Floating OpAmp Sensor Circuit Help Please

Discussion in 'The Projects Forum' started by dosequisxx1, Oct 29, 2013.

  1. dosequisxx1

    Thread Starter New Member

    Oct 28, 2013
    5
    1
    Hi Everyone,
    I'm having trouble designing an opamp circuit that does a 2:1 amplification of a 140Hz sinusoidal signal +-0->1V without a ground plane. Does anyone know a simple method of doing this without specifying a reference plane?

    If possible, I'd like to do this with just some resistors a battery bank (3AAAs = ~4.5V) and a single opamp.

    Thanks!

    :)

    (If I add in a 5V regulator and up the battery bank to 5AAAs to power it, can I use the regulators ground as the ground plane?)
     
  2. #12

    Expert

    Nov 30, 2010
    16,316
    6,818
    A voltage is ALWAYS a voltage compared to some other place. You know where you placed the black lead of your meter/scope/etc to measure the voltage. Use it.

    You might have to AC couple the signal. You might need a differential amplifier. This is what I can do with the information you have provided.
     
    dosequisxx1 likes this.
  3. dosequisxx1

    Thread Starter New Member

    Oct 28, 2013
    5
    1
    Ok so I didn't give enough information, for sure, sorry about that. I have a signal coming out of a vibrating piezo at roughly 140 Hz and 1V... It's generating. I also have a battery bank generating ~4.5V (3AAAs, probably should up it to 5 anyway). I would like to use an opamp to amplify the piezos voltage signal at 2:1 and then read this on a portable oscilloscope (no earth ground, there is an internal ref on the oscope though I'm not sure what it can handle, it's pretty crappy).

    How would you think to develop a reference plane out of that, or do I need to expand the circuit? (ie - should I put in the extra pieces like a regulator and such to generate a stable reference plane)

    Thanks again for the quick reply and the help!
     
    Last edited: Oct 29, 2013
  4. #12

    Expert

    Nov 30, 2010
    16,316
    6,818
    Ground doesn't mean, "Planet Earth". It means, "common". Whatever the scope uses for common is good enough. Just connect the piezo from the input of that amplifier to common. You will need an opamp that can operate on 5 volts. Something like an NJM4250D. 82 cents at www.mouser.com

    However, piezos have a habit of putting out the kind of voltage that kills transistors.
    I'm going to change the drawing. Add a couple of small Shottky diodes so the piezo can't go Kamikaze on the chip.
     
    Last edited: Oct 29, 2013
  5. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    Try this: Only ground is at the scope's input; every thing else is floating...

    Plot is ScopeInput vs sensor output voltage.
     
    #12 likes this.
  6. #12

    Expert

    Nov 30, 2010
    16,316
    6,818
    Yikes. Forgot the center reference.
     
    Last edited: Oct 29, 2013
    dosequisxx1 likes this.
  7. dosequisxx1

    Thread Starter New Member

    Oct 28, 2013
    5
    1
    The title pretty much says it all. Thanks a whole bunch. Like a panama class freighter size of bunch.
     
    #12 likes this.
  8. dosequisxx1

    Thread Starter New Member

    Oct 28, 2013
    5
    1
    Ok so I simulated the circuit, it worked perfect. Then I wired up the circuit exactly as simulated and some things happened.

    1 - It did a 2:1 amplification
    2 - It seems to have only amplified half the signal, as if it were being half-wave rectified prior to amplification.
    3 - It induces an offset from 0 to -Vbattery

    This was using the diagram provided by #12.

    Any ideas why these things may be happening or how to correct them?

    I did notice that the center reference looks pretty much like a half wave rectifier.

    Thanks again for the help.
     
  9. dosequisxx1

    Thread Starter New Member

    Oct 28, 2013
    5
    1
    This is what I'm currently working with.
     
  10. #12

    Expert

    Nov 30, 2010
    16,316
    6,818
    You made the centering divider the same resistance as the feedback resistors, which makes them become significant in the gain equation. They need to be a tenth of the impedance of the feedback resistors, or less.

    You designed the amplifier for a gain of 3.
    You made the load use 100 times the current of the centering resistors. It can't work that way. We expected the scope to be 1 million ohms.
    You are all out of proportion.

    The centering resistors can be as low as 1k. The feedback resistors have to be at least 10 times the value of the centering resistors. The load has to be at least ten times the value of the centering resistors. The diodes should almost never come into play. They are, "just in case" diodes. Just in case the piezo produces more than +/- 3.8 volts, the excess will go through a diode to either the power + terminal or the power common terminal so the piezo doesn't blow out the amplifier input.
     
    dosequisxx1 likes this.
Loading...