6v6gt:
It's a bit late now, but this appears to be the OP's device together with a specification. Apologies if I am repeating something already buried in the thread.
There is even some Arduino code to drive it
That looks like it, or pretty close.
He's not using code to drive it though. This is an Arduino-free project. 
If I'd blown the led by doing what I did, lesson learned. It didn't blow
That is a very crude and crass statement.
It seems you are not capable of actually learning anything. The fact that your LED still functions after a fashion is absolutely no vindication of your stupid actions. You have damaged that LED their is no doubt about that. But I suspect you are not able to realise this, you certainty don't have the test equipment to measure this damage.
So by your criteria any automobile accident that does not result in a fatally can be repeated with nothing needing to be learned.
Please grow up.
6v6gt:
It's a bit late now, but this appears to be the OP's device together with a specification. Apologies if I am repeating something already buried in the thread.
There is even some Arduino code to drive it
Nice find, thank you, I'll save that for later. bookmarks
How do you know you haven't ruptured the hermetic seal.
You have three vary intelligent engineers telling you that
you've done something wrong. You answer that with a statement
that it didn't seem to have done any harm.
One wonders why you come to this bulletin board for advice
and then as a show of foolishness, choose to ignore that
advice.
If you were working in my department, you would have received
your first warning.
Let us hope that you don't someday create a design that people's
lives might depend on.
Dwight
Since I prefer the proactive approach, I would say that the OP's greater "sin", as it were, was to take any action without FIRST verifying that course of action was correct. Where I work, we work with high voltages and high currents. There is no "warning:". You get it right the first time or your dead, plain and simple.
The thing I'm having trouble coping with at the moment is why everyone is freaking out at me for experimenting... It's a spare LED out of a bag of 10 that I got for 5$.
The other thing I'm having trouble with is why everyone is spazzing out about a current limiting resistor when there's no extra voltage to drop with said resistor.
Instead of being a bunch of condescending assholes, you could instead fill the gaps in my knowledge.
To the best of my knowledge, you only get runaway current on an LED when you apply a higher voltage than what the led is rated at.
If I'd been putting 5V on the LED, I would have put a 1.5ohm resistor in series to limit the current. (1.2V/800mA=1.5ohm).
I'm confused how applying a series resistor that will drop voltage needed to make the LED run will benefit me in any way.
Because the forward voltage changes with temperature and the phase of the moon, so you never know just what voltage to apply.
Instead of being a bunch of condescending assholes, you could instead fill the gaps in my knowledge.
We started out more polite, and explained the situation well, but you ignored all that was said. Even now, you don't seem capable of listening/understanding what's been said to you.
A series resistor is always required. No need to repeat more of what's already been said. It's been well explained. The cost of the LED is irrelevant - it's about learning good practices. No one with any knowledge of the subject powers up an LED without current limiting.
And calling people assholes is not on. I don't remember calling you names at any point, despite the fact that you were being stupid and ignorant.
Obviously I've just been wasting my time, since you know so much better. Good luck with it. I'm outta here.
The fully charged battery is 4.2V and you said the forward voltage was "rated for" 3.8V. You never told us the mAh rating of your 18650 battery. If we assume 0.001 ohms (1 mohm) resistance between the battery leads and the led leads. that's
(4.2V-3.8V)/0.001 ohms = 400A.
We aren't freaking out. We are telling you that your approach to learning is stupid.
It doesn't matter how many leds you have or how much money you have. Your approach to learning is stupid because you are too impatient to do any research before taking action. It's your PHILOSOPHY (of electronics) that we are criticizing.
The idea that "I don't know what the fxxk I'm doing and I'm too lazy to ask so I'll just connect this led directly to the battery to see what happens. Why is that stupid ? Because what you don't know is that electronic components explode and project shrapnel at high velocity when subjected to high currents in short time periods.
You could have lost an eye , or both.
I'm outta here...
rollerce:
The thing I'm having trouble coping with at the moment is why everyone is freaking out at me for experimenting... It's a spare LED out of a bag of 10 that I got for 5$.
A paperclip is even cheaper, so I don't know why people freak out about kids sticking them into wall outlets.
And you know what's even cheaper than a $0.50 LED? A $0.03 resistor.
rollerce:
The other thing I'm having trouble with is why everyone is spazzing out about a current limiting resistor when there's no extra voltage to drop with said resistor.
It's in the name. "Current limiting".
rollerce:
Instead of being a bunch of condescending assholes, you could instead fill the gaps in my knowledge.
Condescending assholery would've been more like "Wow what an idiot, why would you even make a thread asking about such a simple project my dog could do in his sleep. Get lost and come back when you actually have a problem"
What I've been reading is "hey, here are some ways for you to learn safely, please don't hurt yourself". Quite different.
rollerce:
To the best of my knowledge, you only get runaway current on an LED when you apply a higher voltage than what the led is rated at.
The best of your knowledge, as you acknowledged, is a gap.
rollerce:
If I'd been putting 5V on the LED, I would have put a 1.5ohm resistor in series to limit the current. (1.2V/800mA=1.5ohm).
The math is correct but the numbers are wrong.
http://led.linear1.org/1led.wiz
See how wrong you are.
rollerce:
I'm confused how applying a series resistor that will drop voltage needed to make the LED run will benefit me in any way.
Learning how to do things the right way is a pretty good benefit. Bad habits in new situations are dangerous. What doesn't kill you, just might in a new environment.
I'm still confused...
If i'd measured the battery to be 4.2v, I would have put a series resistor in. I measured the battery at 3.37v, which i knew to already be less than the forward voltage of the LED, so I saw no harm in powering the led from it for a moment or two.
You keep harping on about the battery at full charge, BUT IT WASN'T! The battery I used was at 3.4v. Please tell me what resistor makes sense to use when powering an LED with a forward voltage of 3.8V in that circumstance. Someone said earlier that they were surprised the LED fired at all, so why would I have put a series resistor in, AT THAT VOLTAGE?
The fact that you didn't have the proper equipment to measure the led current (a Tek scope with a DC current probe) means that you have no idea what the current was during that brief flash.
At the very least you should have measured the current with a DMM.
The fact that you did something with no idea what the current would be is sufficient to conclude it was foolish. Your speculation about what the current could or couldn't be is irrelevant. The fact is that by your own statement,
the flash was blindingly bright, indicating excessive current. How much is irrelevant.
The fact of taking action with foreknowledge that the current is unknown is evidence enough of foolishness or worse. The fact that you continue to argue the point indicates you are incapable of learning.
your question is nonsense (sorry, but it's a fact ! ) you will not power it at that voltage !!!
the Led is rated 3.2V to 3.8V at 750mA (according to the datasheet) - If you want to design something using it, you should use a power source which allows you to give it the power while making sure you won't exeed the current - If you plan to use a 18650, then do the math with a fully charged battery and use the right resistor - While the battery "discharge" (what is the english word ?) , the led will be less bright, and you'll need to charge the battery or replace it with a fully charged one -
Powering it without a resistor with a partially charged battery could have burnt it - It didn't probably because it did not last long, and maybe because the battery couldn't deliver too much current... you've been lucky - But be aware that, if it didn't burn, it might be damaged, and will probably won't run for years !
and...if you think we are assholes...hum... an asshole is useful (I should say... indispensable), an idiot is not 
rollerce:
The thing I'm having trouble coping with at the moment is why everyone is freaking out at me for experimenting... It's a spare LED out of a bag of 10 that I got for 5$.
One of my classmates misread lab instructions and, instead of swapping the input signal connections to a 741 op amp, he swapped the power leads.
The current surge through the now-forward-biased ESD protection diodes caused the IC to literally explode with a loud pop that everyone in the room heard, and a large chunk of the plastic casing broke off and flew upward, almost hitting the ceiling.
A 741 op amp is also less than $0.50. If he had been less lucky and not wearing glasses, his eyes could have been a lot more expensive.
I'm still confused...
If i'd measured the battery to be 4.2v, I would have put a series resistor in. I measured the battery at 3.37v, which i knew to already be less than the forward voltage of the LED, so I saw no harm in powering the led from it for a moment or two.
You keep harping on about the battery at full charge, BUT IT WASN'T! The battery I used was at 3.4v. Please tell me what resistor makes sense to use when powering an LED with a forward voltage of 3.8V in that circumstance. Someone said earlier that they were surprised the LED fired at all, so why would I have put a series resistor in, AT THAT VOLTAGE?
The fact that the LED lit that brightly means that your assumption about its forward voltage value was wrong. Accept that, and try to find out what chain of events or reasoning led to that mistake so that you do not repeat it. What made you think the forward voltage was 3.8V? Did you get it from a website? A datasheet? Did you measure it in some way?
- While the battery "discharge" (what is the english word ?) , the led will be less bright
Discharge is correct, though the proper verb form to use in this case is discharges (While the battery discharges...).
If you'd asked we'd have told you that 4.2V or even 5V is not
enough head room to safely run a LED with 800ma @ 3.8V
and use a current limiting resistor.
You could do 5V with an active current limit.
We tell you that you are doing it wrong and your response,
despite the large number of others telling you it was wrong
and why, you insist it was OK because you, the battery
and the LED seemed to have survived.
We all make mistakes. Most try to learn from them.
It was your continued insistence that you did nothing wrong
that cause the progressive stronger replies.
It has been stated several times to you that the voltage stated
is just an approximate voltage, useful for estimating a current
limiter.
It is not to say that you can connect the same voltage to
is and expect proper operation.
The specification are not there as a learning tool for those
that don't have the slightest idea as to what they are doing.
They are there for an engineer that already knows that
LEDs require current limiting.
The voltage rating was there to assist in a proper designed
current limit.
Dwight
800mA LED really needs a current limit driver chip which acts like a switching constant current supply.
This keeps the LED current constant even when the LED heats up and wants to allow even more current flow which can lead to a runaway condition and burnout.
Example:
Hi,
rollerce have you put one of your LEDs in series with the battery and a 10R resistor.
The MEASURED the volt drop across the LED, your battery and the 10R, it should give you enough LED current to get a good approximation of the LED volt drop and current for those conditions.
Tom... 