Connect with us

Strain Gauge or ...?

Discussion in 'Electronic Design' started by Jack B, May 29, 2004.

Scroll to continue with content
  1. Jack B

    Jack B Guest

    I want to measure (indirectly) the force applied to an arm.

    Think "bicycle crank arm"... force is applied at one end of the arm (as
    with a pedal), and the arm pivots about a perpendicular axis at it's
    other end.

    Givens: I may not measure the force directly at the end of the arm;
    environment is mild temperature-wise (humans can function ok), but lots
    of dirt, dust, vibration; sensor must be small and non-intrusive;
    measurements should be good, but need not be lab-quality.

    Two issues:

    - assuming some sort of strain/flex sensor firmly attached to the arm
    (e.g. epoxy), what sort of sensor? Thin foil-type sensors seem to need
    an "S-"bend to function properly; other commercial ones seem too big.
    Would some sort of piezo device work?

    - how to transmit the data to an analysis device? RF?
  2. John Fields

    John Fields Guest

  3. Dan Major

    Dan Major Guest

    Assuming you are measuring the stress/strain in the arm assumes that there
    will be enough stress (load) to produce measurable strain (bending moment).
    If there is enough load to produce a measurable flex, commercial foil-type
    strain gauges should work after a careful selection. Gauges should be
    configured in a Wheatstone Bridge. Full Bridge configuration is best,
    half-Bridge is next best. Gauges are bonded to the substrate by several
    different types of adhesives. Epoxy might work, also a cyanoacrylic (super
    glue) is used. For more information, check out the Omega Instruments web
    site ( or Vishay Micromeasurements
    ( The latter has
    a literature series about strain gauges and load cell design.
    That would probable be the most straightforward method. The problem with
    doing it this way is to counterbalance the weight of the data collection
    system. But there's not an easy or simple/cheap way to get data from a
    rotating assembly (that was not designed for it from the beginning).
  4. Jack B

    Jack B Guest

    Mmm. If I choose piezo, then it could be a/the C element, eh?... as in:

    ....seems good: low cost, low part-count (?)

    Else... as Dan Major has noted, foil strain gauges are usually put into
    one leg of a Wheatstone bridge... and would require some additional RF
    modulation-thing to transmit strain level.

    And... additional client requirements:

    there could be many of these units in close proximity; cross-talk is
    forbidden; units will come and go in no predictable pattern.
  5. Activ8

    Activ8 Guest

    Couple places make preapproved RF modules. FHSS - so more than one
    remote can talk to the master. is one. I can't remember or find my notes on the
    other, so take some key words from AeroComm site and FHSS or
    frequency hopping (or hopped) and google it. The one I can't
    remember sells the modules for around $90 ea in 100 qtys, IIRC.

    Don't assume the data rates quoted are all yours - part of it is
    overhead. Lots, actually.

    Now these buggers might be a bit large. John's got a good idea, but
    don't let that frequency sweep outside the allowed channel bandwidth
    of whatever band it is that you're not licensed for, etc., blah,
    blah, range, interference, band, ISM...

    What I think would be best is one of those RFID systems like those
    used for tire pressure monitoring. Use that to get the data off the
    rotating machinery and somewhere where you have more room for

    Strain gauges: Sensym, if they're still around, makes them. IIRC,
    some of their pressure sensors had built in bringe circuitry so it
    stands to reason that some of their strain gauges do also.
  6. John Fields

    John Fields Guest

    Piezo is _not_ a good idea, since you'll only get an output when
    pressure changes. That is, for a steady-state change in pressure,
    after an initial charge transfer, the output will drop to zero.

    For a piezo film transducer, what you'd need to do would be to
    mechanically deform the PVDF dielectric to get the capacitance change,
    but when you did that you'd get the accompanying electrical output
    from the transducer screwing things up until it settled down and all
    you'd be seeing was the remaining capacitance.

    What I had in mind was more like a temperature compensated capacitor
    with an air dielectric separating two parallel plates which got closer
    to/farther away from each other as pressure was applied to the
    outboard end of the arm.

    That way, if you had a fixed and a movable plate you could have a
    simple screw adjustment which would move the fixed plate to adjust the
    distance between them (and the frequency out of the oscillator) and as
    the movable plate moved (with pressure), the frequency of the
    oscillator would change. Having the screw adjustment would allow you
    to have all the sensors operating in their own bands and thereby get
    rid of the crosstalk problem.


    For something other than a one-off, the (radiating) oscillator scheme
    starts to look less and less attractive as the number of sensors
    increases, if only for the multiplicity of receivers or the spectrum
    analyzer needed to capture the output frequency of all the
    oscillators. Plus, just add all the in-situ calibrations needed for
    the sensors, and it all starts to quickly turn into a can of worms.
    So, for your situation, I'd think seriously about the strain gage
    route, but it's hard to say without more data.

    For instance:

    Are you allowed to multiplex the data from the sensors?
    How good does the data have to be?
    How do you want to drive the sensors? (AC, DC, Mains, Batteries?)
    Do you have to do sensor driving, data acquisition, signal processing
    _and_ data transmission all at each of the sensors, or can you break
    it up into separate physically disparate tasks?
  7. gwhite

    gwhite Guest

    Of course, SRM has been doing this for over 10 years. Post in if you get a working model cheaper than SRM -- SRM
    is very expensive.

    The PowerTap hub is less expensive, but still pricey:
    Go to products|powerTap

  8. Terry Given

    Terry Given Guest

    I have read a couple of IEEE papers (industry apps IIRC) about passive RFID
    tags that use the SAW equivalent of a transversal filter, passively
    responding to coded RF chirps (look ma, no power source). They specifcally
    talked about strain gauges etc. used to passively modulate the "response"

    betcha cant buy them though :(

  9. On Wed, 2 Jun 2004 21:47:19 +1200, "Terry Given"

    You wouldn't have any references to those papers by anychance would
  10. Activ8

    Activ8 Guest

    sounds like the matched filters used for CDMA synchronization, with
    the strain gauge twist (oops, a pun) added. That's a cool trick.
  11. Terry Given

    Terry Given Guest

    yes I do.

    IEEE transactions on Industrial Electronics
    Vol. 48 No. 2, April 2001
    Special section on microsensors and microsystems in harsh environments
    "state of the art in wireles sensing with surface acoustic waves"
    wolf-eckhart Bulst, Gerhard Fischerauer, Leonhard Reindl

    They specifically refer to a wireless SAW based passive torque sensor:
    "mechatronic extension of a tap holder for process monitoring"
    G.Fischerauer, F. Schmidt, M. Voss, R.Bader
    proc. IEEE IECON '98 1998 pp2378-2381


  12. Many thanks, I will try to chase up a copy.
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