Practical Circuitry (Part 2)
Tricks With The SN76477
Last time we talked about how to find new outputs on the SN76477. In this installment, let's switch gears and talk a little about how to get more sounds from the noise source.
The white noise source in this chip is not the back-biased transistor you might be familiar with, but a binary pseudo-random noise generator. I don't want to get too heavily into the theory of this (see Don Lancaster's TTL Cookbook, Howard W. Sams, 1974, pp 277-283 for a good treatment), but all we really need to know is that the noise source is basically a shift register, with various bits moving down the register at a rate determined by a master clock. This clock can be either internal or external to the chip.
Let's consider an internal clock circuit first. Normally the resistor connected from pin 4 to ground sets the clock rate, and that's that. However, we could just as easily replace the resistor with a transistor (see figure 5). In this configuration the transistor acts like a throttle, and determines the current flow through pin 4, hence varying the clock rate.
Before getting too heavily entrenched in details, I suppose that I should say something about why we want to change the clock rate. The best answer I can give is "try it, you'll like it!". The sound is an incredible swooshing noise, and is very similar to phasing or flanging. The noise takes on a new tonality, and sweeping the clock changes the spectra in an eerie and dramatic manner.
The circuit in figure 5 is specially adapted for a percussive voice, with the noise amplitude being controlled by the envelope generator and VCA. In addition, the envelope is tapped off of pin 8 and applied to the base of the transistor. The result is a voice which sweeps as the envelope dies away.
You will also note that the envelope is tapped via a 10 Meg resistor. The reason for such a large value is to avoid loading down the envelope generator capacitor. We could have buffered this voltage first using something like an op amp, but the loading caused by the 10M resistor is negligible, and is certainly cost-effective in this situation. Since the Beta (or DC current gain) of transistors varies from one unit to another, it may be that 10M is too large to allow the particular transistor you pick to turn on sufficiently. Feel free to experiment with other values; any value from 1M to 10M is permissible, but don't drop below 1M or excessive collector current may flow through pin 4 of the SN76477.
Before leaving figure 5, you should be aware that this would make a nice modification to the Percussive Noise Voice (John Blacet, POLYPHONY, Nov/Dec 19779, pp 12-13, corrections in the Jan/Feb 1980 issue, p 5).
Sweeping noise effects really appeal to me, and ever since discovering how to do them, I have been constantly, at work developing new ways to employ this technique. One of the limitations of the circuit in figure 5 is that the internal noise clock is slightly limited in its sweep range; when you need lots of sweep, try the circuit in figure 6.
Here we avoid the internal noise clock completely, and use an external clock instead. Actually, the "external" clock is really internal to the chip, being the VCO. This is a good way to save parts, space, and wiring hassles. However, if you were planning on using the VCO for something else, you could always clock the noise with virtually any other type of square wave oscillator.
Let's analyze the circuit; consider the VCO first. R9 and a capacitor selected by S1 (C2, C3, or C4) set the basic VCO frequency. A control voltage applied to J1 is summed through IC2 and applied to the control voltage input of the VCO, which sweeps the VCO frequency. Since IC2 is an inverting stage, an increasing voltage at J1 yields an increasing frequency. R3 offsets the VCO if desired, or can be used as a manual sweep control. Trimmer R6 sets the zero point of the VCO so that 0V applied to J1 gives the minimum VCO frequency. Diode D1 prevents any inadvertent negative voltages from creeping into the VCO. The VCO's square wave output is taken off of pin 13 and sent to the external noise clock input at pin 3. Pin 4 is tied to the +5V line, which programs the chip to accept an external clock.
The rest of the circuit is the same as that described in the last installment, with the noise taken off of timing capacitor C1. We have to take out the noise this way since the chip's normal output, pin 13, is already committed to the VCO.
To adjust trimmer R6, turn down R1 and R3, and advance R6 until the noise just begins to be audible. This sets the lowest noise frequency.
Once the circuit is calibrated, start experimenting. You could apply an ADSR to J1, or an envelope follower, LFO, sequencer, etc. - you get the picture; we're talking about a staggering amount of sounds, and they're all extremely useful. If you thought white noise had to be a static, one-dimensional effect, this circuit will definitely turn your head around.
Next month, we'll tie together what we've covered so far, plus more, with a complete project, the "Super Controller Module". I think you're going to like this one!
Acknowledgement. My thanks go to Craig Anderton for turning me on to the use of a transistor to control the internal noise clock.
This is the only part of this series active so far.
Feature by Thomas Henry
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!