0
\$\begingroup\$

I made this preamp circuit for a dynamic mic and am mostly happy with it, but I'd love to get suggestions on how to make it a little better. I'd like it to have as little coloration as possible, and it's not as flat as I would like. Also, I'm not familiar with specific op-amps past the NE5532 and LM741. Would there be better choices here?

Here the link to it if anyone wants to mess around with it/simulate: https://www.circuitlab.com/circuit/7fekr56s3dtt/preamp-circuit/

enter image description here

\$\endgroup\$
18
  • 1
    \$\begingroup\$ You have use a 741, have a total gain of 1000 or 60dB, divided into buffer, 20x gain, 50x gain, buffer, and passive tone control. Can you please ask a specific question, instead of just asking for suggestions for improvements - which likely would end up being endless list of redesigns of the whole circuit. \$\endgroup\$
    – Justme
    Commented Sep 21, 2023 at 17:18
  • 5
    \$\begingroup\$ Using 741 op-amps is the most glaring problem given your supply voltages \$\endgroup\$
    – Andy aka
    Commented Sep 21, 2023 at 17:19
  • 1
    \$\begingroup\$ Use a low noise op-amp suitable for your supply voltage. Get as much gain in the first stage as possible for noise reasons. Why the two buffers? If the wiper of R19 is set all the way to the left, you're likely to have an oscillator. \$\endgroup\$
    – John D
    Commented Sep 21, 2023 at 17:27
  • 1
    \$\begingroup\$ Find one that can run from about 5 volts to 15 volts and with decent bandwidth whilst maintaining fairly decent output rail-to-rail capability. \$\endgroup\$
    – Andy aka
    Commented Sep 21, 2023 at 17:27
  • 2
    \$\begingroup\$ @EliY I'm no expert. But some notes: Douglas Self says for moving coil microphones (one the market in 2019) "the output range is from 40 to 2500 μV, a remarkably wide span of 62 times or 36 dB." (This is spec'd at 5 cm/sec velocity.) He also notes that moving coil mics fall into two areas of a bimodal chart suggesting manufacturing design clustering. The SM57 datasheet specs sensitivity. Do you know what range of SPL you expect to handle and what you need to get at the output for the highest SPL you need to handle? \$\endgroup\$ Commented Sep 21, 2023 at 17:57

1 Answer 1

1
\$\begingroup\$

Putting a variable-frequency lowpass filter between the last opamp output and the circuit output means that the output impedance seen by downstream circuits will vary as the frequency is adjusted. It also will vary wildly as the output level is adjusted. A better way to go is to have everything done before the final voltage follower, so you have a consistent, low output impedance driving the outside world. Move OA4 to the R20 wiper.

For a little bit of circuit protection, it is common to have a 100 ohm resistor in series with the output. Yes, this raises the perceived output impedance, but the value still is low; and, it is constant.

Assuming you change to different opamps, there is no reason for OA1 (and C4, R4, and R6).

For commonality of parts, you can change the gain of both stages to 33.

National Semiconductor (now a part of Texas Instruments), Burr-Brown (also part of TI), and Linear Technology (now a part of Analog Devices) make "audio opamps" with noise and distortion specs that border on witchcraft.

And finally, consider this -

Two cascaded gain stages is the right way to get the gain-bandwidth product you want. But with modern devices, there is no reason for the input and output voltage follower buffer stages. If you move the lowpass filter and output level adjustments to be between OA2 and OA3, the circuit reduces to just two opamps. (Connect C14 and R20 to Vref rather than GND.) Your feedback network impedances are high enough that the opamp output stages will have no trouble driving both them and the next stage (or output).

\$\endgroup\$
5
  • \$\begingroup\$ This is very in-depth and helpful. Thank you! \$\endgroup\$
    – Eli Y
    Commented Sep 21, 2023 at 18:24
  • 1
    \$\begingroup\$ An audio preamp design is pretty much a rite of passage for a circuit designer. My first was in the early 70's for a TV station announce booth, before the 553x opamps were invented.. Two 4-tranistor diff amp gain stages followed by a true Class-A output stage. Measured very well, sounded great, and ran for decades. \$\endgroup\$
    – AnalogKid
    Commented Sep 21, 2023 at 19:14
  • \$\begingroup\$ That's amazing! And good to hear that I am on the right track. I've been a mix engineer for a while now and I have major imposter syndrome when I don't understand exactly how gear works on a circuit level. If you have time... Is this more along the lines of what you were thinking? circuitlab.com/circuit/y3a9nxqcf8dm/… \$\endgroup\$
    – Eli Y
    Commented Sep 21, 2023 at 20:35
  • 1
    \$\begingroup\$ Yup. Some might not like having a gain stage after a lowpass filter, but newer opamps are so quiet that I don't think it is a problem. If you don't need the output amplitude of a rail-to-rail output stage, there are better parts than the AD822. Golden-eared audio nutjobs love the OPA2134 for its ultra-low distortion, but at 20 kHz and 30 dB of gain, it has only 20 dB of gain headroom The 5532 has better gain, bandwidth, and noise specs, plus excellent distortion performance. \$\endgroup\$
    – AnalogKid
    Commented Sep 21, 2023 at 21:45
  • \$\begingroup\$ Thank you so much! All of this is incredibly helpful \$\endgroup\$
    – Eli Y
    Commented Sep 21, 2023 at 22:26

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.