Connect with us

Virtual Ground to Op Amps - Will This Work?

Discussion in 'Electronic Design' started by Jim Stevens, Nov 17, 2012.

Scroll to continue with content
  1. Jim Stevens

    Jim Stevens Guest

    I would appreciate comments about the linked circuit below. It is two
    op amps powered on different sides of a virtual ground, with the
    outputs connected across a resistive load.

    http://www.4shared.com/photo/llPcodei/virtual_ground_dual_current.html

    Any suggestions on which common op amps would be suitable for this
    circuit?

    Yes, it is a little odd. Just curious to see what happens.

    Jim Stevens
     
  2. Jasen Betts

    Jasen Betts Guest

    biasing of the right two op-amps looks kind of wierd.
    hmm, 3V supply, 12mA output probably not a cheap one, but really it
    depends what the goal of the circuit is.
     
  3. If the circuit has no purpose, then may as well choose the very
    cheapest one, since it will meet all the other requirements.

    The circuit, as drawn, might be a problem set by an instructor to
    illustrate non-ideality in op-amps.

    You'll see offset, noise and non-ideal voltage swing/saturation. A few
    LM358s might suffice for that kind of an exercise if you don't let the
    ideal battery drop too much in voltage.


    Best regards,
    Spehro Pefhany
     
  4. P E Schoen

    P E Schoen Guest

    "Jim Stevens" wrote in message
    The virtual ground is reasonable enough to provide a +/- supply for op-amps
    from a single battery, but the ones shown are not connected that way, and
    the circuits do nothing meaningful. You might want to simulate this circuit
    using LTSpice (www.linear.com). It would be a good learning experience. What
    are you trying to accomplish?

    Paul
     
  5. Jim Stevens

    Jim Stevens Guest


    And, I beleive, the degree to which they collapse depends on the value
    of R1/output current.

    With regard to the other comments so far, 12mA (and more) at 3V can be
    done with something like a TS941. Although I was hoping to use
    something on hand.

    As for "uselessness", wasn't that what they first said about the
    laser? It's just learning by tinkering here.

    Jim Stevens
     
  6. Jim Stevens

    Jim Stevens Guest


    Can you please explain briefly what features of the circuit cause this
    to occur?

    Is the "offset" you refer to due to the output of the negative amp not
    reaching ground?

    Noise is not an issue here, and I realize the swing will be odd.
    Hopefully odd enough to be somehow interesting.

    Jim Stevens
     
  7. Jim Stevens

    Jim Stevens Guest

    Why nothing meaningful? I didn't show the signal inputs on pin 2 of
    each amp, assuming it would be implied. It is just a 2Vpp audio
    frequency sinewave to both.

    Hence the outputs are identical, but with a DC offset from each other
    governed by the load resistance.

    Several experts here have said there is a problem with the circuit,
    but I would like to understand exactly why. It's an experiment, not a
    hi-fi system.

    Jim Stevens
     
  8. Guest

    Where to start... The input opamp is a fairly common configuration to
    split the rails for the bias point of subsequent opamps, though often
    it's set up as a low-pass filter to reduce wideband noise. R3 does
    nothing good, and limits U1's output current (not a good thing).

    U2 and U3 use U1's output for their positive and negative rail,
    respectively. This means the output of the upper rail cannot go below
    this and the lower cannot go above this. Therefore, you *must* have a
    DC current in the output. This isn't generally a good thing for an
    audio circuit. I guess there may be reasons for this but you haven't
    said what you're trying to do so this whole discussion is pretty
    silly.

    The U3 is referenced to the negative rail so its output can't go below
    that. IOW, if your input is a sine wave with 0V DC bias it's going to
    clip. Ditto U2 and its negative rail (the output of U1). Both are
    going to clip on the negative half cycle. Even if there is a DC bias
    on the input signal, there is no DC bias where neither opamp is
    clipping.


    Nope. At least one is always driven into the rails. Even if this were
    the case, why would you want such a thing? One opamp can do this,
    BTW. Just sum in a DC voltage.
    If it's an experiment why ask here? Why not see for yourself that it
    does nothing good? If you're trying to accomplish something specific,
    give the complete picture and people will be glad to help (though it's
    probably better over in .basics).
     
  9. whit3rd

    whit3rd Guest

    .... which results in the outputs of each op amp being pulled
    by the other outside the absolute-max limit on the output pin.

    The 'output' op amps are both driven in such a way that they
    can lock-up (if offset voltage is considered, the balance
    equation for the output state includes TWO solutions).
    Arguably, the load resistor can pull up the lower op amp to
    keep it in linear operation.

    I don't see any input.

    The ground-establishing op amp has 110k ohms (series resistance)
    on its (+) terminal; a 110k ohm series resistor to the (-)
    terminal will keep the temperature dependence low (bias current
    times imbalance of resistance adds to voltage offset, and bias
    current is strongly temperature dependent).
     
  10. Jim Stevens

    Jim Stevens Guest

    On Sun, 18 Nov 2012 09:04:08 +1100, Jim Stevens

    Now that I have a better idea of what to expect, I will build one up
    this week and verify how it performs.

    It's hard to predict an unusual circuit without alot of practical
    experience ... more than I have anyway.

    Thanks for the comments.

    Jim Stevens
     
  11. Guest

    If you can't predict how it's going to work, why are you wasting time
    building it? You have the cart before the horse (in so many ways).
     
  12. Jim Stevens

    Jim Stevens Guest

    This sounds like a good idea. Can you be a little more specific about
    the method you are referring to?

    As previously mentioned, this is just a learning process. There is
    really no goal apart from that. I suppose everyone here is over that
    stage.

    Jim Stevens
     
  13. P E Schoen

    P E Schoen Guest

    "Jim Stevens" wrote in message
    You should know enough about op-amps to figure out basic operation. You can
    assume ideal operation and determine approximate behavior in a circuit. It
    appears that you are trying to achieve a sort of bridge circuit output that
    can swing between + and - rails from a virtual ground reference. But clearly
    the two output op-amps can only swing between their respective rails and
    virtual ground. Thus the voltage across the load will always be positive. I
    doubt that is what you want. I can't see it being a learning experience to
    build this circuit, except maybe using a simulator. Usually, the only reason
    for building an actual prototype circuit is to observe behavior that is not
    as easily predicted or modeled, such as instability and oscillation. Or
    possibly behavior of op-amps under other than recommended conditions, which
    can cause latch-up and other possibly destructive operation.

    You might want to check out some proven op-amp circuits and learn from them:
    http://www.amazon.com/Op-Amp-Cookbook-3rd-Walter-Jung/dp/0138896011
    http://www.analog.com/library/analogDialogue/archives/39-05/op_amp_applications_handbook.html
    http://www.ece.ucsb.edu/Faculty/rodwell/Classes/ece2c/resources/an-31.pdf
    (National)
    https://instruct1.cit.cornell.edu/courses/bionb440/datasheets/SingleSupply.pdf
    (TI)
    http://en.wikipedia.org/wiki/Operational_amplifier_applications

    and learn LTSpice:



    Paul
     
  14. Jim Stevens

    Jim Stevens Guest

    On Sun, 18 Nov 2012 09:04:08 +1100, Jim Stevens


    Please excuse the profileration of posts, and my lack of initial
    clarity. Here is a revised circuit showing the signal inputs and other
    changes.

    http://www.4shared.com/photo/29tKWyFG/Virtual-Ground-Amps-2.html

    Input signal is line audio (test tone).

    Note the outputs are not ground referenced.

    Jim Stevens
     
  15. Phil Allison

    Phil Allison Guest

    "Jim Stevens"

    ** So you have a signal voltage that directly drives the virtual earth
    inputs of two op-amps above and below supply rails.

    Brilliant.


    .... Phil
     
  16. Guest

    None whatsover.

    Opamps do not like non-zero differential voltages applied to their (-)/(+) inputs, they just rail. So U2 rails to the node labeled GND and U3 rails to node labeled NEG on battery. It's the only way to drive the currents thru R4/R6 to 0.
     
  17. josephkk

    josephkk Guest

    The circuit as drawn won't work worth squat.

    ?-)
     
  18. josephkk

    josephkk Guest

    Not quite as messed up as the first one, however the output is 100% common
    mode with a DC bias current. None of the input signal appears on the
    output in a useful way.

    Try powering both output amplifiers from the battery directly and make one
    inverting and the other non-inverting. You might get some useful output
    that way.

    ?-)
     
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

-