LED Flasher

Discussion in 'The Projects Forum' started by skyler521, Mar 14, 2008.

  1. skyler521

    Thread Starter New Member

    Mar 14, 2008
    2
    0
    I am trying to make a LED flasher that will flash a set of 6 LEDs on one side and then flash 6 LEDs on the other side. i want the first side to blink twice then the second side to blink twice. (left left right right LL RR LL RR). i have looked for this kind of circuit and made about 4 of them and none of them have worked properly.. the closest i got was a circuit that would just flash one time on each side.

    I am using 10 mm LEDs that need about 4.5 volts to run properly...
     
  2. timc4444

    New Member

    Mar 13, 2008
    5
    0
    I believe a PIC Controller is appropriate for this. There was a guy on You tube that made a arrowstick that he had programmed various patterns in.
    posting the mesage he sent me below

    The design is pretty simple.

    The light is made up of 8 light heads. Each light head has 9 rows of 4 leds and resistors. So thats 288 amber leds and 72 resistors. The value of the resistors on my light is 10 ohms, but could be differnt for you, depending on your supply voltage and what kind of current your leds can handle.

    The leds are just soldered onto pref board that is painted black.

    The control uses a micro-controller to control the patterns. A PIC 16F628A in my case.

    I used darlingtons to drive the light heads, as a micro-controller cannot handle the current the leds draw on its own.

    A 7805 volatge regulator supplys the PIC with a constant 5v and a LM350 supplies the lights with a regulated 9 volts. The light works off anywhere from 12-15v so it can be used in a car.

    I don't really know what it cost me, as I didn't keep track that well. I would guess around $100.
     
  3. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    You can use a CD4060 oscillator/counter and a gates IC to make each side blink twice.
    You can also use an oscillator driving a CD4017 sequencer and some diodes as gates.
     
  4. Søren

    Senior Member

    Sep 2, 2006
    472
    28
    You haven't looked hard enough, I've posted this link several times: Emergency Vehicle Strobe

    Just use 3 strings of two LED's on each side and you can then downgrade Q1 and Q2 if needs be.

    The current is a much more important measure than the voltage when it comes to LED's.
     
  5. Søren

    Senior Member

    Sep 2, 2006
    472
    28
    Btw. just drop me a mail if it acts up when you try building it.
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    This flasher circuit is similar to Soren's, but has a reduced parts count.

    It's unusual in that a 4070 Quad XOR IC is used not only for the clock circuit, but the MOSFET gate triggering circuit as well.

    As you can see, the basic timing circuit is comprised of the 4070 Quad XOR, the 4017 Johnson counter, to establish the timing, a 150k resistor and a 1uF capacitor. The RC components result in approximately 4 flashes/sec.

    I chose to use power MOSFETS due to their relatively low Rds(on) and resulting low power dissipation.

    Everything else is LED wiring and current limiting.

    Note that in either circuit, you should be certain to use 0.1uF (100nF) ceramic or tantalum bypass caps across the Vdd/Vss pins on both ICs.
     
  7. Søren

    Senior Member

    Sep 2, 2006
    472
    28
    Hi,

    For 3 strings of LED's?

    Time for a reality check:
    With a Darlington BjT and ordinary LED's, the loss amounts to 0.1W
    And since the power need is so low, plain old BC337's or similar will do, reducing the loss to less than 0.02W
    Hardly worth the effort and highly increased risk of ESD induced damage IMO - and even if it was, the IRFI510G's R_ds_on is 0.54 Ohm - quite huge compared to what is considered low these days - lots of better alternatives exists.

    Not using a gate resistor will certainly reduce package count, but it will increase dissipation, if it works at all outside a sim - CMOS4k has a tendency to be unable to charge gate capacitance under less-than-ideal circumstances.

    I would really-really recommend adding in gate resistors (of 10..100 Ohm).


    In my circuit, 1nF to say 2.2nF (or whatever is in the junk box) will do, if the layout is made sloppy and the supply bypass cap is located far from the IC's.
    It has been in use for years in several places around the globe (on fire fighters vehicles), with just the supply decoupling (which you omitted in yours) - proper layout keeps things ín check.
    A value 100nF (or even larger) might be needed if you try to charge a MOSFET gate directly from a CMOS4k output of course, but why anyone would try that is beyond me?
     
  8. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Have you actually built and tested that oscillator? I haven't, but the circuit has negative feedback. I would think it would oscillate at high frequency, if at all. In an LTspice sim, it oscillated at 1.68MHz.
    It needs hysteresis to oscillate at low frequency. You could incorporate the unused gate and make a 2-gate oscillator, but Fairchild says it will not necessarily oscillate.
     
  9. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Here's how to make a clock oscillator with two CMOS XOR gates. U1A is noninverting, and has positive feedback, making it a Schmitt trigger. The second gate is an inverter.
     
  10. skyler521

    Thread Starter New Member

    Mar 14, 2008
    2
    0
    thanks for all your help, i hope to get started on this stuff in the next week, been kinda busy...
     
  11. shimniok

    New Member

    Mar 21, 2008
    4
    1
    For what little this is worth... I did this with one LED per side using Basic Stamp 2 ($$$). I would think this would be fairly trivial with some type of AVR (a tiny?) but need to be handy with C and have the programming env setup and a way to program it (not too awful hard). The logic circuit SgtWookie posted seems easier.
     
Loading...