I am working on a board design which is very limited in space. I am taking input from a passive guitar pickup and sending it through a high impedance buffer op-amp, a low pass filter, and then a non-inverting op-amp. The final output is a cleaned up, half-wave rectified, and amplified version of the original signal. This signal will be measured by the ADC of an Arduino, which will use the intensity of the received signal to do some things.

At no point will the signal go to a speaker, and I do not care at all about the tonal quality of any audio signal.

enter image description here

From my research, I have read that film capacitors are best for audio DC blocking applications. In this case this is capacitors C9 and C10. As space is limited, I'd like to use a different type of capacitor since film capacitors are quite large.

Since I don't care about the quality of the audio in any way, is it acceptable to use a non-film capacitor? If so, which type of capacitor would be a good space-saving alternative?

Edit: I should probably mention the caps in question (C9 and C10) are 1uF.

  • 4
    \$\begingroup\$ C10 makes no sense at all there. \$\endgroup\$
    – tobalt
    Jan 24, 2023 at 6:28
  • 1
    \$\begingroup\$ Why worry about the audio quality of film caps if you're not concerned about audio quality? Your question is a bit odd. Anyhow, for practical purposes in a guitar circuit like this, any capacitor will work fine for you. It's only the cork sniffers that think they can tell the difference. Use ceramics, they'll be fine. FWIW I play guitar and fix amplifiers/pedals semi-professionally. \$\endgroup\$
    – Kyle B
    Jan 24, 2023 at 7:01
  • 1
    \$\begingroup\$ If you use ceramic caps, NPO or C0G offer the fewest surprises. But compared to other ceramic caps they are only available in small capacitances. If you can't find NPO in the capacitance you need, you can maybe get by with X7R ceramic caps. All ceramic caps have some shock sensitivity. If you tap it, or drop it, it will convert the mechanical impingement into a voltage pulse. This could be bad for some audio applications. \$\endgroup\$
    – user57037
    Jan 24, 2023 at 7:26
  • 1
    \$\begingroup\$ @mkeith (It's NP0 (zero), the number stands for temperature coefficient) \$\endgroup\$
    – Lundin
    Jan 24, 2023 at 7:34
  • 1
    \$\begingroup\$ @J.P. It has to be DC-biased not to saturate the opamps. That is what C9 is there for. It biases the AC signal to a DC level of VCC/2. After your circuit, you can remove this bias again (i.e. after U3C), by adding a capacitor there and a resistor to the new DC level (e.g. GND). If you remove the DC bias before U3C, you will saturate U3C. Also, you should place a large capacitor in series with R8, too, to make it work well. Without this additional cap, U3C will also amplify the DC bias (which you don't want, I suppose) \$\endgroup\$
    – tobalt
    Jan 24, 2023 at 15:02

1 Answer 1


I don't see why you need more than 0.01uF for each. With the 2.2MΩ resistors you'll have a cutoff frequency of about 15Hz which is below audible frequencies let alone those from even a bass guitar.

Pin 10 has no DC path to any particular voltage. You probably want something like 1MΩ to ground. As it is it may drift upwards to Vdd or downwards to GND though variations in pin 7 will cause it to change; pin 10 voltage will be bounded by the supply rails +/- a diode drop.

For the 10nF you can use C0G or X7R ceramic capacitors, and probably the latter would be fine for your application where you don't care about signal quality.


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.