Why'd my LEDs pop?

Oh, I know. :-) I hope he really did try it for himself. I noticed it quite by accident when I connected a blue LED to the output of an op-amp to clamp the peak output to the Vf of the diode, letting the op-amp limit the max current thru the LED. The op-amp was biased at 2.5V and I was sort of surprised to see the LED start glowing as soon as I plugged it to the breadboard. Not blindingly bright (indicating the obvious self limiting of the current), but clearly on. Just can't understand why Phil has to have a conniption fit over it.

Brian has always been an idiot. Add him to your kill filter, its the best thing you can do.

YOu are correct. This is basic ohms law.

A houshold lightbulb is very different. It is purely resistive (theoretically) piece if wire where as an LED is semiconducter.

OK, the problem with putting a whole heap of LEDs in series is that each one may be slightly different. A small imperfection cause go one to go short circuit then the rest are going to get more current.

Add to this, if your power source varies, then so will the current flow through your string of LED's

Before we can help you with the correct sollution, you need to tell us all details. What is the use? What is the expected power source and how much can it vary (tolerence). What brand LED's are you using and what is there stated foward voltage and current.

"E d" wrote

there are a number of rude posters here, just like anywhere. but anyway....

Brian Goldsmith.

That makes sense. I suspected the power supply was giving a dirty DC signal when testing a voltage reg with it. I tried the reg with a 9V battery and it gave a certain voltage out, with 4 AAs it gave the same voltage but with the DC power pack the voltage was different. The peaks wouldn't have shown in the LEDs as they actually put out less light after a certain voltage.

I've changed the design a bit now so that each LED will have it's own resistor and they will all be in parallel, that way if one dies the others will keep working.

Michael

Its appears to be greater in this group than other's for some reason.

http://217.33.241.242/download/Firstsight%20Vision/Tech%20Tips/LED-lighting.pdf

Thanks, I'll have a look at it now.

Michael

Unfortunately the occasional flashes of brilliance PA exhibits are grossly overshadowed by incessant ramblings of incoherent spasms of abuse and foul language. Obvously there's a wiring mix up in his brain or needs re-programming?

Cheers, Alan

Maybe a "Hard reset" would work? ;-)

Generally he refuses to answer his unit calling if the voice is unfamiliar ... however its not that hard to gain entry if you have time .

the voltage at which they light isn't certain, and varies...

I think I get it now. They are still similar to a standard light bulb in that they use a certain current at a certain voltage but that current changes very rapidly with small changes in voltage. It can vary greatly between LEDs in the same batch. Hence you need to make a circuit to supply a controlled amount of current, not voltage. And a plain old resistor is a good way to do that. :-) I'd always thought the resistor was just to get the voltage down and could have been replaced with a reg.

Thanks for all the replies.

Michael

there are articles on why you should not run a LED of a regulator. LED are supposed to be thought of as having a constant voltage drop across them just like any forward biased diode.

What if your source voltage varies? It is hard to maintain a constant current without using some form of regulation.

Simple: use a LM317 variable regulator as a current regulator. In the NatSemi app notes you can find it, but it works by having a resistor in series with the output terminal that is selected to have a 1.25 V drop at the current you want to regulate at: so 1.25/.02=62.5 Ohms. The adjust terminal is connected to the opposite end of the resistor which is then the output of the current regulator.

This works by having the LM317 keep a constant 1.25v across the resistor which gives a constant 20 mA.

I'm using an lm317 to regulate the voltage.

Have a look in lm317 datasheet at constant current source Tom

This will essentially provide the same result. To provide a seperate constant current source to each LED would be silly, when a resistor in series with a regulated source voltage will do just fine in your application.

You are looking for an explanation... Another potential explanation is the effect of capacitance. Rather than thinking about the static DC operation.... consider what happens when the thing is first turned on.... A mass of electrons fly through the circuit, to build up the operating bias point for each LED, this can be modelled as a capacitance for each LED, but if one has less capacitance than the others, then it will reach this point sonner, and hence become the soul resistive load in a chain of 'shorts' . Of course, this all happens in a microsecond or so.

To the people in this group who are unfortunately exposed to the pointless abuse from others...I suspect they are simply socially inept, oblivious to their human defects seen by others, and continually thought of as sad, angry, friendless people.

They get a cruel sense of fun when abusing / flaming others, so dont bother arguing with them

"Simone Merrett"

This one deserves an honorary PhD in Crapology.

....... Phil

I was considering that but I'm running 12 LEDs in parallel so if one dies the others will get more current.

Michael