Note that I've linked to manufacturer pages for the few chips that I use; you can download the PDF datasheets from these sources. A scanned graphic of the handwritten schematic is linked off the title above. In the next lecture, you'll learn how to commit rough schematics such as these to the computer with Protel EDA.

This circuit runs on a single +5 Volt supply and drives a speaker directly. The sound that's heard comes from a pair of 555's. The frequency of these signals is changed periodically by the stuff in the rest of the circuit; one of the 555's is occasionally hard-reset and held off. Counters are used with slower 555's to generate bits, which are in turn used to reset, trigger, or perturb other sources and processes. The 4051 multiplexer is used to provide more bits at its output, as selected by lines from the counters and 555's. More detail below...

The only two IC's that make audio waveforms that are actually heard are the two 555 chips U1 and U9. The output of U1, however generates subharmonic octaves in the 4024 binary counter U6; one of these is selected for the audio source by CMOS multiplexer U10 (a 4052). Pin 5 of the 555 is a voltage-control input; a voltage applied here (between ground and Vcc [here +5 Volts]) will modulate the frequency of the 555 astable oscillator (freq. goes down as the voltage goes up) or change the width of the pulse produced by a 555 wired as a monostable "one-shot" (pulse widens as voltage goes up). U1's frequency is also voltage-controlled (accordingly at pin 5) by the slow linear waveform output by U2 (consider it a triangle wave, although it is actually an attacking and decaying exponential from the charging and discharging of the 1 uF timing capacitor). The signal from this capacitor must first be buffered by voltage follower U4, however, before driving pin 5 of U1, as it is at very high impedance.

U2 is also voltage controlled by the pulse wave coming from the divide-by-127 output of binary counter U14, which is driven by the 555 U9 (which makes the other audio waveform). The frequency of U9 varies with the state of multiplexer U12 (a 4052), which switches different timing capacitors in and out. U2's frequency is also voltage controlled by the buffered (in voltage follower U5) ramp (actually it is an exponential transient) coming from the 555 U7, which is configured to be a "one-shot" or monostable multivibrator (e.g., instead of continously oscillating, it gives one long pulse [or "ramp" in this case]) upon receipt of a trigger [a negative-going pulse on its pin2]). This ramp also modulates the frequency of 555 U11, a slow oscillator (e.g., LFO or "low frequency oscillator") that changes the behavior of U15 (more below on this). U3 is another LFO that clocks binary counter U8, producing state changes at its various outputs. One of these lines puts U9 into ocassional reset (muting its audio output). Other lines on this counter are used to select the subharmonic on U10 or switch U9's timing capacitor on U12. U15 (the 4051)is an 8-line multiplexer, with the selected line determined by outputs of conunters U14, U8, and U6. The input of this multiplexer is defined by counter U6, and the analog reference level is switched by the LFO U11 (toggling the analog reference on a CMOS mux in this way is highly unconventional, and results in some interesting effects; but remember, it's not standard practice!). Depending on the signals at its address lines, input, and Vref, different outputs of U15 will be addressed and change state. These outputs are used to do various things in the ciruit; e.g. toggle other muxes, reset counters, etc.

U13 is an inverting summing amplifier that mixes the U9 and selected U6/U1 outputs, then routs them to a speaker through emitter follower Q1 (a TIP31 power transistor). The noninverting pin of U3 is biased up so the OpAmp circuit will work with a single +5 Volt supply (assuming 0-5 Volt logic signals at the input). Note that this output stage isn't recommended practice, as Q1 provides an average nonzero DC voltage at the speaker, which burns it up in the coil (if the voltage were any higher, "burning it up" would have a more real connotation...). We can get away with it for the little 5 Volt circuit depicted here, but ultimately, we need to isolate the DC output of Q1 from the speaker with a large capacitor (e.g., a 200 uF electrolytic), with positive end pointed toward Q1. Alternatively, the speaker could be driven differentially relative to the bias voltage (vs. having one end of the speaker grounded, as it is now), or a split supply can be used for the OpAmps, enabling the DC to be blocked by putting capacitors at the summing input resistors of U13 (then we'd need a complementary pair of output transistors for Q1; e.g., an NPN and PNP configured as push-pull buffers). Note that we're using a low-voltage, quad, single-supply OpAmp here, the TLC274, which can pull to the negative rail. A better choice would be the TLC2274, as these can pull to both the positive and negative rail, giving more headroom for positive signals. Again, for this application, this is of little consequence in the end, as the TLC274 gives enough dynamic range.