LEDs without the use of current limiting resistors

Hi, this is my first post and I'm sad to introduce myself with a question. But I'd really appreciate an answer.

Well, I'm in a project and I was searching some info on the internet when I found this code: http://dlnmh9ip6v2uc.cloudfront.net/datasheets/Components/LED/_7Seg_Example.pde

The thing is that I would love to know wich rules/relations/equations does this guy use to get the current average just with the amount of time that the led is on and off, in this part of the code:

//Display brightness
//Each digit is on for a certain amount of microseconds
//Then it is off until we have reached a total of 20ms for the function call
//Let's assume each digit is on for 1000us
//If each digit is on for 1ms, there are 4 digits, so the display is off for 16ms.
//That's a ratio of 1ms to 16ms or 6.25% on time (PWM).
//Let's define a variable called brightness that varies from:
//5000 blindingly bright (15.7mA current draw per digit)
//2000 shockingly bright (11.4mA current draw per digit)
//1000 pretty bright (5.9mA)
//500 normal (3mA)
//200 dim but readable (1.4mA)
//50 dim but readable (0.56mA)
//5 dim but readable (0.31mA)
//1 dim but readable in dark (0.28mA)

I wonder that this may be related to joule/second = w, but I'm not sure.

Thanks in advance.

The rules he is using is pig ignorance coupled with a lot of stupidity.
This is crap.

See this:-
http://www.thebox.myzen.co.uk/Tutorial/LEDs.html

I am driving LEDs with a constant current driver, and wanted to increase the current ( pulsed ) to increase the brightness.

I asked my LED manufacurer what the safe pulsed current is for the particular LED, and they said 30mA - its rated at 25mA normally.

So its definately not safe to start thumping huge current pulses through the LED, ( or from the Arduio pin )

GMikes tutorial sums it up nicely, but as he says, different LEDs have different pulse tolerances, so check with the manufacturer.

Unless you have a constant current drive, include the 2c resistor.

I dont know where he is getting his figures from, but He seems to be using a 6.25% duty cycle, so maybe is is using a calculation based on duty cycle and the parts forward voltage rating. Lets say your LED is fully lit with 20ma at 3.5, and you are giving it 5V, at 5v lets say you are using 100ma, but with a duty cycle of 6.25 % its using 6.25ma.

Those numbers are made up just to try to make sense of what he is probably doing, Its not too clear from his documentation.

I can tell you that I have run LEDs without resistors, and right now I have a couple sets of 192 LEDs running with out a single resistor, and they have been running for months now without issue. The really crazy thing is that no matter how angry mike gets, my LEDs still shine brightly.

Heres 192 LEDs, 0 resistors, 1 arduino, 1 4-pack AA batteries (6v).

You are saying its OK to pulse it at 100mA, my manufacturer says 30mA - hmm .

Some TV remote controls use the internal resistance of the battery to limit the current through the IREDs, perhaps your running 192 LEDs from a couple of AA batteries is doing the same, how many LEDs are on at the same time ?

Whichever, its bad practice.

I run my 25mA rated LEDs at 18mA, and out of the last 25,000 LEDs since October, I have had only a couple of failures, so I stick to that current.

I am pulsing the latest project ( 50% ) at 21 mA, still well inside of the recommended current.

Boffin1:
You are saying its OK to pulse it at 100mA, my manufacturer says 30mA - hmm .

Some TV remote controls use the internal resistance of the battery to limit the current through the IREDs, perhaps your running 192 LEDs from a couple of AA batteries is doing the same, how many LEDs are on at the same time ?

Whichever, its bad practice.

I run my 25mA rated LEDs at 18mA, and out of the last 25,000 LEDs since October, I have had only a couple of failures, so I stick to that current.

I am pulsing the latest project ( 50% ) at 21 mA, still well inside of the recommended current.

I never said it was ok to pulse 100ma to anything, I described what I think the author of the document that the OP posted was trying to elaborate, but was unclear on how they made their calculations, I gave an example of made up numbers, and i clearly stated that I made the numbers up.

Your manufacturer is talking about "safe limits", most of the datasheets that I've looked over show LEDs at a 20ma max. I dont know how much current the LEDs Im using are running at, but they are running at 5V from wall wart USB battery chargers, one is 500ma, the other is 800ma. When both are hooked up to a kill-a-watt, it shows 1 watt running both (it floats between 0 and 1) so my guess is it draws about 1/2 amp. I do know that the red LEDs are 1.8-2.2 VF) so 5v should be way too much for it.

I have another set that I have just completed, and I've run it from a 6 volt lantern battery, and from 4 AA batteries.

I have read that you can safely double the safe limit when using a duty cycle less than 50%, but I havnt done any testing with that. I have run 74hc595s way over their 70ma limit, for a few days without issue.

I never said it was ok to pulse 100ma to anything,

So have you measured what you supposed good LEDs are actually taking?
There are two things getting damaged the LEDs and the electronics driving them.

I dont know how much current the LEDs Im using are running at,

So how do you know you are not damaging things?

I have read that you can safely double the safe limit when using a duty cycle less than 50%,

Safely doubling safe limits is an oxymoron. You can read all sorts of rubbish on the internet, it doesn't make it true.

I can only tell you what will work 100% of the time for 100% of the people. I know this because I have spent my whole life designing reliable electronics. Any idiot can design unreliable electronics that works for maybe two or three years. Or that works and significantly shortens the life of parts.

You can of course choose to ignore this, I can not stop you from being a total idiot. What I can do is point out that you are an idiot and point out why you are an idiot and why you are suffering from self delusions. Hopefully I can stop sensible people from following your idiotic total lack of logic.

  1. You do not have enough equipment or expertise to know if you are damaging a component.
  2. You appear not to know about peak current and power dissipation and how that relates to average current and average power dissipation.
  3. You think that any activity that doesn't immediately kill you is a safe activity.
  1. You think that any activity that doesn't immediately kill you is a safe activity.

LOL, who first coined the phrase "That which does not kill us only makes us stronger" ?

Lefty

Grumpy_Mike:

I never said it was ok to pulse 100ma to anything,

So have you measured what you supposed good LEDs are actually taking?
There are two things getting damaged the LEDs and the electronics driving them.

I dont know how much current the LEDs Im using are running at,

So how do you know you are not damaging things?

I have read that you can safely double the safe limit when using a duty cycle less than 50%,

Safely doubling safe limits is an oxymoron. You can read all sorts of rubbish on the internet, it doesn't make it true.

I can only tell you what will work 100% of the time for 100% of the people. I know this because I have spent my whole lave you measured what you supposed good LEDs are actually taking?
There are two things gife designing reliable electronics. Any idiot can design unreliable electronics that works for maybe two or three years. Or that works and significantly shortens the life of parts.

You can of course choose to ignore this, I can not stop you from being a total idiot. What I can do is point out that you are an idiot and point out why you are an idiot and why you are suffering from self delusions. Hopefully I can stop sensible people from following your idiotic total lack of logic.

  1. You do not have enough equipment or expertise to know if you are damaging a component.
  2. You appear not to know about peak current and power dissipation and how that relates to average current and average power dissipation.
  3. You think that any activity that doesn't immediately kill you is a safe activity.

I havnt measured the current, but Im fairly confident that the arduino is delivering 5v, and I know that 5V is higher than the forward voltage, and technically should be doing damage. It should also be noted that LEDs wear, even when using them at the recommended specifications, they will eventually fail. so your statement about 100% of the people 100% of the time is 100% false.

  1. false.
  2. false.
  3. false.

What is true, is that you want to call me an idiot, and make a lot of false statements to support your claims that im an idiot. Shame on you mike, you should not behave that way.

I hope you get over your cold or whatever.

What is true, is that you want to call me an idiot,

I don't want to call you an idiot, it saddens me that when faced with the truth you can't see it.
That is my definition of an idiot.

So let's just look at one claim I made:-

  1. You do not have enough equipment or expertise to know if you are damaging a component.

You said this was false.
So you do have equipment and expertise to assess a damaged component!
If you truly have a decapsulating machine and a scanning electron microscope then please post the pictures of your driver output stage after 6 months of over driving an LED with out any form of current control.
If you do not have this equipment then what equipment do you have that can assess internal component damage?

This is starting to look like my threads.... :grin:

popcorn.gif

The thing is that I would love to know wich rules/relations/equations does this guy use to get the current average just with the amount of time that the led is on and off, in this part of the code:

From the datasheet for the mcu and the led.

Leds, especially high brightness leds, behave far more like resistors (at high current levels) than like diodes, in that their voltage drops go up significantly with the current through them - as a north-eastern-ish V-I curve.

That, coupled with diodes surprising resistance to pulsed current, makes them fairly resistant to current shocks.

Together with the mcu's output resistance, this "no resistor" approach works.

Engineering isn't about "you cannot do that", or "that's wrong". Engineering is about knowing why you cannot do it so you know when / where you can do it.

People holding a simplistic view just lack common sense to understand that.

Engineering isn't about "you cannot do that", or "that's wrong". Engineering is about knowing why you cannot do it so you know when / where you can do it.

People holding a simplistic view just lack common sense to understand that.

You know dhenry that is probably one of the few bits of wisdom from you that I can agree with. But even then such wisdom when responding to beginners to electronics is bound to do more harm the good. Too much cargo electronics going around these days as it is.

Lefty

Mike, I dont have that equipment, nor the expertise to use them, but i have used a microscope before to examine dies. You can probably just use a regular microscope for this scale, and The LED even has a clear lens to see right into it. I still wont be bothering to do that. Instead I built the thing, and have been running it as a test to see how long it will take to fail, and it has shown no signs of failure yet. Its been months with no sign of dimming, no smoke... it just keeps working.

Dhenry, thats some sound theorizing. I remember before when you mentioned batteries having an internal resistance, and shift registers possibly working as current limiting devices. Your theories have merit, but I have no idea how to test the theory.

its also possible that parts are under rated or materials or manufacturing have changed, and made improvements that never got re-factored in or tested.

If you have steady state current, open a LED wire connection up and put a meter in series, see what the current flow is.

If its multiplexed, put a shunt in series, and with a scope see the voltage is across the shunt, calculate the current.

I like where your heads at. When in doubt, measure it.

I dont have a scope, I do have a couple cheapo digital multimeters.
Not only is it multiplexed, but its charliplexed. 192 LEDs on 16 arduino pins.

I have a spare arduino, is there a way I could make a scope out of an arduino?

See if this works for you.
http://www.negtronics.com/simplot

but I have no idea how to test the theory.

You don't need to: the datasheet lays it out clearly for you.

The typical diodes have a log-linear v-i curve, and the typical leds have a logn-linear v-i curve at low current levels and a linear curve at high current levels.

The resistor really is there to provide some negative current feedback to prevent a thermal runaway on the diode.

I take this as an implicit "I'm happy to field all future "my Arduino and/or LED array is knackered because I don't understand current limiting or datasheets because I'm an artist/hobbyist" type questions" from now on, dhenry.

Well volunteered.

You use a lot of words that dont make a lot of sense, are you intentionally being confusing?

It sounds like you thing that someone here hasnt read a datasheet or heard of a current limiting resistor being used with an LED, but im pretty sure everyone here knows about that stuff, so this just sounds like an insult with no merit, why even post it?

Thanks for the link, im going to read up on simplot today.