2
\$\begingroup\$

I am trying to design an op-amp input stage to capture a guitar through an ADC, specifically the ADC on an Arduino Due (3.3V, 12 bit). I am aware there are numerous ways to configure and bias the op-amps. I have designed a simple circuit based on a inverting amp followed by an active LP filter. The system will be running off a 9V battery and the power will be provided to the op-amps and uC (regulated) from this.

The aim being:

  • Bias the signal around Vdd/2 (mid of ADC input range)
  • Amplify the signal to use best of the ADC's range
  • Lowpass filter the input signal at Nyquist rate (fs/2) to avoid aliasing

However, I have been unable to find much information on what the benefits or drawbacks of different configurations and methods of biasing are, specifically for this kind of signal. I have not yet measured the voltage output of my guitar so the specifics of the required gain properties are not known yet! Aslo, will I require a high input impedance for a guitar?

Any advice or comments would be greatly appreciated.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ Updated schematic to bias second op-amp \$\endgroup\$ – CatsLoveJazz May 9 '13 at 20:25
1
\$\begingroup\$

Generally a guitar amp would have about 100k (or greater) input impedance - this is because the tone controls and volume controls are about that sort of range. Yours has 10k input impedance. I'd make R8 100kohm and R7 1Mohm

Your 2nd stage is not needed - it has unity gain for the relevant frequencies and your filter cap might just as easily be placed across R7 but obviously 100x lower in value because R7 is 100x bigger than R10. But having said R7 needs to be 1Mohm (above) the cap needs to be 1000x lower like 15pF.

I'd also use an input jack socket with an earth switch integral then you can disconnect the battery when not plugged in.

\$\endgroup\$
  • \$\begingroup\$ Thanks so much, that's very helpful info for me! Out of interest, what would would the pros/cons in using a non-inverting configuration instead be (other than the signal would be inverted which can easily be taken care of in software) ? \$\endgroup\$ – CatsLoveJazz May 9 '13 at 21:50
  • \$\begingroup\$ I'd probably use a non-inverting config except for the fact that your resistor to V- input would go to the centre point of a poor mid-rail impedance wise. It's swings and roundabouts. Inverted signals are neither hear nor there - no problems inverted or not-inverted \$\endgroup\$ – Andy aka May 9 '13 at 21:52
1
\$\begingroup\$

Many problems:

  1. You forgot to power the opamp! This is not a rail to rail opamp, so the power scheme needs to be thought about carefully.

  2. You say you have a non-inverting amp followed by a active filter, but that is not what the schematic shows. Both stages are clearly inverting.

  3. You mention a 3.3 V supply, but then the schematic shows a battery used only for setting the bias point of the first amp. This seems needlessly complex and klunky. It would be simpler to use the 3.3 V supply directly to derive the bias voltage from. The TL081 has very high input impedance, so this could be a voltage divider drawing only a few µA, which wouldn't put any meaninful burden on the existing supply.

  4. The bias voltage divider is correct for a 9 V source, but "9 V" batteries vary a bit. The error scaled down to half of 3.3 V shouldn't cause too much trouble, but why use a separate voltage source and then divide it down to try to match 1/2 the supply? It makes more sense to use the supply itself with two equal resistors, like 100 kΩ each, to make the bias voltage. See point 3, above.

  5. The second amp is biased at 0 with a DC gain of -1. That will produce negative output voltages, which is unlikely what you want.

  6. The low pass fillter has about 10 kHz rolloff. That implies a sample rate of 10s of kHz unless you know something special about the frequency content of the signal. Do you really intend to sample at 50 kHz or so?

\$\endgroup\$
  • \$\begingroup\$ Thanks for the comments! 1 - I neglected to mention that the system will be running from a 9V battery eventually and the uC will be powered by a 3.3V regulator! 2 - I meant inverting, sorry 5. I presume this means I have to bias the V+ terminal of the filter op-amp too? 6 - I would like to sample up to this frequency as the guitar will produce frequencies this high - I thought! \$\endgroup\$ – CatsLoveJazz May 9 '13 at 20:16

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.