555 Bistable Random Power-Up State

Discussion in 'The Projects Forum' started by apqo1, Oct 16, 2013.

  1. apqo1

    Thread Starter Well-Known Member

    Oct 5, 2008
    50
    3
    I'm working on a bistable latch circuit to switch a MOSFET on/off with a momentary push-button switch. I've experimented with every BJT and FET-based "soft latch" circuit I can find; some don't work at all and others are intermittent or difficult to debounce.

    I finally settled on the attached 555-based circuit. It works beautifully while powered, and I can't make it misbehave no matter how hard I try to bounce the switch. However, there's one problem: it sometimes starts in the "on" state when power is applied.

    I've tried increasing decoupling cap C2 to as high as 220μF. This improves it a little, but if power is removed for more than ~15 seconds, it will still sometimes start up in the "on" state.

    How do I guarantee that the circuit starts in the "off" state when power is applied, regardless of how long it's been removed?

    Thanks.
     
    Last edited: Nov 25, 2013
  2. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    Use a resistor and capacitor to delay the reset pin coming up to "high" during power up.
     
    KJ6EAD and apqo1 like this.
  3. apqo1

    Thread Starter Well-Known Member

    Oct 5, 2008
    50
    3
    Well, duh! Thanks, #12.

    The 555 datasheet says RESET voltage is 0.7V typical. I put 1MΩ/0.1μF on RESET for ~5 millisecond delay reaching 0.7V at power-up.

    Works perfectly! New schematic attached.


    EDIT 1: Substituted 100kΩ/1μF to achieve the same RESET delay with parts already used elsewhere in the circuit.

    EDIT 2: Added scope capture of the result. CH1 (
    yellow) is Vin; CH2 (blue) is RESET. Nice to see the math work perfectly: 0.7V in 5.2 msec.
     
    Last edited: Nov 25, 2013
  4. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Just using a cap on pin 5 (which is almost mandatory for 555 use anyway to reduce noise issues) could have got you the same result.
     
    apqo1 and #12 like this.
Loading...