Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!decwrl!decvax!mcnc!gatech!hao!ames!pasteur!trinity!max
From: max@trinity.uucp (Max Hauser)
Newsgroups: sci.electronics
Subject: Re: VCO waveform generators
Message-ID: <1083@pasteur.Berkeley.Edu>
Date: 28 Feb 88 17:36:42 GMT
References: <18403@topaz.rutgers.edu>
Sender: news@pasteur.Berkeley.Edu
Reply-To: max@eros.UUCP (Max Hauser)
Organization: UC Berkeley
Lines: 33
Summary: Exar, current vs. voltage control

In article <18403@topaz.rutgers.edu> awalker@topaz.rutgers.edu (*Hobbit*) writes:
>...  Are there any VCO chips that do square/
>triangle/whatever outputs that require a minimum of external parts?  Cheap?

Indeed many of the good parts still being used were designed in the 1970s 
and so reference data from 1978 may still be useful. With the caveat that
it has been a few years since I looked at the range of oscillator chips
available, I recall the very successful line from Exar Integrated Systems,
designed by Al Grebene et al., sort of a second generation of VCO circuits,
following the earlier very successful oscillators in phased-locked loops at
Signetics (560, 561, 562, 565, 566, 567). The Exar parts include the 205,
a wide-bandwidth square/triangle generator capable of Megahertz frequencies,
and the 2206, a submegahertz oscillator often used for audio, with
square, triangle and sinewave outputs. If someone has experience with more
recent easy-to-use oscillator chips, please post (I'm curious too).

It's worth noting that virtually without exception, circuits marketed as
"voltage-controlled" triangle-wave oscillators are forms of relaxation 
oscillators and operate internally, at the core, as *current*-controlled
oscillators, to which the designers have usually added a voltage-to-current
converter circuit. This is because it is easy to derive fundamental timing,
over a wide controllable range, in a triangle generator from the ratio of
a current and a capacitance; and yet a voltage, not a current, is the
usual external frequency-control signal that users want to apply.

Often the limits to accuracy in the VCO's voltage/frequency relation derive
from the voltage-current conversion and not from the relaxation-oscillator
core. Also, in many circuits it's possible to override the V-I converter and
impose your own external current control if you really want high control 
accuracy or wide range. Sometimes the designers provide explicitly for this,
but sometimes (as with the Signetics 565-566-567) you have to get tricky.

Max Hauser / max@eros.berkeley.edu / ...{!decvax}!ucbvax!eros!max