[sdiy] Need help with simple VCO design

[Harry Bissell]

The main core of your oscillator is the capacitor in the feedback of the opamp...
an integrator. You inject current into the inverting input (current flow direction makes the
cap charge up/down)

In a VCO, there is a fixed resistor for you to pass the current through. Whatever the input voltage is, and that fixed resistance determines the current (therefore frequency). Ohms law
will tell you the current knowing E and R.

In a CCO, that fixed resistor is replaced with a current source. Think of this as a resistor
whose value can be changed dynamically. Make that current real small (therefore the frequency is very low - that would be as if you made the resistor tens of megohms...) or very large (that would be like tens of ohms perhaps).

For a fixed supply voltage, the range of currents is almost unlimited. You could go from picoamps
(now circuit parasitics are a problem, leakage, offsets, anything non-ideal) to AMPS (now heat is the big problem). Practical values are probably from uA (microamps) to low milliamps (above maybe 1mA the temperature starts to change.

Easy to get any current you want from a fixed supply. A fixed resistor makes the range limited
(usually)

H^) harry



----- Original Message -----
From: Julian Bunn <Julian.Bunn at caltech.edu>
To: Harry Bissell <harrybissell at wowway.com>
Cc: synth-diy at dropmix.xs4all.nl
Sent: Sun, 27 Dec 2009 13:20:07 -0500 (EST)
Subject: Re: [sdiy] Need help with simple VCO design

Thanks Harry. I'm not sure I understand the distinction between a true VCO and
a current controlled VCO (CCO?). Certainly I am limited on the input to a voltage
swing between 6 and 12V. I should make some more measurements on the real circuit
to confirm my Spice model.

Julian

Harry Bissell wrote:
> It looks like the core is a true VOLTAGE controlled oscillator.. the
> resistors in the input stage are how the capacitor is charged. Most really
> wide range designs are a CURRENT controlled oscillator.
> 
> Those fixed resistors mean that eventually you will need a really HUGE
> input voltage to get more current in the core. The up side is that you might be able to use
> really large voltages as an input, some of these true VCO designs can have as much as
> (ohh...) 50V on the input. Of course your power supply limits that completely, you never see
> more than maybe 12-15V.
> 
> I think the circuit might run to one rail when stalled, and it takes a long time to
> get to the trip point of the schmitt trigger stage. This might be similar to how a
> 555 based astable oscillator works (the first cycle starts from 0V and later cycles go between
> 1/3 and 2/3Vcc).
> 
> Thats my guess. This is a classic design but sooner or later you run out of supply voltage.
> 
> Replacing the resistors in the input stage with a controlled current source makes it have
> a much wider range but makes the rest of the core much more difficult.
> 
> My quick guesses....
> 
> H^) harry
> 
> 
> 
> 
> ----- Original Message -----
> From: Julian Bunn <Julian.Bunn at caltech.edu>
> To: Synth-diy at dropmix.xs4all.nl
> Sent: Sat, 26 Dec 2009 21:53:47 -0500 (EST)
> Subject: [sdiy] Need help with simple VCO design
> 
> This is the schematic of a VCO I'm playing with:
> 
> http://farm3.static.flickr.com/2488/4217027179_c41d9bedd3_o.png
> 
> It is (very) loosely based on a circuit that appeared in "Applied Ideas"
> in Electronic Engineering magazine in Jan 1975.
> 
> Using a Lin voltage input of between 0 and 6V, the circuit produces
> a ramp/triangle/saw (depending on the position of the "Shape"
> pot) of between ten or so Hz up to around 10kHz, which is fine
> for my purposes.
> 
> However, I don't get the range from the Log input: it flattens off
> too quickly, and I'm not sure why. I basically just bolted on the
> log current generator without really thinking too hard about it.
> 
> Another problem is that, at low input voltages, the VCO is slow
> to start - it takes a perceptible time after the input voltage is
> applied before the oscillations start - and this behaviour is
> confirmed in Spice. The time to start is much greater than the period
> of the oscillations.
> 
> I'd welcome some advice on changes/adjustments I could try to
> eliminate these problems.
> 
> Thanks,
> Julian
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> 

-- 
Harry Bissell & Nora Abdullah 4eva