varying intensity of led

Discussion in 'General Electronics Chat' started by deb, Mar 25, 2009.

  1. deb

    Thread Starter New Member

    Mar 25, 2009
    Can any one please help me in the topic below............

    I want to find out a circuit by which i can increase the intensity of a led or bulb to maximum and then to minimum with out using any programming logic.

    I am confused that pwm will work well or not.
    please suggest me..........
  2. Audioguru


    Dec 20, 2007
    I made some mood lights many years ago with a two-opamp ramp generator driving a transistor that faded and brightened red, blue and green LEDs. Each ramp generator drove its colour at a slightly different frequency than the other colours so millions of colours and brightnesses shined slowly on the ceiling at night.

    A power transistor like a TIP32 can be used to light powerful or many LEDs with up to 350mA of current.
  3. Darren Holdstock

    Active Member

    Feb 10, 2009
    I had the back of an envelope available and an itch for a quick design:
    The spec is a bit vague, but I'm used to that. Ideally we'd have the load requirements, the frequency and nature of the control signal, and the desired chopping frequency of the PWM signal. The 2-op-amp circuit above should cover most angles. The current through the LED is the voltage across Cf divided by Rsense. Rin and Rh just provide a bit of hysteresis -1k and 1M respectively should work OK if the first op-amp input signals aren't too noisy. Rf and Cf depend on the PWM frequency. The PWM chopping signal is derived from a triangle wave, Vpwm, and should be at least 10 times the frequency of the control signal, Vcontrol. The latter can be any waveform or even DC, and the output is at 0% when Vcontrol equals the lower peak voltage of Vpwm, and is at 100% when Vcontrol equals the upper peak voltage of Vpwm.

    If you've got an analogue control signal available anyway, then miss out the first op-amp and put a control voltage directly into the non-inverting input of the second op-amp.

    This topology will just work; it is, however, the easy bit of the design. Choosing op-amps and passive values is the bit that takes the time; sadly I have very little of that, maybe kind peers will post their takes on op-amp choice, PWM frequencies, and passive values. One thing I should note though is that Rsense should be a low value (<100 Ω is best) or gain peaking will occur as Rsense sets up a pole with the parasitic capacitance of the op-amp inverting input. If the calculations spew out Rsense >100 Ω, then add another couple of resistors to pot-down the voltage from Cf into the op-amp input to bring Rsense back below 100 Ω. Not too low though.

    The second, 1-op-amp circuit is a PWM-controlled current source and is cheaper and more efficient than the 2-op-amp circuit, but don't build it unless you've got everything sussed, or it will just go up in smoke. The 2-op-amp circuit is much more forgiving, and should be hard to destroy if the LED driver op-amp can handle the load current.

    Good luck deb, hope your project is deeply satisfying.
  4. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
  5. deb

    Thread Starter New Member

    Mar 25, 2009
    thanks u all for ur help
    this is a very good site...............
    I will also help others by answering their questions according to my best of knowledge.............