Shell: Theory of Operation

The Shell waveshaper module is a sequence of three operations, each with a control knob:

  • First, is a loose frequency doubler. A frequency doubler takes an incoming sawtooth wave, constructs a 50% square wave and subtracts that: the result is a sawtooth at twice the frequency, which only contains even harmonics of the fundamental wave. However, Shell has a trick here: we use amplification rather than comparator operations, and diodes, so that the curve of the doubled sawtooth has a distinct bend, favouring lower harmonics and introducing some odd harmonics.
    • The Side knob (Muddle on the 4HP panel) controls the ratio of incoming saw to square-ish, allowing more of the fundamental and odd harmonics in.
    • DC modulation of the incoming signal will modulate the pulse width and mix.
    • To probe the frequency doubler: the centre of the Side/Muddle pot is best.
  • Second, is a dampening or blanking or squelch circuit connected to alternating switches which turn on or off the frequency doubler output. This uses a window comparator on the incoming signal: in the case of an incoming Sawtooth signal it means that the centre section of the signal will be held down to 0V. When this width signal is set to, say, 33% (ie. 1/3), it means that third harmonic may be boosted (or lowered, depending on the other settings) among other effects (and other harmonics divisible by three may also be particularly effected). The output will have four zero-crossings, which means that the fundamental of the wave will not be the loudest harmonic.
    • The Gap knob (Width on 4HP panel) knob controls the width (around the 0V of the incoming signal) that will be blanked.
    • The modulation input provides mild modulation of the gap width. However, this is by design not an extreme effect.
    • DC modulation of the incoming signal will alter the width of the blanking.
    • Test point TP1 is the blanking signal. Test point TP2 is the doubled saw after blanking (though it has some bleed from TP3).
  • Finally, the inverted input is also put through a dampening or banking or squelch circuit control, switched in opposite phase to the frequency-doubled sawtooth’s squelching. This adds some slope to the middle blanked section, the inverted version of the incoming saw.
    In other words, the Gap/Width control controls an alternation between the frequency doubled version of the input saw, and the inverted saw, so that low positive and negative parts of the incoming wave are 0 to inverted on output, and high positive and negative parts of the incoming wave switch to the frequency-doubled waveshape.
    This in effect can change the relative phases of various harmonics as well as their levels.
    • The Tilt mix adjusts how much of the inverted incoming signal is added to the gap.
    • Test point TP3 is a good location to see this (though it has some bleed from TP2)

Of course, waves other than the sawtooth can be used. If a sine or triangle wave are used, you will get two blanks per cycle, as the blanking occurs around the zero crossing.

The Alt Out output is the positive portion of the In signal, but limited by the positive portion of the Out signal, for a (usually) thinner, brighter tone.

Signal Levels:

In: The circuit is designed to work best with a +/-5V incoming signal, however, as with most waveshapers, modulating the incoming level (with a VCA) is the primary form of modulation.

Out: This produces quite a range of output signal levels, for a +/-5V signal: from +/-4V to +/7V. These will vary not only by changing the incoming signal level also by adjusting the knobs.

Alt Out: This is always a lower-power signal than the Out output due to the definition of the waveshape, and on some settings be quite low-level.

Design a site like this with WordPress.com
Get started