shielding load cell signal

Discussion in 'The Projects Forum' started by przemmo, Jan 9, 2009.

  1. przemmo

    Thread Starter New Member

    Jan 6, 2009
    8
    0
    Hello,
    I have a cell load and need to supply it and acquire the differential signal (please find the attachement). The question is where the sensor end of the shield should be connected?
     
  2. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
    Connect only ONE end of the shield! If both ends are grounded, it won't be a shield anymore - it will be a "ground loop" and you'll get all kinds of noise on your line. A ground loop is essentially a shorted transformer winding.

    Frequent practice is to ground the shields at a physically common point. IE: In a panel with cables leading to remote sensors and controllers, shields would be grounded only at the said panel.
     
  3. przemmo

    Thread Starter New Member

    Jan 6, 2009
    8
    0
    So in this paper http://www.bustec.com/support/papers/difinput/connecting_signals.pdf Figure 2 we have a ground loop? Yhis paper treats about single ended signals( the signal source is referenced to groung IE Fig 1, fig 4, or is insulated). A bridge gives differential signal, and in such a case only one end of the shield should be grounded (better at the common potential)? Where can I find any informations about shielding differential signal ?
     
  4. Alberto

    Active Member

    Nov 7, 2008
    169
    36
    Sometimes connecting both end of a shield to ground reduces the noise, some othertimes noise increases significantly.
    In the first case you are not creating a ground loop in the second case you are. Since creating a ground loop connecting both ends of the shield to ground depends on where the connection are made and this depends only from the layout of the circuit, so the best way to solve the problem is: use a scope to check out the noise level with one end to ground versus both ends grounded.

    Alberto
     
  5. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    As Alberto says, eliminating noise is sometimes trial-and-error. As the cabling has to carry the excitation, ground and the two LC outputs, the shield has to run the same potential as circuit ground. Where it becomes common with circuit ground (the star point) can be significant.

    Results from the trials are immediate, though.
     
  6. przemmo

    Thread Starter New Member

    Jan 6, 2009
    8
    0
    So at first I'll try to ground it only to the amplifier's ground. Will let You know.
    Thx
     
  7. pandit

    New Member

    May 27, 2009
    1
    0
    Hi, Where u have connected the sensor shield wires(+sense, -Sense)???

    Did u ground the both lines....??


    Pandit
     
Loading...