Connect with us

Differential Configuration for a Floating Signal

Discussion in 'General Electronics Discussion' started by AdamZ, Jun 30, 2016.

  1. AdamZ

    AdamZ

    18
    1
    Mar 17, 2016
    Why is it necessary to use biasing resistors for a floating signal in differential mode of a DAQ card (analog)? What would happen if it was kept floating?


    NI link:
    http://www.ni.com/white-paper/3394/en/
     

    Attached Files:

  2. AnalogKid

    AnalogKid

    2,297
    646
    Jun 10, 2015
    The article explains in three places why it is necessary. Which parts do you not understand?

    ak
     
  3. AdamZ

    AdamZ

    18
    1
    Mar 17, 2016
    It says:

    "If bias resistors are not used in a differential...when measuring floating signal sources, the measured signals can be unstable or at positive or negative full-scale range of the instrument."

    But it doesn't give an explanation as to why it would become unstable. I'd like to understand the mechanism.
     
  4. BobK

    BobK

    7,607
    1,648
    Jan 5, 2010
    What is the voltage between a floating voltage the ground of the instrument?

    Bob
     
  5. hevans1944

    hevans1944 Hop - AC8NS

    4,214
    1,994
    Jun 21, 2012
    Differential amplifiers, even those with FET inputs, have bias currents at their input terminals. These currents need to be returned to circuit common or they will charge up the capacitance between each of the input terminals to circuit common and drive the amplifier toward either the positive rail or the negative rail, depending on which bias current is greater than the other. With a very high input impedance FET front-end, with very low bias currents that are carefully matched, the process can take days or weeks to occur. When you go to troubleshoot the problem, it goes away as soon as you probe the inputs because the probe discharges the "stray" capacitance.

    I have built instrumentation amplifiers (in the 1970s) with bias resistors as large as 100 meg-ohms, but these were experimental R&D projects. One to ten meg-ohms is a reasonable value to use and will have no effect on the signal from a reasonably low-impedance source. The resistors should be matched in value for best common-mode rejection performance.
     
    AdamZ likes this.
  6. AdamZ

    AdamZ

    18
    1
    Mar 17, 2016
    Perfect explanation, as usual!
     
  7. hevans1944

    hevans1944 Hop - AC8NS

    4,214
    1,994
    Jun 21, 2012
    Oh, I forgot to mention that smaller-valued bias resistors are better, as long as the application can tolerate the decreased differential input resistance. Smaller valued bias resistors decrease the differential offset voltage created by the bias currents, and also decreases the Johnson (thermal) noise created in the bias resistors, hence, lowers the input noise voltage, too. We only went to 100 meg-ohm bias resistors because we were working with extremely high-impedance sources. Everything associated with the inputs had to mounted on Teflon insulating posts, but that's an entirely different problem.
     
    AdamZ likes this.
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

-