i was playing about with a modular synth recently and got carried away with the lfo making nice dubbie sounds, which was a waste considering it was a beautiful 20 unit doepfer system that i’ll not get to play with again for a loooong time!

i wanted something with a bit more meat to the sound than the usual 555 timer sirens, and more control over the lfo, as most of the sirens i’ve built just cut the sound, leaving the lfo running in the background, and i wanted to be able to reset the lfo each time so i could start it at the top of its wave pattern.

the lfo consists of a pretty basic integrator and schmitt trigger that produces a triangle wave (and a square wave, but its not utilised here) from two op amps. the lfo has speed control, a switchable rate capacitor, and depth control on its output. this feeds another fairly standard, but widely used, vco seen in lots early 80′s synths such as the amdek pck-100 and ds7 drum synth, and is practically the same circuit as described in the lm324 data sheet, but with a few component changes. the pitch of the vco is modulated by the ouput of the lfo plus a separate tone control pot, and has square or triangle outputs, and finally a master volume on the output of the vco.

this gives us the basic 4 controls of a dub siren, speed, depth, tone, volume, plus an lfo rate switch and vco wave pattern selector. it would be quite easy to add an extra cv input to the circuit, and when hooked up to the ripple synth, it makes some weird sounds!

the circuit requires a dual power supply, but i used a voltage bias circuit to create a virtual ground at 4.5v, when run off of a 9v battery or dc input.

the original design of the lfo and vco together worked great, but there were nasty clicks from the vco when triggering it. it was explained to me that one of the reasons for this was the vco wave form was creating a voltage spike at the output as it was starting mid wave so to speak and the output was jumping from ground to a higher voltage. it made sense to me at the time, but i think i’ve explained that badly…

the solution was either to reset the vco as well as the lfo each time so as not to create these spikes, or to add a voltage controlled amplifier. the first option worked to a degree, but the output still clicked a little too much for my liking when triggering.

i adapted the basic vca schematic from the lm13700 data sheet with a little help from dave, and hooked it up to a simple envelope generator triggered from the same switch as the lfo reset.

the result was perfect. the clicks were taken care off by the vca, as it controls the volume output of the vco, and by fading the volume in, even very quickly over a fraction of a second, it takes the clicks and voltage spikes out of the equation by rising the output from ground instead of a sudden jump.

the reset function allows the lfo wave to start at its highest voltage, and can also be switched off so that the lfo keeps running when the vca’s output is low.

the lfo’s triangle wave doesn’t actually go the full distance between ground and 9v when it sweeps, so by reseting the lfo, it starts outside of this confine, giving longer sweeps on its first fall before settling into the dialed in parameters.

if you remove the reset function, you can add pots to the envelope generator instead of fixed resistors. this will allow you to control the attack and decay times of the vca. which is good fun, but only works without the reset as the decay is triggered by releasing the trigger switch. this causes the output of the lfo to go low, and silencing the vco before it has chance to fade out with the decay.

a spring loaded dpst toggle switch was used to trigger the vca’s envelope and the lfo reset. the lfo reset is a two position toggle, with down and off settings, and the two position toggle switch controls the vco output from a nice warm triangle to a harsher square wave.