Calculate value of a capacitor for low pass filter?

Discussion in 'General Electronics Chat' started by siliconviet, Mar 31, 2014.

  1. siliconviet

    Thread Starter New Member

    Mar 30, 2014
    Dear all,

    I have a question on low pass filter.

    I am working on VOX board (please see the picture below), and experiencing noise from Microphone.
    Voice frequency is from 20-20000 Hz. So I want to add a capacitor C10, to cut off frequency above 20000Hz.
    What value of capacitor should I take? Should I know the resistance of a microphone for calculating?

    Best regards:)
    • VOX.jpg
      File size:
      73.1 KB
  2. Alec_t


    Sep 17, 2013
    If you can hear it it's almost certainly < 20kHz, so a cut-off above 20kHz will have no noticeable effect.

    Assuming your mic is an electret type and R5 (10k) is its drain resistor, the signal power will be reduced by 6dB when the C10 reactance X is 10k.

    X = 1/(2*pi*f*C)
    C= 1/(2*pi*f*X)
    For f= 20kHz and X=10k, C= 1/(2*pi*20k*10k) Farads = 1/(2*pi*20) uF = 8nF (you could use nearest standard E12 value =8.2nF)
    siliconviet likes this.
  3. AnalogKid

    AAC Fanatic!

    Aug 1, 2013
    You are asking about a single-pole low pass filter, and there are better places in your circuit for it than the input. A couple of things about your circuit. You might want to sit down for this.

    First, an electret microphone does not have a symmetrical output impedance. This causes distortion when the mic is loaded with anything other than a resistor. So hanging a cap directly on the mic output is not recommended.

    Second, an LM358 has significant crossover distortion. If it is the amp you're stuck with, you can help this by adding a fixed resistor from the output to GND.

    Third, the only DC bias for the circuit is the DC output level of the electret, something that rarely is specified and never is well controlled. Depending on the circuit power voltage, the positive peaks might be clipped by the 358. If you're stuck with a single power supply, there are ways to fix this with two resistors and one cap.

    Fourth, voice frequencies run from around 50 Hz to under 8 KHz, and normal human hearing is barely better. That 20-20000 thing is for young children that never have been to an action movie, worked on a farm, been to a live music event, own an iPOD, or anything else that makes a sharp loud noise. Did some work for an acoustics lab, learned a lot.

    Fifth, R12 and C9 form a 72 Hz highpass filter. And, the 0.1uF output cap will form an even higher hipass filter with the input to the next stage. If the input impedance of the next circuit is 10K, this output filter cutoff is 160 Hz. If the impedance is 1K, the cutoff is 1.6 KHz. I'm just sayin...

    Sixth, The LM358 is does not have very high open-loop gain. At your high gain value is 48 dB, the 358 open-loop gain crossover frequency is around 3 KHz.

    And finally, your noise filter. A better place for the cap would be to replace C7. As shown in your drawing, C7 forms a lowpass filter with the series feedback resistors. Depending on the gain adjustment, its effect is a lowpass filter at 6 KHz (32 dB gain) to 128KHz (48 dB gain). If you change C7 to 200 pF and connect it directly across R11, you will have a more stable cutoff frequency. True, the max attenuation above cutoff will not be as deep and will vary with the gain adjustment, but the corner frequency will stay put. Life is choice.

    If it sounds like I'm dumping all over your circuit, that is not my intent. And what you have certainly will amplify the microphone output. But if your goal is a microphone preamp that is flat to 20 KHz, at 48 dB gain, with less than 1 or 2% distortion, this circuit is not it.

    siliconviet likes this.
  4. siliconviet

    Thread Starter New Member

    Mar 30, 2014
    Thank you for your analysis. Your comments helped me to understand better the circuit.:)