Creating a bias for an op-amp and AC signal

Discussion in 'Embedded Systems and Microcontrollers' started by DonnellyCircuits, Nov 1, 2015.

  1. DonnellyCircuits

    Thread Starter New Member

    Nov 1, 2015
    4
    0
    I have an MCP601 op-amp and a piezo disc and I want to amplify the signal from the piezo but also have a biased signal in order to read the negative voltages. The device is supposed to be used as a guitar tuner to pick up the vibration of a string.

    Can someone explain how I would get this effect, I have read http://www.allaboutcircuits.com/textbook/semiconductors/chpt-8/op-amp-practical-considerations/ but I don't really understand it.

    The following circuit is a portion of the whole thing. The rest connects to an ADC and then to a Raspberry Pi. I expected to get a bias but when it's attached the amp just amplifies the bias and I get a constant 5V.
    Can anyone tell me how to alter my circuit so that I get a 2.5V bias and the piezo moves around that value?
    [​IMG]

    Thank you
     
  2. JohnInTX

    Moderator

    Jun 26, 2012
    2,344
    1,025
    There are many ways to do this. The easiest, I think, and the one I selected for my ongoing guitar amp project is a 'Rail Splitter' IC. Datasheet attached.
    You also can visit sites like ESP that have lots of amplifier schematics/projects using various other ways.
    If you have a single supply, some way of dividing it by 2 and calling it 'ground' for the signals to swing around is necessary. The tricky part is to do the split with enough current capacity do run the circuit.

    Not sure why you want to go to the ADC for a tuner, though. Don't you just want to count waves to get the frequency of the input?

    Have fun.
     
    Last edited: Nov 1, 2015
  3. DonnellyCircuits

    Thread Starter New Member

    Nov 1, 2015
    4
    0
    I've looked for some schematics but they don't seem to make sense, or when I actually use them, work.
    Do you have a schematic of this?

    I'm new to electronics but I didn't think this would be such a stretch, I've been doing this for a while now.
     
  4. JohnInTX

    Moderator

    Jun 26, 2012
    2,344
    1,025
    Yeah, its not too clear..

    You would connect Vin to Rpi5V and GND to GND. Now, Vo is Vin/2 and buffered to drive things.
    The op amp power is the Rpi5V and GND. All of the signal 'grounds' are Vo from the rail splitter. This is almost the same as powering the op-amp from a +/- supply and having signal grounds as 0V (again, halfway between the power rails of the op amp). The difference is that the 'no signal' level with the rail splitter is 2.5V and the signal swings +/- around that level. You can remove the offset in the ADC arithmetic. With a +/- supply, the signal swings positive and negative with respect to 0V. Your ADC will not like that nor would it like, in the first case, the effect of cap-coupling the output and referring it to GND to remove the offset. The signal would again swing negative with respect to ADC GND.

    If you are just counting intervals between wave peaks, you could just remove the 2.5 V offset in the original drawing. The amplifier would amplify only the positive peaks. Louzy audio but probably enough to determine frequencies.
     
  5. DonnellyCircuits

    Thread Starter New Member

    Nov 1, 2015
    4
    0
  6. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    You have V1 connected to the wrong place.

    Here is how I would do it:

    The non-inverting input is biased to 2.5V with an input impedance ~1meg (like you started with).

    By putting a DC blocking capacitor in the feedback path, the gain is still (100K/10K +1) = 11 for mid-band frequencies, with a roll off below ~1/(2*Π*R1*C1).

    270a.gif
     
    Last edited: Nov 1, 2015
    F Df and JohnInTX like this.
  7. F Df

    New Member

    Apr 22, 2016
    1
    0
    Can do you explain
    Can do you explain it?
     
Loading...