Constant Duty Cycle Oscillator

Discussion in 'The Projects Forum' started by Ben.B, May 16, 2011.

  1. Ben.B

    Thread Starter New Member

    May 16, 2011
    4
    0
    Hi, apologies in advance if this has been covered to death before, but some light googling and forum searching returned no results

    I'm looking to build a circuit that will drive several high-bright leds at adjustable frequencies (0.5-~100hz) with a constant duty cycle of around 50-60% (square wave). I tried messing around with a 555 timer on the Falstad circuit simulator driving a mosfet/transistor to drive the leds, but when you adjust R1 and R1 to alter the frequency, changing the duty cycle seems to be unavoidable.

    Is there an alternate, more suitable oscillator for this application? (I'm open any simple solution, my knowledge of ICs and the like is fairly limited!)

    Many thanks
     
  2. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    If you are content with 50:50 duty cycle then you can use any oscillator and run the output through a "divide by 2" flip flop. This will give half the frequency out, and exactly 50:50 duty.

    If this is for experimenting, I would suggest a signal generator that has a control for frequency and a control for duty cycle. Most of the decent sig gens do this.

    Otherwise you can use a 555 and a couple of pots, and manually adjust both freq and duty. It helps if you have a 'scope to see the output. :)
     
  3. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
  4. Ben.B

    Thread Starter New Member

    May 16, 2011
    4
    0
    Hi RB, thanks for the reply, is there any way you could give a more worked example of using the flipflop? This is getting way out of my experience here :p

    Thanks :)
     
  5. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    A CD 4013 is two of them in a package. Clocked flip flops do a ÷2 on the frequency, but the cycle is perfect.

    Just curious, did you even look at the link I gave? A conventional 555 doesn't do 50% well with a hysteretic oscillator, but the CMOS version does, and is uber simple to build. One variable resistor, one capacitor, one chip.
     
  6. Ben.B

    Thread Starter New Member

    May 16, 2011
    4
    0
    Edit Edit: I'm referring to the hysteresis 555 here (I'm far too used to facebook live-refreshing!)
    Bill that is absolutely perfect! Problem solved, I'll mock up a circuit later. One thing I just noticed when simulating the circuit, at the higher frequency end a very small adjustment of the pot ends up causing a massive change in frequency. Is it possible to buy pots that are logarithmic, such that it takes more travel/turns to increase the resistance from 0?

    Edit: Sorry, I replied without refreshing the page and as such didn't see your reply until now :(
     
  7. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    About the pots, yes. They are called (wait for it) logarithmic pots. Audio volume for humans is also logarithmic, which is why they exist.

    You probably should stick a small resistor to prevent the resistance from going to 0Ω. The small resistance will set the upper frequency limit and prevent the 7555 or TLC555 (the latter is Radio Shack's answer to a CMOS 555) from trying to get hot. A standard 555 will get hot.
     
  8. Ben.B

    Thread Starter New Member

    May 16, 2011
    4
    0
    Thanks again Bill. You learn something new everyday eh?
     
  9. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    You are welcome. It's why I write the articles. That one has been in volume 6 for several years already.
     
Loading...