Connect with us

Figuring out LED specs

Discussion in 'Electronic Basics' started by Mike Chambers, Jan 20, 2005.

Scroll to continue with content
  1. I have a bunch of loose LEDs. Some of them are completely clear. Is
    there anyway that I can figure out the specs for the LEDs? I no longer
    have the packaging?

    i.e. how many volts they use, and their current?

    I am afraid if I hook them up to check them out, I will burn them out.

    mike c
  2. First you have to sort by color. Put the leads directly across a 3V
    lithium watch battery (no resistor) which will safely light red LEDs,
    and won't kill them if you reverse-connect them. If it stays dark, stack
    two lithium cells for 6V; same lack of reverse-voltage concern. Been
    sorting LEDs this way for some time. As for ultimate current limits per
    device, well, suck 'em and see.

    IRLEDs will be a tad more difficult...

    Mark L. Fergerson
  3. Leon Sorokin

    Leon Sorokin Guest

    actually if you put 3V across a red LED, you'll probably fry it. most reds
    are 1.7V, high brightness are 1.9V. amber, and green are around 2.0-2.1,
    whites and blues are 3.3-3.6 usually.

    as long as you get some current through the LED, u'l be able to tell what
    color it is. 3V will destroy or damage anything other than blue or white. to
    be safe, test everything with a 1.5V battery or a well drained 3V. most LEDs
    wont be full brightness at 1.5V, but 85%-90% will at least glow.

  4. You've been lucky. Without current limitation (perhaps your battery was
    weak), pretty much any LED will burn out soon after you exceed the Vf
    and get it turned on. Granted that a quick hit probably won't do it
    (I've done dangerous things too ;-) I do find that many LED's are quite
    durable to this kind of thing, but there are plenty that will just go
    poof in the blink of an eye.
    AFAIK, 6V is likely to destroy just about any LED without some kind of
    current limiting. Most LED's will light dimly if you supply as little
    as 1 or 2mA.
  5. The right way to do this is to send a specific current through the led,
    and then measure the voltage over the led.

    5mA is a safe current.

    Use a voltage source which gives significantly more volts than the led.

    So a 6 - 12 volt DC source is suitable.

    Calculate the resistor you need to put in series with the led to set the
    current to 5mA.

    Assuming a 12 V supply, 12 minus the led voltage which is max 3 or so,

    9V/0.005=1800 = 1k8 resistor.

    Ok, so you should connect the power supply 12 V in series with a 1k8
    resistor and put the led you want to test in series with the 1k8

    Then measure the voltage over the led. Check visually if it lights up and
    in what color.

    You have built an "unknown led" tester, from only a battery and a
    resistor, and a voltmeter.

    What happens if you happen to connect the led backwards?

    The led has a back voltage and you will find out in a safe way how much
    it is. The 1k8 resistor limits the current to 5mA in both directions.
    The led does not work as a lamp when turned the other way around.

    So you can try each led twice, once in each direction.
    That way you find out the polarity of the led too.

    If you want to use a higher current, like 10mA, that is okay.
    Most leds work well at currents up to 20mA.

    To raise the current to 10mA you lower the resistor to 1kOhm
  6. Guest


    You've gotten some conflicting advice here. Just wanted to add another
    vote to the ones who are saying "put a current-limiting resistor in
    series with the LED". Do not hook it up directly to a battery or other
    voltage source without a resistor.

  7. Dumb Luck has always played a large part in my life; I always use
    batteries from "dead" or otherwise defective watches (I suppose I might
    have mentioned that), so maybe you're right.
    Haven't had any die from testing so far.
    Again, Dumb Luck on my part. I'll go with the other suggestion about a
    current-limited supply and voltmeter if the OP can manage it, but I felt
    comfortable recommending watch batteries for quick 'n' dirty sorting
    since I inferred the OP was more interested in not throwing the unknowns
    away than assembling "proper" test gear. Might be wrong again...

    Mark L. Fergerson
  8. For decades, I've had a 9volt battery with a 1K resistor soldered to one
    terminal. The resistor limits the current, and the lead of the resistor
    makes it easy to hook up the LED.

    It's not only handy to make sure the LED is working (or at least a visible
    light LED), but handy to get the polarity right.

  9. Me too.
    Like I said, you've been lucky. I've killed a few small ones by just
    juicing them for a split second with 5V.
    The only problem is that there is risk involved when not using a

    have fun
  10. Mark mentioned using a lithium. They aren't going to deliver enough current to
    fry much of anything.

  11. Have you looked at the battery specs for a CR2016 or CR2025 or CR2032?

    I don't think it would be possible to burn out a red LED with one.

    Alkaline? Yup -- that would be bad. But these lithium button batteries used in
    watches just aren't designed to deliver appreciable currents.

  12. If you are willing to use 6 AA alkalines, you can arrange things so that you use
    a current limited driver:

    ,-------+---------- VP2 (+)
    | |
    | --- ~ LED
    --- \ / ~ under
    B1 - --- test
    6V --- 4 |
    - AA +---------- VP2 (-)
    | |
    | |/c Q1
    +-----| 2N2222
    | |\e (NPN)
    | |
    --- +---------- VP1 (+)
    B2 - |
    3V --- 2 \ R1
    - AA / 120 Ohms
    | \
    | |
    '-------+---------- VP1 (-)

    If you now place the LED into the circuit, it will limit the current to it.
    Typically, below 19mA. If you now measure the voltage for VP1 (setting your
    test leads as indicated) you can compute the current through the LED as:


    If you measure the voltage for VP2 (again, as indicated), you will have already
    measured the operating voltage of your LED. It's just VP2, at the above
    computed current.

    The 6V of battery B2 should allow this to work for a wide variety of LED colors.

    You can adjust the current to something else by changing R1. For example, by
    using a 100 Ohm for more current or a 180 OHM or 270 Ohm for less current.

    You can also achieve something very similar with a single 1.5 AA alkaline
    battery, using a hand-wired transformer (easily done with a couple of feet of
    magnet wire on a tiny toroid) and two resistors and a capacitor and a simple
    1N4148 diode. If you are interested, I can post that, too.

  13. Forgot to mention that it also needs the NPN transistor. Oh, well.

  15. The CR2016, according to Energizer's data sheet, is about 38 Ohms rising
    offscale past 120 Ohms shortly after 70mAh of use. They point out that it can
    be used in pulse applications and supply about 6.5mA.

    Assume that 38 Ohms is fixed, independent of current being supplied (it's not,
    it gets worse when you pull more), then if your LED needs 2V and the battery is
    supplying a peak of about 2.9V (about right for most of its life) you will get
    about 0.9V/38ohms or slightly less than 24mA. Which is about right for 2V on a
    red LED (the old style red LED is about 1.55V+21*I, which in this case is about
    2.05V at 24mA -- about 2V.)

    I doubt the CR2016 would supply even that 24mA, though.

    Now, if you look at the fatter CR2032, it's all the same. The spec shows about
    38 ohms, again. They suggest a slightly higher pulse current as an example,
    7.3mA, but the device is otherwise very similar except for its greater energy

    None of these things deliver the energy very quickly.

    The 2CR5 you mention is NOT a button battery. And it is designed to deliver a
    20mA continuous current, an ability to actually provide 1.5A continuous, and
    with peaks of up to 3.5A!! -- as compared to a button battery's design for about
    0.1mA and ... maybe up to 0.5mA. That's quite a healthy difference. Notice
    that the 2CR5 weighs some 38 grams as compared to about 2 grams for the button

    Comparing those two families is TOTALLY crazy-minded!!

  17. Another simple scheme, if you can get a jfet, is to use it as a current

    | |
    | |-+d
    | | N-Channel JFET
    | .-g->|-+s
    | | |
    | 9V | .-.
    --- | | | R
    - | | |
    | | '-'
    | | |
    | '------o
    | |
    | .---------.
    | | |
    | | Your |
    | | LEDS |
    | | |
    | | |
    | | |
    | '---------'
    | |
    (created by AACircuit v1.28 beta 10/06/04

    The JFET automagically limits the current. Start with a 10k resistor,
    work your way down to something that makes a normal red LED glow

    N-JFET with a Vgs(off) of between -1 and -8, like a BF245A would work
    well. Heck, with that JFET, connecting the gate to the source would
    limit current to max 6.5mA, since that's the Idss. Using a resistor is a
    good idea, however, with different JFETs.

    Robert Monsen

    "Your Highness, I have no need of this hypothesis."
    - Pierre Laplace (1749-1827), to Napoleon,
    on why his works on celestial mechanics make no mention of God.
  18. Now, do all that with the JFET using only a 1.5V AA battery, two resistors, a
    capacitor, an 1N4148 diode and a simple, hand-wound toroidal. ;)

    V+ V+ V+
    | | |
    | | |
    | )| .|( (about 50" of
    | )| T1 |( magnet wire
    | )| |( for both
    | + )|. |( windings)
    --- ,' |
    - B1 | | D2 R2
    --- \ +---|>|---+---/\/\---,
    - / R1 | 1N4148 | 220 |
    | \ 2200 | | |
    | / | | |
    | | |/c Q1 | --- ~
    | '--------| 2N3904 --- C1 \ / ~
    | |>e --- 100uF --- LED
    | (B1 > 1V) | | |
    | | | |
    gnd gnd gnd gnd

  19. Guest

    I agree with Jon, some time ago I worked with some Red, white and blue
    LEDs using CR2016 and CR2032 (very common batteries in LED torches) The
    batteries won't fry an LED as the batteries can't supply enough

    It's common to use the CR2032 for red ones, and two CR2016 for white
    and blue ones, I found though that the last option will end with the
    LED's life faster.

    It's important to remember the LED's are current divices, so I think
    that the simpliest thing is to use a resistor to limit the current
    below 20mA.
  20. Here's a circuit for testing LEDs that works very well.

    Use a +5 volt source. Any power supply will do the trick.

    Connect a 150-ohm resistor in series with the LED.

    If you look at an LED from above, you'll note the circular shape; the flat
    spot indicates the cathode.


    Cathode of LED to - terminal of power supply.
    Anode of LED to 150-ohm resistor.
    Other end of resistor to + terminal of power supply.

    Turn on power supply, and you're in business!

    If you get an LED that won't light up, there's a possibility that it's an
    infra-red LED and its light won't be visible. Some video cameras can see
    it, though, so try looking at it through one.
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day