1
\$\begingroup\$

I was designing a microphone pre-amplifier circuit for Wireless Noise Monitoring application. I used an electret microphone AOM-5244L-R-53724 (datasheet: http://www.puiaudio.com/pdf/AOM-5244L-R.pdf)and an amplifier connected to a 9V power supply. The multisim simulation of the circuit with a function generator gives the expected gain of around 7.8. But the same circuit in the breadboard with the microphone gives the gain of only around 5. Please let me know what improvements are needed in the circuit to get the expected gain of 7.8?

enter image description here

Output from the oscilloscope:

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ Have you checked the actual values of your resistors? If they are 10% tolerance, for the two you used to set the gain you could be up to 20% out. Also, are you sure your opamp swings rail to rail? \$\endgroup\$
    – Tim M
    Commented Mar 4, 2015 at 16:34
  • \$\begingroup\$ What type of capacitor are C1, C3 and C4? If electrolytic, what are their polarities? What is the voltage on pin 6 of U1? How did you measure the gain? Can you show us a photo of your breadboard circuit? \$\endgroup\$ Commented Mar 4, 2015 at 17:05
  • \$\begingroup\$ What is the output impedance of your signal used to test the gain - might it's output by reducing when you connect it to the circuit? This will give an apparently lower gain. Also, what frequency test signal are you inputting? \$\endgroup\$
    – Andy aka
    Commented Mar 4, 2015 at 17:57

2 Answers 2

2
\$\begingroup\$
  1. 1.8 kΩ seems rather small for a electret pullup, especially to 9 V. Have you checked to see what the datasheet recommends?
  2. You are actually asking for a gain of 6800 in the passband. I can't even guess where your 7.8 figure came from. This is bad idea for a single stage. Even at 400 Hz you are asking for a overall gain*bandwidth 2.7 MHz, so the opamp needs to be good to 10 MHz or so for the feedback to have some headroom. You should only need a few 100 to 1000 gain from a ordinary electret microphone to line audio levels.
  3. 6.2 MΩ is really high. I didn't lookup that opamp, but even if its input impedance is significantly larger, there can be other noise that gets into the signal at that impedance. Just leakage thru dirt on the PCB can be significant.
  4. With such a large gain, this configuration will take a long time to stabalize after power up. The (R1 + R2)C1 time constant is one minute! And, you would need to wait a few time constants for the system to settle close enough to the final operating point.
  5. The high pass rolloff frequency set by R2 and C1 makes sense for audio, but the low pass rolloff set by R1 and C2 is 380 Hz. Is that really what you intended. Personally, I'd probably leave off C2 altogether.
\$\endgroup\$
8
  • \$\begingroup\$ He's actually asking for a gain of 7.18. 1 + R1/R2. Standard non-inverting amplifier. \$\endgroup\$ Commented Mar 4, 2015 at 18:00
  • \$\begingroup\$ I think it is a 6k19 resistor not 6190k \$\endgroup\$
    – Andy aka
    Commented Mar 4, 2015 at 18:01
  • \$\begingroup\$ @What: 1 + (6.19 MOhm / 910 Ohm) = 6800 like I said. Perhaps Andy is right and he is assuming a different value for R1, but it's shown on the schematic as 6.2 MOhm ("6,190kOhm"). \$\endgroup\$ Commented Mar 4, 2015 at 18:25
  • \$\begingroup\$ At least on my screen that's decimal point, not a comma. There is a grid point down and to the left that may look like a tail to you. \$\endgroup\$ Commented Mar 4, 2015 at 18:33
  • \$\begingroup\$ @What: That's possible. Of course a schematic at decent resolution to be able to see here would have been good. \$\endgroup\$ Commented Mar 4, 2015 at 18:43
1
\$\begingroup\$

But the same circuit in the breadboard with the microphone gives the gain of only around 5

Here's a potential problem. The electret microphone will have an output impedance of maybe a few hundred ohms. When you connect your signal generator across the electret to inject a signal, the 50 ohm output impedance (fairly normal for test equipment) signal will be attenuated a little bit and this might make you think the gain is falsely low.

Try disconnecting the electret mic and see what happens. If, in fact your signal generator is 600 ohms output then the 1k8 bias for the electret (even if electret no present) is also going to significantly erode your gain thru attenuation of the input signal from the sig generator.

\$\endgroup\$
3
  • \$\begingroup\$ I am not using the signal generator and the microphone together. I just did the multisim simulation with the signal generator. I replaced the signal generator with the microphone and played different frequency sine waves (from MATLAB), starting 20 Hz to 20KHz. Gain was measured from the oscilloscope, keeping one probe at the input and another at the output,and measuring the peak to peak voltages. All the capacitors are electrolytic except the 68 pF (ceramic). The resistors are 5% tolerance. If I disconnect the microphone and use signal generator instead, the expected gain is obtained. \$\endgroup\$ Commented Mar 8, 2015 at 15:24
  • \$\begingroup\$ Are you saying it's ok? \$\endgroup\$
    – Andy aka
    Commented Mar 8, 2015 at 16:54
  • \$\begingroup\$ The circuit works fine with the signal generator so I think there is nothing wrong with the circuit and the expected gain (7.8) is obtained. Using the microphone in the same circuit lowers the gain to around 5. What might be the reason that is not giving me the gain of 7.8 when microphone is used instead of a signal generator. The microphone I am using is link .Your help is really appreciated. \$\endgroup\$ Commented Mar 8, 2015 at 17:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.