Why square waves are important in audio

July 7, 2021
 by Paul McGowan

18 comments on “Why square waves are important in audio”

  1. So Paul, correct me if I’m wrong here.
    Basically, you’re saying, any amount of distortion with even a 3V input, of an amplifier, will distort and meaning, the square wave, will end up looking like squiggles.
    Is this correct?

  2. Shouldn’t every oscilloscope most precisely display a square wave? Thus what do audio designer not understand when their designs of preamps and power amps are unable to reproduce a clean square wave? Could it be that the mysterious voicing process of an audio chain leads to adding distortions which have to mask the inherent and audible imperfections?

  3. It makes sense to test circuits using a square wave, but here are a few, mostly quite pedantic, observations:

    A square wave is not normally created by adding odd harmonics to a sine wave fundamental. A square wave test signal is more conveniently created in a simple function generator that periodically and regularly alternates between two fixed levels. It is quite correct that the spectrum of a square wave consists of a fundamental frequency plus odd harmonics, but that’s not how we generate it.

    Music is not really modified sine waves and, if it were, modified sine waves are not sine waves. The waveform created by an instrument, or a voice, is generally complex and determined by the physical nature of hardware, meaning the strings, hammers, bows, pipes, tines, soundboards or larynx as appropriate. Occasionally we have a nearly pure tone from (for example) a flute or a trained singer that approximates a singe frequency, but this is not the general case.

    A square wave is a poor way of assessing non-linear distortion, because the signal consists of only two levels. It tells us nothing about non-linear behaviour.

    1. While you’re correct in the way a square wave is generated in a function generator, Mark, I would argue with you on your last sentence. Square wave performance in a circuit tells us a great deal about how that circuit is behaving. For example, ringing of the square wave is almost guaranteed to translate to harsh sound: a brightness that follows any transient musical events.

      You’re correct in that square waves do not exist in music but as a tool, this waveform is very helpful.

      1. It makes sense to test a circuit using square waves, and I wasn’t disagreeing with your basic position here. One thing that a square wave doesn’t help us with is characterising a non-linear transfer function, since the test signal exercises only two points on the curve.

          1. Good afternoon Paul!
            Now I have a better understanding of what you said about square waves both here to Mark, and on your video.
            If I said that I didn’t miss working with tools like acello scopes, I would be lying.
            I tell you this because, I do miss working with tools like the acello scope.
            Sure I trained my wife on how to replace tubes.
            But that, how does a blind man teach someone to work with a tool like the acello scope, if he can’t see it?
            Or in less, they make an acello scope that a blind person can use.
            If you know something about that, please let me know where I can go, to pick one up.

              1. Good morning Paul!
                Please correct me if I’m wrong.
                I happen to know anymore these days, testing tools are being used with computers.
                If an acello scope can be connected to a computer to look at square waves and other signals, there has got to be a way to bring a brail display in to that equation.
                Because if a brail display can help a blind person pen point exactly where they are in a document or something other on the computer screen, wouldn’t you be able to trace a signal by using a brail display to look at the acello scope?

          2. I would add that as long as the output square wave is kept below clipping, it illustrates a number of things. The rise and fall times give an idea of slew rate. If there is a spike at the leading and falling edge of the square wave, it could suggest a marginally stable feedback circuit. There are other symptoms that might be determined from the wave shape, but no need to beat it to death.
            One reason an oscilloscope can display a near perfect square wave is its bandwidth. Typically dc to 50mHz these days.

        1. Well, square waves aren’t harmful to your speakers (mostly tweeters) but clipping is. What’s confusing about that is that when an amp clips it makes a square wave. Clipping is bad for a number of reasons but mostly because it is at full power and because it happens at the amp’s output it generates a shit ton (technical term) of higher harmonics at very high power levels.

          1. That is what I meant Paul. Square waves that are caused by clipping. I think it creates a continuous high power output to the tweeter that tweeters cannot handle. Is there no way to filter out that damaging square wave caused by clipping so it never reaches the tweeter?

            1. Several amps have been built over the years with limiters on their front end to prevent the amp from going into clipping. I don’t know of any way other than limiting the amount of input so you never clip. It’s really only a problem with small amps. Big amps are never a problem as if you are getting to close outputting hundreds of watts you’re probably doing other damage long before clipping occurs.

            2. Good morning Joe!
              I don’t know if this is still being practice today, but there is a way to stop clipping from frying your tweeters.
              But what’s happening, is a good deal of direct currant is being shot up there to the tweeters when the amplifier is being driven to clipping levels.
              But a long time ago, some speaker manufacturers, used audio transformers as apart of the crossover network to rectify that problem.
              I don’t know if this makes any sense to anyone, but there it is.
              That is the best of my understanding of things like this.

  4. What wasn’t mentioned is the frequency response of a square wave. Lots of amps do OK at 1 kHz but not so good at 10 kHz much less higher up.

  5. Little known fact:
    Square waves were originally designed back in the early 80s by Colecovision’s Eric Bromley to be used in an 8 bit surfing video game.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram