help with this preamplifier

Discussion in 'General Electronics Chat' started by alarassi, Mar 15, 2010.

  1. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    hello guys​

    attached is a preamplifier i'm using in my project!i want to feed the output of this circuit into PIC for ADC and then further programming! i have few problems with the circuit​

    this preamplifier bias the output voltage to avarage half of the supply voltage, let's assume in my case to about 4.2V! the problem is that since i use mic to detect sound, this reference voltage of 4.2 varies with the sound detected by the mic( in other word when i detect sound of different frequencies and intensities this voltage varies up and down but when there's no sound on the mic it becomes stable) it varies in range about -,+ 0.1-0.9 added to the reference voltage.
    is there any way to stabilize the reference voltage while detecting the sound? (using it to detect heart sound)
    or does it affect on the process of ADC if it's unstable?​

    another things, do i need to add buffer to the output of the circuit or no need since the active LPF gives me low impedance?​

    how is the way to calculate the HPF filter that produced by the capacitors in the circuit? i need it to be about 30 at -3db and i want to know how to calculate it!​

    last question, is my schematic ready to be fed into PIC or i have missing parts?​

    the amplifier will be replaced by another type that accept 5V supply voltage so that i can feed it into PIC​

    (the original preampilfier schematic has been givin to me by the member Audioguru)​

    thank you​
     
  2. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Audio is an AC voltage. It causes the output to swing above and below the DC reference voltage .

    The buffer is useless. Remove R7, R8, C3, C4 and U2A.
    Your circuit does not have a lowpass filter. Make a 2nd-order Butterworth Sallen and Key lowpass filter that is fed from the output of the preamp opamp.

    The formula for an RC filter is one divided by pi x R x C.
    The mic is about 3.3k and it is in parallel with the 10k resistor so they are 2.5k ohms. The opamp has FET inputs so its input resistance is infinite. The two 100k divider resistors are in parallel for the signal so they are 50k ohms. The total resistance of 52.5k ohms and the 0.22uF coupling capacitor produces a highpass filter that is -3dB at 13.9Hz.

    The 1k resistor is in series with the 10uf capacitor so it produces another highpass filter that is -3db at 10Hz.

    The total of the two filters reduces the output to nearly -6dB at about 12Hz.

    The gain is 1+ (20k/1k)= 21 from about 100Hz to about 200kHz.

    Depending on which opamp you use, the output with a 5V supply will be about 1V to 4V and will be AC with a 2.5VDC bias.
     
  3. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    so that means it doesn't affect the reading of ADC of the pic when it swings up and down! right?

    oh sorry, i think i made a mistake in drawing the schematic, C3 actually should be connected to the output of U2A as feedback to form 2nd-order Butterworth Sallen and Key lowpass filter! in this case should i have buffer or the active LPF gives me low impedance so that i don't need buffer?

    you advised me before to change input capacitor value into 0.22uF and also to change the feedback capacitor to ground into 10uF so that i can have total cutoff frequencies is about 25Hz at -3dB (last reply in this thread). http://forum.allaboutcircuits.com/showthread.php?t=34563
    my need is to have a total of HPF about 30Hz out from the preamplifier as well as LPF

    i'll use OPA2134 or OPA4134




     
  4. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    no because i want the output to be in a range of 0-5V so that the ADC of the pic can read the voltage.

    thank you
     
  5. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I don't know what your ADC will do.

    Then the corrected lowpass filter is a droopy Bessel not a sharp Butterworth.
    An opamp has an extremely low output impedance so a buffer is not needed.

    But why didn't you continue in your last thread? I don't know which sounds you have and why you are blocking high audio frequencies.

    Then do it.
     
  6. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    i think mine is butterworth filter!! i'll double check it again!

    the sound is heart sound n i want to have HPF with fc=30Hz,
    i want to know, is it possible to have about 30hz at -3db in this preamplifier without affecting the biased output reference voltage? if yes show me the way and i'll calculate it myself.

    thank you
     
  7. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    oh i'm confused here, R1 is not 1M, it's 10K ohm!!
    do u mean that i shouldn't connect the mic to the resistor R1? and instead of that connect R1 directly to ground?? this thing might affect the mic output right?
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Dataman,
    What are you talking about?
    R1 is not 1M, it is 10k. It powers the Jfet transistor inside the electret microphone.
    R1 is decoupled from the supply by R4 and C5 and the signal from the mic is very low level anyway (10mV).

    Alarassi,
    Heartbeats are around 16Hz so a highpass filter at 30Hz is too high. Use 10Hz or less for less tilt in the waveform.
    A 2nd-order Butterworth lowpass filter has its feedback capacitor double the value of thew capacitor to ground but your capacitor are the same so it is a droopy Bessel filter. A Butterworth filter can use the same R and C values if the opamp has a gain of 1.6.

    The output is AC audio so the output voltage goes above and below the output DC reference voltage.
     
  9. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    i really appreciate your help but i really need to know how to calculate them to have the total value at -3db!
    is it by followint the normal formula of RC passive filter ? or it has another way!
     
  10. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    One RC filter has an output at its cutoff frequency that is down -3dB which is 0.707 times the voltage or half the power.
    Two RC filters that are buffered so they do not affect each other have a cutoff frequency that is down -6dB which is half the voltage or 1/4 the power and the response is droopy.

    At the cutoff frequency the phase of the output is changed which causes the display to be tilted.
     
  11. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Nobody knows the exact current drawn by an electret mic because they are all different so without a coupling capacitor the opamp might be cutoff or saturated.

    If you use larger coupling capacitors then they will pass the low frequencies you need. You need a highpass filter if you have many earthquakes or have crazy people jumping around.
     
  12. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    hello guys

    i used the circuit given in the first page and it works really nice but something i couldn't understand! how the voltages get biased to half of the supply voltage!?
    i know it's through the three resistors given there but what i mean is, what is the technique used? and how its calculations?
    i tried to find on the net and understand but i really couldn't !!
    could any one explain for me or at least guide me to any sources that i can learn from!

    another thing, i didn't find the internal connections or structure for the mentioned amplifier on its datasheet!! i don't mean the connection of pis but the FET connections! how can i get it ?

    thank u
     
    Last edited: Apr 7, 2010
  13. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    ok guys, i searched and got the answer! simply by using resistive divider :D

    another question! what are the charactiristics that make single OPA1234 or dual as good amplifier for audio application!? i know it's low noise and distortion and others but what i mean here is why it's low noise? what features they have to make them low noise op-amp?

    thank you
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I never designed an opamp. I just look at their spec's and select the opamp that works well in my application.

    The old opamps like the 741 and LM324 were made for amplifying DC and low frequencies. They were noisy (hiss) and cut high audio frequencies that didn't matter. The LM324 and LM358 were designed for low power so their output transistors did not have enough current which produces crossover distortion.
    Better opamps like the TL07x and OPAx134 were designed for good audio.
     
  15. nomurphy

    AAC Fanatic!

    Aug 8, 2005
    567
    12
    Anyone with a heartbeat of sixteen times a second isn't going to be around long, 1.6Hz would be more reasonable.
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Something vibrates at about 16Hz during each heartbeat. Maybe it is the skin trapped under the stethoscope head. The Dark Side of The Moon album begins with a 16Hz heatbeat. My 3" computer speakers just played it and had a difficult time producing 16Hz. My 8" speakers produce it fine. The album is 37 years old.
     
  17. nomurphy

    AAC Fanatic!

    Aug 8, 2005
    567
    12
    The song you reference has what sounds like a typical heartbeat, thump-thump, thump-thump of about 1.5Hz (count the seconds between), it's definitely not a 16Hz "heartbeat" which would definitely be ventricular tachycardia or ventricular fibrillation.

    The 16Hz could be some resonance or artifact of the blood flowing into the atria and ventricals or along the arterial wall.
     
  18. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Since I had two blocked heart arteries fixed last year with angioplasties, my heart beats at about 58 beats per minute when I am not excercising. My home made electronic stethoscope plays the 16Hz vibrations of each heartbeat loud and clear.
     
  19. nomurphy

    AAC Fanatic!

    Aug 8, 2005
    567
    12
    So, as you know, 58 beats/minute is just under 1 Hz. Your perceived 16Hz vibrations are not heartbeats, but most likely blood movement heard as a vibration.
     
  20. alarassi

    Thread Starter Member

    Feb 6, 2010
    52
    0
    audioguru

    i have made the circuit and connected it to PIC and i had the correct output! thank you ;)

    i'm thinking to have an extra feature on my project by adding a headset to listen to the heart sounds!
    the one i constructed feeds the signal values into PIC but here i want to add a headset circuit so that i can listen to and watch the signals.

    can i do that without affecting the current constructed circuit's values?
    if yes, how can i do it?

    thank you
     
    Last edited: Apr 10, 2010
Loading...