Connect with us

Building an LED microflash

Discussion in 'Electronic Design' started by Zephod Beeblebrox, Aug 3, 2013.

Scroll to continue with content
  1. I am having a little fun thinking around corners.

    I have a huge array of LEDs and I want to pulse them on and off quickly (once).

    A trip switch will trigger the LED flash briefly.

    I need help with the circuit design. I'll try to show what I need done...


    The LEDs are all in parallel and are 3v. The input voltage will be from a 3v DC supply (a coin cell).

    A capacitor will charge and a charge indicator light will light up (could be a red LED). The LEDs remain off.

    A switch closes the circuit over the capacitor causing the capacitor to discharge 3v into the LED array which glows briefly. I will have to work out the right capacitor value myself. When the capacitor has discharged, no morecurrent flows across the LED array regardless of whether the switch is nowopen or closed.

    Can it be done simply and can anybody help with a diagram?
     
  2. Please, I'm serious. This is for a personal project. I do need help - real help, not jokey, snooty time-wasting stuff.
     
  3. Phil Hobbs

    Phil Hobbs Guest

    (Context restored.)
    Fair enough, but you're the one who picked the jokey nym, not me.
    Serious folks on SED generally use their real names.

    Putting all the LEDs in parallel will probably work OK for pulsed use,
    though it's fairly far from optimal since there's nothing much to ensure
    equal current sharing.

    Knowing how many "a huge array of LEDs" means would help.

    Knowing how long "briefly" is would also help.

    Knowing what you're actually trying to do, e.g. make something cool
    looking, doing time-lapse digital photography, or whatever, would help
    as well.

    And since you're asking for somebody to design something for you, you
    might be better off in sci.electronics.basics. (I've cross-posted and
    set followups there.)

    It's okay if you don't know the answers in great detail, but a bit of
    context would help us to help you.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs
    Principal Consultant
    ElectroOptical Innovations LLC
    Optics, Electro-optics, Photonics, Analog Electronics

    160 North State Road #203
    Briarcliff Manor NY 10510 USA
    +1 845 480 2058

    hobbs at electrooptical dot net
    http://electrooptical.net
     
  4. Ok. Continuing in sci.electronics.basics
     
  5. Guest

    There should be no problem at all. We manufacture high-intensity white light LED lamps (with simple
    built-in charger) with LEDs all in parallel. The
    parallel connection provides the high intensity.
    Basically, there should be a dropping resistor
    for each LED, and the driving voltage should be
    bit high, say 6 Volts, to take into account the
    voltage drop in the dropping resistor. For added
    protection, add a high Watt resistor that will
    limit the total current into the parallel array.
    That is, suppose 20 LEDs are in parallel, each
    running at the nominal current of 20mA. So the
    high Watt resistor would limit the total current
    that flows into the array, and each individual
    low Watt resistor will limit the current into
    the LED connected to it. The calculations are
    simple. Hope that helps.
     
  6. Martin Brown

    Martin Brown Guest

    How large is huge? In here it tends to mean advert sized display boards.
    How short is briefly? 1ps, 1ns, 1us, 1ms 1s
    Coin cell isn't going to hack it as white LEDs need a bit more voltage
    than that to reach their full brightness. Having the LEDs in parallel
    isn't great either as the weakest one will get most of the current.

    You need about 4v across one to get maximum light output output 20mA and
    for a single pulse of current you can overdrive them somewhat provided
    you don't exceed manufacturers maximum ratings. I'd suggest trying two
    coin cells in series or better a pack of 4 AA cells at 6v.

    See this basic tutorial which may help you
    http://www.electronics-tutorials.ws/diode/diode_8.html

    With pulse drive the lead inductance will save you from disaster. On DC
    the weakest LED will suffer from too much current and the stronger ones
    will be dim. Might not be too bad if they are all the same batch and you
    weed out any obvious stragglers but not in general a good idea to do
    this without a small series resistor in series with every one.
    I'd use a green one - looks brightest for least current.
    But how long does it have to recover between flashes?
    Use a changeover switch so that the capacitor is switched from the
    battery to the LED bank.
    Battery --R-- -----Leds
    | \ |
    | _|_ |
    | ___ C |
    |__________|________|


    Try R=1k and C=1000uF as a first guess.

    The charge indicator LED wants to go across the capacitor with a largish
    series resistor and maybe a diode in series so that it only lights up
    when the capacitor is fully charged.

    If you describe what it is for then you might get a much better answer.
    Ring flash for macro photography designs in amateur electronics magazine
    EPE have much more sophisticated drive, timing and trigger.

    Basically using LEDs for flash requires running them aggressively close
    to their maximum permitted drive voltage and power dissipation.

    These days you can buy single batwing LEDs good for 7W continuous
    dissipation on a decent heatsink.
     
  7. Fred Abse

    Fred Abse Guest

    Count the heads ;-)
     
  8. John S

    John S Guest

    I think that's a bit misleading, John.

    1 joule (a watt-second) _per_ microsecond is 1 MWatt.

    1 joule _per_ second is one watt.

    But, that's just my opinion.

    Cheers,
    John S
     
  9. Phil Hobbs

    Phil Hobbs Guest

    IIRC an actively-quenched flashtube can get down to the few-microsecond
    range. There's an afterglow, of course, but the peak is big enough that
    that doesn't matter much. Doc Edgerton did fine with .303 bullets,
    which are a teensy bit faster than air guns.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs
    Principal Consultant
    ElectroOptical Innovations LLC
    Optics, Electro-optics, Photonics, Analog Electronics

    160 North State Road #203
    Briarcliff Manor NY 10510 USA
    +1 845 480 2058

    hobbs at electrooptical dot net
    http://electrooptical.net
     
  10. John S

    John S Guest

    Yes, but maybe to OP did not. Perhaps I did not phrase my reply in a
    tactful manner.
    I can't disagree with that.
     
  11. whit3rd

    whit3rd Guest

    Your LEDs will take maybe 50 mA for a few microseconds? If there's a dozen of 'em,
    the capacitor will have to supply 50 mA x 5 us of charge, and droop by less than
    or equal to Vmax - Vswitch-Vload after those five microseconds.
    So, it'll take a fast switch (probably NMOS power FET), and a capacitor that can handle
    microsecond pulses.

    A fast switch with higher voltage and lower current (LEDs in series) and an inductor
    to store energy, is the more usual strobe-light design.
     
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

-