AW: Re: AW: 2-comparator sawtooth genera

[Haible_Juergen#Tel2743]

> The overshot is not slight in this case. An LM311 will *never* be able
> to stop the discharge of the cap in time (2.2nF, 3V, 50ohm). A 
2-comparator
> circuit is not useable for a sawtooth generator where the goal is an as
> short as possible discharge time, I think.

True. I never aimed for maximum discharge speed. As I said, if *that* is
the goal, a thyristor solution might be the best.

> I had not thought about the current through the switch until now,
> but I guess it exceeds the 4066's specs. 12.5 ohms and 5 volts gets
> you 400mA initial current, which is rather large.

That's the point! I never considered devices to carry such a current!
And If You use low current devices, the discharge time is longer, and
the overshoot is smaller.

But even with those fast discharge times, the worst thing You could get
is a *full* discharge of the capacitor - but the circuit would still work,
though with increased amplitude. As I tried to show in my previous mail,
this overshoot, too, would not effect the stability of the VCO in a
negative way.

Systemizing the different possibilities, I could think of the following
alternatives:

(1) Discharge slower than comparator response => works as intended,
     saw cycles (almost) between the two comparator thresholds.
(2) Discharge slightly faster than comparators => some amount of
     overshot, but nevertheless stable amplitude and frequency.
(3) Discharge much faster than comparators => full capacitor discharge.
     Saw cycles between reference voltage of cap and the 1st threshold.
     2nd threshold is without effect, but the oscillator still works ok.
     BUT we have to take care that the usual buffer opamp is capable
     of the whole saw range (i.e. up to ref voltage! This point is a little
     unpleasant, I have to admit, 'cause the idea of the 2-comparator-
     design *was* the ability to use one supply rail as reference voltage 
...)

So, Rick, thanks for Your questioning. I see, the whole thing is not
as straightforward as I thought. I guess it worked so well in my circuit,
because I never used high discharge currents.

JH.