[sdiy] LFO question

[Magnus Danielson]

From: ChristianH <chris at scp.de>
Subject: Re: [sdiy] LFO question
Date: Wed, 29 Jan 2003 12:21:36 +0100

> Hi,

Hi Christian,

> just a foggy idea: what about using a micro controlled current
> source/integrator, in order to fill the steps. Since the controller
> always knows the desired frequency, it could adapt the slope to match
> the step size. That would make filtering for a smooth output easier.
> 
> However, using only a current source to generate wave forms (without any
> feedback into the controller) would most probably suffer from symmetry
> problems, causing DC level to run away.
> 
> And using 2 DACs for coarse voltage output and a DAC controlled analog
> linear interpolation would only be justified if you're going for a more
> complex LFO than the ordinary tri/sine/square one. That is, unless you
> have multiple cheap DAC ports anyway, e.g. PWM generated analog outputs.

What you could do is this...

You have a DAC which feeds an integrator. The frequency is set by the MPU by
the time it changes the DAC output current. It will be easy to do MS-20 style
LFO, i.e. adjusting slope balance. By adjusting the current amplitude you would
automatically allow yourself to control the LFO depth, but the downside for
doing that is that at lower frequencies the depth-control will not have much to
work on. A CMOS mux could be used for coarse depth, letting the finetune go
into the DAC and away you go.

Interesting idea there... keep that thought!

Cheers,
Magnus