Loading ...
Sorry, an error occurred while loading the content.

Capacitance force sensor (was Gripping Fragile parts with Robotic Arm)

Expand Messages
  • Peter Balch
    ... I ve just got the AD7746 evaluation board and it s a very impressive chip. It s quite something to be able to measure femto Farads. You can easily see a
    Message 1 of 1 , Jan 17, 2013
    • 0 Attachment
      > From: "jamericanfreddy" <jamericanfreddy@...>
      > Here is the page for one type ,they do have others
      > https://www.sparkfun.com/products/7902
      > Now to make the sensor you need a 2 tiny copper plates and some soft
      > insulator material to go between them

      I've just got the AD7746 evaluation board and it's a very impressive chip.
      It's quite something to be able to measure femto Farads. You can easily see
      a change when you wave your hand several inches from the electrodes.

      I've made a force sensor with some 10mm neoprene foam. (I'm not convinced
      that any rubber will be good - it's likely to be non-linear, drift with
      time, change under constant load, etc., etc. I reckon coil springs will be
      better. But I've got some neoprene foam to hand.)

      What experience do you have with shielding?

      Clearly you need some. Say you were building a bathroom scale. The
      capacitance of the person standing on top is significant (I can see it on my
      test rig).

      Also, a bathroom scale has, say, four sensors. Did you work out whether the
      AD7746 can support 4 sensors? It has two channels; each channel can be
      differential or you can swich off the "negative" input and make it
      single-ended. But can you switch off the "positive" input to also make it
      single-ended? I haven't found that in the spec sheet.

      (I guess I could measure differential; then single-ended positive; then
      subtract to get single-ended negative but the noise would double.)

      Noise clearly is a bit of a problem with such a sensitive chip. And the
      update rate (60mS) is sufficiently slow that you can't afford to average
      over lots of samples.

      Finally, did you manage to talk to the evaluation board with your own s/w or
      did you build your own board?

      Peter
    Your message has been successfully submitted and would be delivered to recipients shortly.