Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

“Hey! Turn it down!”

You’ve probably heard these words at some point, especially if you’re like me and enjoy listening to music LOUD! If you decide to oblige and actually “turn it down” then what must you do? Reach for the volume control, of course!

What is a volume control? Put simply, it’s a circuit which attenuates the amplitude of an audio signal. Usually the desired behavior is for the amount of attenuation to vary from none, down to some high level of attenuation so that the audio is too quiet to hear. Many audio devices are digitally controlled, but there are still plenty of systems out there with a knob that you can rotate in order to adjust volume.

Figure 1 - The Potentiometer

This knob is actually an electronic component called a potentiometer, or pot, as shown in Figure 1. Pots are a special type of resistor with a third terminal, called a wiper. The wiper moves back and forth as you rotate the knob, changing the amount of resistance seen at the wiper terminal. This creates a voltage divider kind of behavior, making pots ideal for use in volume control circuits.

However, not all pots are created equal. The three most common types, or tapers, are the linear taper, logarithmic (or audio) taper, and reverse logarithmic (or reverse audio) taper. Figure 2 shows how the resistance of each taper changes versus rotation.

Figure 2 - Typical Potentiometer Characteristics

Human hearing spans a large dynamic range, so logarithmic units of decibels (dB) are typically used to describe the power of audio signals. An ideal volume control would have an attenuation characteristic which is linear in dB, since the result is a very natural change in volume as you rotate the volume knob. For this reason, audio taper pots are often used as volume controls. However, as you can see from Figure 2, these pots are not actually logarithmic, but rather a combination of two linear pots with different slopes. This does a poor job of achieving linear-in-dB attenuation, but thankfully there’s a better way!

 Figure 3 - Baxandall Active Volume Control

Back in 1980, Peter Baxandall presented a circuit which has become known as the Baxandall Active Volume Control, shown in Figure 3. It’s simple, flexible, and achieves excellent linear-in-dB attenuation performance with a linear pot by implementing a clever type of shunt feedback.  The high input impedance and low output impedance of this circuit ensure that its attenuation performance is maintained no matter what kind of source and load impedance is connected.

 Figure 4 - Baxandall Active Volume Control Analysis

The transfer function of this circuit can easily be analyzed using Kirchoff’s current law and Ohm’s law. I won’t go through the full analysis here, but this is the result:

 An important note about this transfer function is that Rx represents the percentage of pot resistance after the wiper, and R1-x 
represents the percentage of pot resistance before the wiper. If you plot this transfer function versus percentage of pot rotation, as shown in Figure 5, you can see that a nearly perfect linear-in-dB attenuation characteristic is achieved. The attenuation drops off sharply as rotation goes to zero, since the transfer function approaches negative infinity. This is actually desirable, since in a real circuit this would provide a very low “off” volume which is inaudible.

Figure 5 - Calculated Volume Control Response

The Baxandall Active Volume Control circuit is used as the basis of a TI Precision Design: Active Volume Control for Professional Audio. In this design, I provide a detailed analysis of the theory behind this circuit, as well as full simulation and real-world results and all the circuit board schematic, layout, and bill of materials files. If you have an interest in audio, I highly recommend that you check it out!

Anonymous
  • Ian,

    I just plotted the transfer function in Excel on a linear graph and it looks exactly like what I want for controlling the brightness of an LED. The non-linearilty is adjustable by changing the gain, which is an excellent feature that was not mentioned. The only problem for me is I need a circuit that works off a single supply. Have you looked at the effect of adding an offset voltage at A2?

    Thanks,

    Wayne

  • Hello George,

    Thank you for taking the time to read and comment on this post, and I appreciate your perspective as someone with years of experience in the high-end audio industry.

    You may be correct that "no one complains about the lack of a perfect log response in an audio pot." I'm not an audio end-equipment design engineer, so I don't hear that kind of feedback. While not every audio pot may have the kind of two-linear-taper characteristic mentioned here, many in fact do, and this design fully solves that problem.

    I also believe that amplifiers are necessary in order to solve various issues which are present in audio designs, such as buffering, filtering, and power amplification. For example, a colleague of mine designed an active crossover circuit for two-way studio monitors, detailed here: http://www.ti.com/tool/TIPD134. 10 amplifiers are used on that board in order to achieve the desired performance, not to mention the power amplifier in the following stage. My two-amplifier volume control solution is also incorporated into the design before the crossover. Despite all this silicon, these speakers sound great and have outstanding measured performance at every stage!  

    Thank you again for reading! Perhaps I can run some future design ideas by you in the future.

    Best regards,

    Ian Williams

  • I've been in the hi end audio design business for most of my life. This is a solution to something that ain't broke.

    A good quality audio pot worked beautifully for many years and still does. No one complains about the lack of a perfect log response in an audio pot. The curve is not necessarily cast in stone as 2 linear tapers. There are pots available that have a log response.

    There is also the presumption that these op amps don't alter the quality of the audio. They do, no matter what the specs are. My experience is that less circuitry = more audio, 2 buffers and 2 amplifiers in series with the audio will eat up low level information - reverb tails, ambience, room tone, all part of the realism of a good recording. This is subjective evaluation.

    This is essentially a variable negative feedback amplifier and the amount nfb changes the gain bandwidth product. While differences may measure outside of the audio band, to good ears it changes the quality of the presentation.

    As for the quality of the pot itself, these vary widely and the contact quality and the resistance track quality determines most of the performance quality. So unless you have a pot for this circuit that has high quality contacts and resistive medium, then you are no better off than having the same qualities in a standard audio pot.

    So in essence we've added 9 components and additional power supply requirements for a problem that a professional quality or audiophile quality potentiometer doesn't have.

    So if it ain't broke, don't fix it.

  • Hi Ahmet - I'm glad you enjoyed the post!

    Hi Stephen - you're very welcome!

  • Hi Ian

    Thanks for clarifying this for me.

    regards

    Steve