0
\$\begingroup\$

I want to sample an audio source using an electret microphone and analyze the spectrum of it with a uC. I'm interested in the 100Hz-2.5kHz range, so I calculate with these values when I tried to design the preamp circuit.

One of my try: schematics

When I leave R7, I get ~10nV output! But when R7 isn't used, I get the output pretty well: output without R7

I want to ask four questions:

  1. R7 and C1 is for AC coupling with a corner frequency ~50Hz, also C1 and R1 is a HPF too, but if I put away R7 how can I get the -40 dB/d slope?
  2. If I don't put R7 there, will the opamp's input bias current charge C1 and cause an offset drift? How can I handle this without R7?
  3. Will that 1M resistor cause bigger noise?
  4. Is C4 is enough for the output AC coupling? VF1 will be connected to a PGA's input to extend the dynamic range of the sensor.

For the simulation, VG1 generates a 10mV, 440Hz sinusoidal signal.

\$\endgroup\$
  • \$\begingroup\$ Connect R7 to VS2, not ground. VS2 is your signal reference. \$\endgroup\$ – sstobbe Oct 10 '18 at 2:10
0
\$\begingroup\$

I would answer your questions, but you are making several mistakes that you need to fix first.

1) You have a single-ended supply so both + and - op-inputs need a 1/2Vcc reference. The op-amp output would then be centered at 1/2Vcc.

2) If VS2 2.5 is your center reference then ignore line 1 and connect R7 to VS2. Make sure VS2 has a bypass capacitor (10uF)to ground.

3) Option; R7 is shorting out your 2.5 volt reference as the op-amp tries to keep both inputs at the same voltage. Omit R7 and make R2 6.8K 1% to get the same roll off frequency.

4) To get an exact gain of 100 make R1 680K 1%

5) The LM324 is a very low quality op-amp with a high bias current so resistors much above 100K will tend to act like a lower value. It has a poor slew rate so RF frequencies are out of the question.

6) If your signal source is of low impedance (<= 600 ohm) then you can divide your gain and input filter by ten and get better output from the LM324. C1 is now 47nF, R2 = 680 ohm 1%, R1 = 68K 1%. The filters at the output of the op-amp look ok. You can make C4 100uF if 10uF is too low a value to pass 440 HZ into a 50 ohm load.

7) About C4; I will give you a list of roll-off (-3dB) frequencies as a high-pass filter, more than just a DC blocker for single ended supplies. This assumes R6 is 49.9 ohms 1%. If C4 is 10uF it will NOT pass frequencies below 31.894 KHZ. If C4 = 100uF the roll-off point will be 10.086 KHZ. If C4 = 48,000uF the roll-off is 460 HZ.

8) With what you have now C4 is all but killing your 440HZ signal. Since R6 needs to stay the same for impedance matching I would move C4 back to the output of the op-amp with a resistor in parallel with C3. Now C4 can be 10uF and R4 can be 3.3K 1% with a roll-off at 482.28 HZ. Leave your output values the same (R5, R3, R6, C3).

NOTE: If you want a much more quiet low-cost op-amp with gain up to 50KHZ then I suggest using a TL051 JFET op-amp. However it will need a bipolar supply of +/- 6V to +/- 15 V. You would then ground pin 3 of the op-amp, as ground is now your center point. You could also omit C2 (no HF compensation needed) and C4, as you have no DC bias in the output to block unless you need a Hi-pass filter function at the output.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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