How to use a random LED?

[ADWSystems]

How do you determine what current a "random" LED should be used at? By "random" I mean some unmarked LED from a bag or drawer full of LEDs that came from who-knows-where.

 

[ronsimpson]

Post a good picture of the LED(s). What size?
https://www.digikey.com/products/en...t=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25
In the above search we are looking for "LED through hole" not surface mount.
Example: a 3mm diameter LED gets you into a certain type of LED. By playing with the search categories you can remove all that are very different.
This is not accurate! But better than a guess.

 

[ADWSystems]

One of these?

 

[dknguyen]

destructive testing. hook up to a 3Vish source through a pot and turn down the resistance while measuring current and diode voltage drop until you reach desired brightess, too hot, or failure, whichever comes first. if failure, halve that current. if too hot, dial back until it's not.

 

[ronsimpson]

LEDs come in many types. (surface mount in your case)
A simple red LED probably need a voltage in the 1.5V range, while a white LED want about 3 volts.
Do not put a voltage across a LED. They are a "voltage" device but a "current" device.
Because some of you LEDs probably need 3.3V you need a voltage source greater than 3.3V.
I use a current limited power supply. I set the voltage to be 4 volts or less and the current to limit at 10mA for a first test.
If you don't have a current limiting supply you can use a 9V batter and a resistor. A 1k resistor will draw only 9mA if the entire 9V was across it. If the LED is white them it will use up 3 volts of the 9V battery and the 6 volts will be across the resistor. (6mA)
This will tell you which lead is + and -, and what color. For some of you LEDs 6mA is small but it won't burn them out.

You might look at this LED. It is almost like some of yours.
https://www.digikey.com/product-det...-inc/LR-T67F-U1AA-1-1-Z/475-3122-1-ND/4899557

 

[dr pepper]

Without a part No. your not going to find out without spending more time than they are worth.
However instead of binning them you could just assume 10ma draw, test one of each type you want to use for an hour at that and if it lasts just use it at 10ma.
Dont bother using them in new gear or something your going to sell, get new ones for that.

 

[ADWSystems]

Ron,

Thank you but I do know how to integrate an LED into a circuit. What I'm trying to figure out is for a random LED, what the Vf and If are without ending up going dknguyen's route and blowing up the LED.

 

[dr pepper]

Ok your not going to find out If without the manufacturers datasheet.
Bung a 9v battery with a 1k in series on the led and then put your meter across it, the measured voltage across the led is Vf.
If you really want to know If then you'll have to get some new leds, 10ma is usually enough for reasonable brightness from an smd led, spose for some unlikely reason you put 10ma through a led meant for 1ma and you blow it, your no worse off.

 

[ronsimpson]

Vf is much related to color. I would use a 9V battery + 1k resistor to get a idea of Vf and color. The Vf should be measured at If but you don't know what it is. Vf at 6mA and Vf at 20mA is about the same.

If is very much a guess. I look at the case size and what I can find in a catalog. If you can find a maker's name or the case size then this will get you close to the If. Then you can come back and measure the Vf at 20mA or what you think the If is.

edited.
You can apply 20mA and 3.3V and then measure the dye temperature. (hard to do)
edited + edited
A LED is very much a Zener diode. A unknown Zener is measured by applying a current and seeing what voltage you get across the diode. In one direction you get the 0.7V and the other direction you get 5.1V. Now you know. Measuring current (wattage) is hard. You apply power and see how hot it gets. Example: 1 watt burns your fingers but 1/4 watt seems fine. So you don't really know the wattage but it is not 1 watt. Also look at case size.

 

[audioguru]

The color of the LED light determines its forward voltage. A red, orange or yellow is 1.8V to 2.1V. An old green is 2.0v to 2.2V. A modern green, blue or white is 2.9V to 3.6V.
An LED is not a simple hot piece of wire, they cannot make a bunch have the same forward voltage.

 

[alec_t]

If it were me, I'd accurately measure the LEDs and then check reputable component vendors' web-sites to try to identify same-sized and same outline LEDs with known characteristcs. That should get you into the ball-park for maximum current rating. Forget about Vf (but it's already been described how you can find that if you really need to), since LEDs should be current-driven, not voltage-driven.

 

[MikeMl]

I use a 100 Ohm resistor in series with the + terminal of a dual-meter (V/I) lab supply set to an open circuit voltage of 5.0V.

Connect the LED between the free end of the resistor and the - terminal of the power supply. I am assuming you do not know which lead on the LED is the anode, and which is the cathode.

If you get lucky, and the LED cathode is connected to the supply -terminal, the LED should light with reasonable brightness. If it doesn't light, try reversing the LED's leads. If the LED still doesn't light, but the supply shows some current, then the LED is shorted, or it may be an IR LED. The reason I set the supply to only 5.0V is because the reverse breakdown of most LEDs is ~5V, so you are less likely to blow up a reversed LED.

With the LED illuminated, check the supply's current. Say you see 20mA. This means that if you want to use the LED as an indicator, then it needs about 20mA to get that relative brightness. You can also estimate Vf from this measurement.

If it is too bright, you might be able to reduce the current. Try it by momentarily reducing the supply voltage to less than 5.0V. Say that the LED is still bright enough at 5mA s displayed on the supply's current meter, then now you have an idea of how to calculate a suitable current limiting resistor if the LED is to be used as an indicator from a higher or lower voltage.

Further reduce the supply voltage to where the LED just barely lights, and that is good measure of Vf.

 

[dknguyen]

Ron,

Thank you but I do know how to integrate an LED into a circuit. What I'm trying to figure out is for a random LED, what the Vf and If are without ending up going dknguyen's route and blowing up the LED.

You just do it like how I described. That's why I said turn down the pot. You start at a high resistance where nothing will blow and slowly turn it down. Nothing says that you have to turn it down until the LED is destroyed.

 

[JonSea]

All of the above aside, 10 - 20 mA is going to be safe for most small LEDs. Use an LM317 to make a constant current supply of 10mA, connect an LED directly across it and measure Vf.

 

[Rich D.]

That's what I do. I have a small plastic project box with a pot, banana jacks and an LM317. I connect it to a power supply set to whatever volts - sometimes 5 to 10 volts. It doesn't matter as long as it's a few volts above your Vf because the current regulator will adjust for it. I connect a "random" LED, and turn up the current until I see a decent amount of light. Usually that's 5 to 10 mA. Virtually any LED will illuminate with 10mA. I used to have a current meter in series with the power, but then I used a Sharpie° to mark the currents on the pot's dial.

So whatever the current is - that's your If. If you want a brighter LED, turn it up to a higher current and call that your If.
Vf is whatever you read across the LED with a voltmeter, related to that If (current).

There is no one true If or Vf...it all changes depending on the current you select and how bright you want the LED vs. how long a life you want that LED to have.
A well-spec'd LED will have a Vf/If curve - and you can select any point on that curve and get your numbers. For example, a static-powered LED will use one Vf/If pair, but an LED designed into a multiplexed array will need a different Vf/If because it will only be on for say...1/8 of the time, therefore it needs to shine 8X brighter during that brief time. That LED would need perhaps 100mA, and in the period of its illumination, it might show a Vf that is 20% higher.

The rest is up to you...how long that LED will last before it burns up depends on how bright you want it. I assume you can live with a possible failure, or you would spend the 10 cents for a spec'd component.
It's not much different from a car's RPM limits. How fast do you want to go vs. how long do you need to keep the motor running? Higher revs will shorten the life of the motor. Eventually you'll reach a point that the motor will blow up instantly, but there is no single correct RPM to run it.

 

[k7elp60]

I agree that trying to match LED's can be a problem. I have been doing projects for a long time. I can remember when blue and white LED's were not avaiable in the consumer market place. I have been able to connect LED's in parallel on a number of projects. I match the forward voltage within about 50mV at the desired current. The select the current limiting resistor depending on the # of LED's and the supply voltage.
I am uploading a schematic of a LED matcher I recently built. In the schematic there is a pot that I adjust the desired current. This pot is a multi turn 1K. The momentary push button has some scotch tape to hold it down while I was taking the pictures. The push button was added as a safety feature. The black & red tip jacks above the display were added to confirm aqccuracy of the display.
The pictures show the current of 1.49mA and a Vf of 2.66 and a current of 9mA and a Vf of 2.88V. I am also posting a picture of a BOO badge that flashes and aq picture of a cross that has 12 blue LED's in parallel.

 

[Rich D.]

That is one fancy-*** LED tester! Much cooler than mine. But mine does have 4X AA batteries for the option of portability and I'm guessing yours doesn't, so my tester's performance can be outstanding in the field.