Connect with us

shock detection

Discussion in 'Electronic Design' started by Michael, Nov 12, 2007.

Scroll to continue with content
  1. Michael

    Michael Guest

    I need to find a way to detect whether our hand held instrument ($25k-
    $40k) was abused (dropped, bumped, etc).
    It means free fall and shock detection (detect, timestamp, go back to
    sleep). First one is relatively simple. "Relatively" because this
    shock_detection_system needs to rely on it's own battery - the
    instrument may be dropped when main battery is out and this "battery
    out" state may be as long as two months. And occupy no (ok, very low)
    Shock detection is what I need help with. I found out (http:// that 25G is
    "small" shock. This gives some clues - shock detector will not
    necessarily be able to see free fall - two different devices - ???
    "...You have a tough job ahead of you. I know of no technology that
    will get you the kind of low power performance you want without a lot
    of pain. We do not offer anything like that...." - response from the
    accelerometer manufacturer tech support.
    I have EE background and my knowledge of physics is limited.
    To add insult to the injury: we do not know how much shock our
    instrument can sustain without losing performance.
    Can anybody point me in the right direction?

  2. Jim Thompson

    Jim Thompson Guest

    Just measure the deformation of the case ?:)

    ...Jim Thompson
  3. Why would you care about free fall? It's not the fall that kills your
    instrument, it's the sudden stop at the end.
    It's not necessarily an appropriate job for electronics.

    Cheap zero-power technology:™

    Best regards,
    Spehro Pefhany

  4. How do you plan on measuring free fall? This is probably impossible for
    you. Netwon and Einstein say that free fall is an inertial frame of
    reference so there is no method that can measure it directly. Any method
    doing so must use something about the surroundings(which almost surely = air

    In any case if your handhelds cost $25k+ then I'm sure you make enough money
    so that you can hire someone to try and solve the problem.

    As far as impact, which seems to be what you actually need, take a look at

    But again, if your selling a "handheld" for $25k+(what, is it made of solid
    gold?) then you can afford to hire a real scientist to do the work(or are
    you that "scientist" that was hired?).
  5. John Larkin

    John Larkin Guest

    The Brat dropped her Dell laptop yesterday, and lunched the hard
    drive, including a powerpoint presentation that took 8 hours to do and
    is due today. She called for help. I told her to head for Wal-Mart,
    get a new pc, and start over. They had DElls and a Sony Vaio
    available, and called me from the Wal-Mart and asked me which to buy.
    The Sony of course. It has mems accelerometers inside that detect
    zero-G and retract the disk heads before impact.

    1G = not falling

    0G = falling

  6. Calibrated glass tubes are often used - if they break, there was a shock of
    at least X g.
    Either measure it - or don't worry about it because then it is not
  7. Michael

    Michael Guest

    We need to have a time stamp (to be able to properly assign
    blame ;o) ) - glass tubes won't do.
    I do not know:
    a. "how much is too much?"
    b. How to keep this system running for, say, two months between
    recharges on whatever tiny volume available for the battery. The
    minimum I will need (I have found so far is) ~1..1.5Amp-hour cell.
    It's not tiny... :eek:(
  8. John Larkin

    John Larkin Guest


  9. D from BC

    D from BC Guest

    You just need a loose wire.

    The Fonz would punch the side of a jukebox and it would start working.

    D from BC
  10. Guest

    <rolls eyes> You are quite serious when you say that?
  11. Tim Wescott

    Tim Wescott Guest

    There are some large-format piezo accelerometers out there that appear
    to produce a big enough signal to turn on some electronics -- I did some
    preliminary work on this for a prospective customer. That takes care of
    the "big shock" case.

    For free-fall, perhaps some self-centering tilt switches?

    Instead of consuming power all the time, you can get this down to
    consuming power for each event -- but I'll bet a hand-held instrument
    sees a lot of free fall without seeing consequent shock; you may run out
    the battery from all the "whee!" events then miss the "thump!" that you
    wanted to capture.


    Tim Wescott
    Wescott Design Services

    Do you need to implement control loops in software?
    "Applied Control Theory for Embedded Systems" gives you just what it says.
    See details at
  12. Sure-- you've failed to consider that the end user may be attempting
    to use the product in, say, a space station, so the 0g state is
    indistiguishable from normal operation.

    Best regards,
    Spehro Pefhany
  13. Tim Williams

    Tim Williams Guest

    Alright, so use an opto detector of some sort that watches the glass! Or a
    contact (metallize the glass tube?), or who knows what else... ;-)

  14. John Larkin

    John Larkin Guest

    But then, if he drops it, nothing breaks!

  15. Guest

    Argh, beat me to it! But there's really no "dropping" on a space
    station, is there???
  16. Ecnerwal

    Ecnerwal Guest

    Well, the disgustingly easy method won't do timestamps. It is however
    small, cheap enough to stick on a cardboard box, and reliable - with no
    batteries at all.

    This will monitor, but could miss an event depending on logging
    interval, which trades off with battery life. Could probably arrange to
    OEM the guts to cut the size down (take out of waterproof package). Also
    rather low-range for you, I guess (+-3G)

    A bit large for you, but otherwise seems to do what you want:
  17. How bout springs? F = kx, mount two opposite springs(actually 6) on a ball
    then any accelleration will compress one and extent the other... the amount
    of acceleration being proportional to the acceleration... i.e., F = ma = kx.
    (with 3 axis you'll need a better algorith as it would depend on the
    orientation(in this case it does too but I'm assuming lateral motion)).

    I'm sure there are a bazillion ways... In fact somehow they do it with very
    small chips using ingenious methods. I'm sure it can be done for a few

    Knowing the minimum high at which the device could potentially break would
    help. If its in free fall(and not thrown) from 1 feet then it will take
    about 0.35 s to hit the ground so one does not actually have to sample at a
    high rate(well, depends on what this guy wants to do).

    But seriously, I don't see what real damage could come from a handheld made
    of solid gold. What if it is dropped? Why does it matter? Does the gold
    really know its in free fall?
  18. (I'm not actually saying this is a good method... I imagine it would work on
    a large scale with precision components but I'm sure it probably doesn't
    beat your standard accelerometers in cost, size, and accuracy)
  19. Eggs are a good source of accelerometers too... although they don't actually
    test acceleration ;/
  20. It's sort of in free-fall continuously.

    Best regards,
    Spehro Pefhany
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day