Datasheet's frequency vs. real-world frequency for IR LED

Question

I am very new to calculating frequencies, but here goes...

The datasheet for an infrared LED states that it can handle a current of up to 1A when the voltage is pulsed at 300 pulses per second and a pulse width of 10us:

Am I correct in thinking that this equates to a frequency of 300Hz with a 0.3% duty cycle? I get the frequency by:

 1 second/300 = 0.0033 second = 300Hz 

and I get the duty cycle by dividing "on time" by period, if 10 us = 0.00001 seconds

(0.00001 seconds / 0.0033 seconds) = 0.003 = 0.3%.

So then my other question comes when I read this article on Wikipedia that states that many electronics pulse IR at 38 kHz:

http://en.wikipedia.org/wiki/Consumer_IR

And there's a tutorial I read that instructed me to pulse the voltage at about 13 seconds on and 13 seconds off for 38kHz. Isn't that a 50% duty cycle? How can the duty cycle be so much bigger when the pulse "on time" is nearly the same? Is it assuming a much smaller current than 1A?

Answer 1

This is not directly an answer to the question (datasheet interpretation), but some relevant experimental information ("real world" as in the question title) on how LEDs are actually used in commercially successful remote controls.

This is measured from a 1056B01 remote control (supplied with Time Warner, Rogers and other cable boxes).

The LED drive circuit consists of an MMBT4403 in an SOT-23 package driving two paralleled IR LEDs. The LED current will be the same as the emitter current, minus a bit from the base drive (about 15mA).

^{simulate this circuit – Schematic created using CircuitLab}

I disassembled the remote, applied a regulated 3.00V supply in place of the batteries (the holder comes inconveniently apart),

and monitored the voltage across the 2.2 ohm resistor.

So the voltage is about 0.74V, implying 340mA, minus 15mA for the base current. If the LED current splits evenly between the two LEDs, that's 160mA each.

So it appears that the LEDs are run at a nominal current of ~160mA each at around 30% duty cycle during the pulses. I'm not going to get into the overall timing of IR remotes, it's easily found.

Note that IR remotes are a low-cost consumer product used in benign environments (temperature wise anyway), and the overall duty cycle of the LED is very, very small, so they can run them much closer to the line than high-reliability 24/7 industrial, medical or aerospace applications. In other words, be careful about extrapolating this to other applications that don't fit the same mold.

Answer 2

At higher duty cycles you have to reduce the current. Without any other information it looks like you would need to limit the current to 60 mA in order to satisfy the limits on the data sheet.

The LED might work just fine with higher current at 50% duty cycle, but then again it might not, or it might work for a few minutes or a few weeks before failing. The information you provided simply cannot guarantee that anything above 60 mA will be within the manufacturer's limits.

You also will need to verify that you meet the specified power limit, but you haven't provided enough information to comment on that.

Answer 3

That datasheet snippet does NOT state that the LED can handle 1A. That figure is in the Absolute Maxmimum Ratings section. The definition of that section is roughly "these circumstances won't kill this LED immediately". Note the 'immediately' and also (although implicit) that it does not state that the LED will operate in its normal fashion (that is, emit IR light).

Summary: don't read that section.

Now for the rest: the fact that such LEDs are often used in a certain way (50% duty cycle) does not invalidate that it CAN be used in other ways (much lower duty cycle). But do read the "Normal Operating Conditions" section to find out which peak current the LED can handle.