Vibration/Impact Sensor Circuit Question

Discussion in 'The Projects Forum' started by cce1911, Sep 3, 2013.

  1. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    I have built a variation of this circuit. My circuit is the first picture. I have attached the piezo sensor to a plate clamped in a vise. As you can see from the first oscope pic, the piezo behaves as you would expect. That's to say it rings as a damped wave form when I hit the plate with a wrench. The monostable multi-vibrator (M74HC123) works and I've tuned the R and C values to get the pulse width I want (single_hit.jpg). However, the re-trigger feature of the 74HC123 makes the pulse much too long (multi_hits.jpg).

    I want only one trigger per "hit", so how do I suppress the re-triggers? Better yet, how do I suppress any additional trigger for say 100ms?
     
  2. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    You could possibly use the other monostable to disable the primary monostable input for a period longer than the expected mechanical post transient settling time.
     
  3. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    t_n_k,
    Any suggestions on how to do this? I've been scratching my head thinking it might be easier to use mono2 to disable the output from pin 13, but I haven't come up with anything yet.
    -cce
     
  4. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Just make the monostable period 100mS. One "hit" will make a HI output pulse of 100mS duration.
     
  5. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    I got the impression the OP wanted a much shorter single output response pulse.
     
  6. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    The problem is that the "re-trigger" keeps restarting the timer so I get a pulse that is too long.
     
  7. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    How long do you actually want the output pulse to last?
     
  8. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    I only want one ~40ms pulse per impact. The "ringing" keeps retriggering the 74HC123. I've built an LTspice model and approximated the impact with a damped SINE. I tried to use the second mono to turn off the input from the piezo, but with no luck. Next, I tried using 2 LM555's, but I got no where with that either. So, I pulled a 74HC221 out of the LTspice .lib and ran a simulation using it. This looks like it may work. I don't expect the impacts to be less than 150ms apart and I'm hoping they will attenuate faster than that.

    Is there a way to dampen the signal coming from the piezo disc?
     
  9. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Some monostables such as HEF4528 can be configured in non-retriggerable mode.
    However it's still likely you might need some front end signal modification to reduce the likelihood of multiple outputs to a single impact test.
     
    Last edited: Sep 5, 2013
  10. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    The 74HC221 is non-retriggerable, but I'm out of ideas on how to modify the front end signal. If I had an example of a scenario that required chopping a signal after the first half of a wave form, I could research it. Unfortunately, I'm a ME not a EE so my circuit design experience is extremely limited.
     
  11. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    The first challenge is to understand what the likely variation in the transducer output will be as tests are undertaken. Is the mechanical configuration fixed or is it likely to change and thereby give varying responses - in both amplitude and form. You are probably also limited by the low power supply of 3 volt in what other front end active circuit options are open to you.
     
  12. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    There could be a significant variation in the transducer output. The impact is caused by a bullet hitting a steel plate. The bullets will vary in weight and velocity.

    I'm free to use any voltage necessary, the only reason I was using 3 vdc is because that's what the example circuit I found used. I found a few discussions about using piezo transducers to detect impact, but none of them were interested in measuring the time between successive impacts. That's my challenge and why I need to suppress the input signal after the first rising edge.
     
  13. cce1911

    Thread Starter Member

    Jun 11, 2013
    41
    0
    I decided to go with the 74HC221 and skip modifying the front end signal. So far, it seems to be working.

    Would anyone suggest what type of capacitors would work best in this application?
     
Loading...