Hysteresis on Buffered Comparator Input Signal

Discussion in 'The Projects Forum' started by OdysseusWorm, May 28, 2013.

  1. OdysseusWorm

    Thread Starter New Member

    May 28, 2013
    Hello, thanks for taking the time to read this.
    I am a newb and could use a tip if someone's already figured this out or knows a better way.
    I'd like to use an RC timer to switch a relay. I need hysteresis to prevent short cycling, but I can't have the hysteresis charging the cap back up. I ran the cap voltage through a buffer to protect it, which behaves perfectly at node 'a'. (besides the short cycling without hysteresis, the circuit does the right thing if node a is shorted to node c - straight comparison).
    The comparator op-amp also behaves correctly at node b.

    My problem is that is seems like the comparator input signal dons crazy-pants in the hysteresis pot (tho I made it a very large resistance to keep the op amps from 'fighting' each other). Erratic as heck, doesn't sink down all the way....

    Any ideas smarter than mine?
    Is there a clever way to put the hysteresis on the threshold pin instead?

    -the cap on the 'threshold' voltage is because I had the 12v supply rail switched by the relay coil itself (latches in running mode) and i didn't want the threshold voltage going nutso when the comparison resulted in killing 12 v. Just for debugging, I've got the 12v supply permanent to eliminate that cause.
    -I'm using LM324 quad op-amp for buffer and comparator, though I tried using a 311P comparator (buffered output) instead with same result.

    thanks very much for any tips, sorry if there's already a well-known solution.
  2. OdysseusWorm

    Thread Starter New Member

    May 28, 2013
    I read the material regarding op amps, comparators, and the schmitt trigger on this awesome website called allaboutcircuits, and found the answer.
    I didn't understand the feedback, which is what I was seeing destabilizing the circuit (you should've heard that poor little relay :).

    By leading the input signal to the inverted input and the hysteresis arrangement to the non-inverting input, the amplifier stabilized due to positive feedback, and the high output controllably bumped the threshold on the non-inverting input (thereby immediately saving the comparison from noise).
    Since the schmitt trigger now output a High when I needed a Low for the relay, i just inverted it at an unused channel of the 324 op amp.
    Maybe not elegant, but functional. :D
    I could even use the relay to kill the 12 v supply without causing another feedback issue (cause the relay is mechanically slow enough)! - I didn't like the drain on my truck battery.

    The fixed circuit is attached.
    - hope this helps some other newb like me.

    P.S: read the website.