

I think a more accurate description of this is...

"Based on the electrical properties of the LED if you only supply the voltage that it has been designed for then the LED will only draw the current it needs... If you supply more voltage than it was designed for then it will allow/force the LED to draw more current from the circuit, to the point of burning up, unless something else limits the current...

The pins on the AVR are "Current Limited" to a value that cannot create enough heat in the LED to burn out most LEDS...

(But I do think that they "sink", send-to-ground, more currrent and if you wire the LED's the other direction they can burn up...)

(Heat produced is Amperage/Current Squared multiplied by the Resistance... More current, more heat... POOF!)"

May 12, 2012
by Rick_S
Rick_S's Avatar

Jim,

Could you please show where in the datasheet it states that when sourcing at 5V the output current is limited on a given pin to less than would damage an LED. To my knowledge, and this has come up a lot, there are no current limiters in the micro-controller. The only limitation I know of is the internal resistance of the micro-controllers circuitry.

It seems odd to me that, if the micro had built in current limiters, then why in every Atmel note I've seen where LED's are used, a current limiting resistor is used as well.

I understand when multiplexing, the resistor may not be needed because the duty cycle is so low that full current draw may never be reached, but when just turning on and leaving an LED on, I believe a current limiting resistor is required.

Rick

May 17, 2012
by mongo
mongo's Avatar

True... You can make LED's seem a lot brighter by hitting them with higher power for very short durations. A small duty cycle will still look like the light is on, though it is flashing many times a second.

Take a look at the cars that have factory LED tail lights. If you sweep your eyes across them to simulate fast movement, chances are you will see the flashing as the tail lights flicker across your field of view. This is the same concept, rather than a constant current draw on the LEDs, they are strobed at higher power to make them appear brighter. As long as they can sink off the heat from the overpower moment, they survive quite well.

Another example is the light bars on police card. Those are some seriously BRIGHT LED's. They are mounted on heat sinks to absorb that shock.

Post a Reply

Please log in to post a reply.

Did you know that a thermometer can be made "faster" by using a bit of math? Learn more...