Variable width 5V pulse needs to turn LED on.

Discussion in 'The Projects Forum' started by ProZac, Nov 23, 2010.

  1. ProZac

    Thread Starter New Member

    Nov 23, 2010
    8
    0
    Hello all.

    Im building an ignition component for a performance automotive application, and have struck a little brick wall. The actual ignition/coil firing side of things works perfectly, the problem is with a diagnostic funtion I need.

    The igniter chip I'm using has a flag that gets set +5v when the current going to the coil reaches 4.5A. The IC itself will provide up to 8A of coil current, so the width of this flag pulse can vary depending on how much current is allowed to build up in the coil (which depends on the dwell time, the duty cycle of the ignition pulse).

    I need this flag to turn on an LED as soon as it appears (when the current allowed to build in the coil reaches 4.5A). The LED needs to stay on after this point, aslong as the flag is regularly present... i.e the actual value of coil current isnt important, just that its being allowed to increase to more than 4.5A.

    Ive tried a couple of things with an LM741, using it as a voltage buffer, with a diode to charge a capacitor quickly, which then discharges through the LED when the flag is no longer present... But am not having much luck,

    Any help would be greatly appreciated.

    Cheers all :)
     
  2. ProZac

    Thread Starter New Member

    Nov 23, 2010
    8
    0
    Monostable 555 Timer to the rescue, problem solved :).
     
  3. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    A 555 timer set up in a "Pulse Stretcher" configuration would be the most simple and direct.

    The "advanced" version would be using a microcontroller to light the LED when the signal is received, but also keep track of how often it "misses" for measured RPM, etc. That would be a rather complex project though, but this is how standard auto ECUs work for internal diagnostics while running (accessed through OBD port)
     
  4. ProZac

    Thread Starter New Member

    Nov 23, 2010
    8
    0
    Cheers.

    Yeah, 555 pulse stretcher is the route I've gone. The PCB the project is on has to be pretty compact, so trying to keep complexity (and cost!) down as low as possible.
     
  5. wayneh

    Expert

    Sep 9, 2010
    12,087
    3,027
    You could have considered discahrging the cap through a higher resistance, if you had a comparator already on board. Watch the voltage on the cap against a fixed reference, turnig the LED off after a delay. If you'd have to add a comparator anyway, you're better off with the timer strategy.
     
  6. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    I found this drawing in bottom of stack- forgot where it was to go. In using a 555 one-shot timer there will be some resets when 555 times out & next input arrives. As shown an invertor allows triggering on leading edge of input pulse & also provides a nibbling away of the charge on timing cap via Rx, about 10k, adjusted to just not allow reset at lowest RPM. A second 555 connected to pin 3 of first 555 would show presence of missing pulses..
     
  7. ProZac

    Thread Starter New Member

    Nov 23, 2010
    8
    0
    Thanks heaps for that, why didn't I think of that! Very helpful, was looking for a way to keep the LED on constantly, without having to wait for the next pulse to re-trigger.
     
  8. ProZac

    Thread Starter New Member

    Nov 23, 2010
    8
    0
    Update for anyone who finds this thread in a search for something similar, a still yet cheaper and wasier solution is to use an 'IttyBitty" timer chip, as it is only setup for monostable mode.

    http://www.micrel.com/_PDF/mic1555.pdf
     
Loading...