I built a VCA with a that2180 using a 470nF tantalum capacitor and a 20k resistor as audio input.

As source I feed an output coming from an Arduino with a opamp which creates a 0-3.2volt output sine. The sine looks fine on a scope. I want to remove the DC offset, so any VCO can be connected, not just my own.

I connect the output of the opamp at the arduino to the that2180 cap. The cap seems to distort the sine at the input. When I bypass the cap all is fine except for the offset, I'd like to remove the offset ofcourse.

On the leftside of the oscilloscope picture you see the result when the cap is added, on the right the input when the cap is not there, so straight at the output of the Arduino opamp. (the oscilloscope is connected at the input of the vca. On the rightside of the oscilloscope picture just before the 20k resistor without the cap (and showing the correct sine but 0-3v) and on the left with the cap given the funny sine (but without the DC component)

I've tried adding a 1k resistor to ground, but no joy. I've tried using 4,7uF elco instead of tantalum, still no same quality sine. I've simulated the sine source, opamp and tantalum cap in LTSpice, which look good there. I've tested also with a proper signalgen, no issues.

What can I do to remove the dc offset and not lose the quality of the sine? Or rather why does my arduino opamp output give such a weird output when the cap is added? Should I level shift at the arduino opamp (not having a negative powersupply at the board though)



  • 1
    \$\begingroup\$ Excuse me, but what have you connected to the pin labeled Vee? Manufactures expects there's al least 4V below the ground level, the specified voltage range is plusminus 4 to plusminus 18 volts.. If it happens you have Vee=0V this cannot handle negative voltages properly. Omitting input capacitor makes input voltages have DC component and that can make the circuit to work . \$\endgroup\$
    – user287001
    Apr 4, 2020 at 22:21
  • \$\begingroup\$ Try creating a resistor divider and connecting that to the output side of the 470uF cap via a 100k+ resistor. This will allow you to vary the offset voltage and see what happens. I agree with user287001 that its probably an issue with the amplifier and not the cap. \$\endgroup\$
    – Drew
    Apr 5, 2020 at 6:54
  • \$\begingroup\$ The Vee is indeed -15v. The VCA circuit is working fine. It's just the Arduino analog out that works between 0-3.2v and I want it to work between -1.6 and 1.6v. But I was hoping the Arduino board could just create that output by itself without having the need for a negative powersupply at the opamp of the Arduino. As it looks now; it seems I'm going to have to build an opamp with a pos+neg powersupply and levelshift the signal unless you kings and queens have a better solution :-) (which i'm kinda hoping for :-) \$\endgroup\$ Apr 5, 2020 at 7:45
  • \$\begingroup\$ Where the oscilloscope was connected to for the shown oscilloscope screenshots? \$\endgroup\$
    – user287001
    Apr 5, 2020 at 11:51
  • \$\begingroup\$ the oscilloscope is connected right at the input of the vca, on the right just before the 20k resistor without the cap (and showing the correct sine but 0-3v) and on the left with the cap given the funny sine (but without the DC component) (i'll put it in the OP as well) \$\endgroup\$ Apr 5, 2020 at 14:50

2 Answers 2


You are using electrolytic capacitors for your AC coupling. Electrolytic capacitors don't perform well when they are reverse polarized.

You have an AC signal imposed on a DC offset. When you pass that through the capacitor, you lose the polarization - which side of the capacitor is "more positive" changes as the AC changes. That results in distortion.

Try a 100nF non-polarized capacitor instead of your large value electrolytic capacitors.

Try that first. See if it fixes your distortion.

If the non-polarized capacitors helps but the cutoff is too high, then you can try using a large value ceramic capacitor - but they have their own downsides.

The capacitors and the load posed by the amplifier will of course create a high pass filter. With 100nF and 20k, that will be 80Hz at most. If the amplifier has any impedance at all then the cutoff will be lower.

What you are talking about is actually a "coupling capacitor."

Decoupling capacitors are the ones you find on the power supply pins of ICs.

  • \$\begingroup\$ I'm sorry for the error indeed, I mean coupling capacitor ofcourse. (I'll edit the title) In the mean time I found a solution as well. Please read below. \$\endgroup\$ Apr 5, 2020 at 20:00

I did find a working solution: a voltage divider before the signal enters the cap.

After 'researching' my own breadboard I added a potmeter of 10k in order to take back the sine input level before it enters the capacitor (one end to ground, other end the input, wiper to cap). I took it back so that the signal measures 408mV Pk-Pk, now it's clean as can be. I measured the 10k potmeter, 1.2k and 9k, replaced the pot with resistors and so creating a voltage divider before the signal enters the cap.

So attenuating the incoming signal did help, and now the coupling of the signal works correctly. Also, changing the source to a signal generator and playing different waveforms and freqs thru the VCA causes no problems. So i'd say it's working correctly now.


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.