[sdiy] Idea - Triangle wave DCO core

[Mattias Rickardsson]

On 14 March 2012 22:14, Scott Nordlund <gsn10 at hotmail.com> wrote:
> Is the offset even a problem? Won't it just drift to the rail, then clip a tiny bit each cycle (equal to the drift over one period) and be stable from then on? I'd think the only requirement would be to design it so that the triangle wave spans the full range of what the integrator can put out.

Maybe the triangle wave could be limited by a pair of minimum/maximum
voltages (say +/- 5V) where no more charge builds up in the capacitor,
but kind of spills over like running water would do in a bucket when
it's full? :-) The oscillator should then be calibrated to just about
touching these limits when running normally, and it would then run
almost identically even if some component values would change with
time or temperature, or if the charge current would end up too high.
For such an overflow mechanism I guess we'd mostly need a diode and
some voltage reference... or do I need a stronger cup of tea this
wonderful monday morning?

/mr