How to implement a sinewave oscillator

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
6 replies · 2K views
JohnsonJohnson
Messages
2
Reaction score
0
Hi, I was wondering on how to implement a sinewave oscillator with a variabke frequency from 20hz to 20khz.

What types of circuits are there to do this and what sort of parameters would affect the frequency range?

Thanks
 
Engineering news on Phys.org
Hi there
welcome to PF :smile:

Google Wein oscillator circuits
there are dozens of examples, some better than others
find a link or two that has one that suits your requirements and share the link here discussed

they work OK, have built one or two in the past
if you are wanting reasonably calibrated precision, then buy a signal generator/function generator
They will give you a referenced frequency and a controllable output

google function generators on ebay
as with the osc circuits, share a link or two and recommended guidance can be givenDave
 
The Wien Bridge Oscillator is the one to use. Make sure the particular circuit has either a small lamp or a pair of back-to-back diodes as a gain control. This is necessary to ensure a sine wave. I believe Hewlett and Packard started up their business in a garage when one of them invented this circuit. Prior to that, there was no simple circuit able to do the job
 
JohnsonJohnson said:
Hi, I was wondering on how to implement a sinewave oscillator with a variabke frequency from 20hz to 20khz.
What types of circuits are there to do this and what sort of parameters would affect the frequency range?
Thanks

There are many different oscillator topologies - each with pro and cons.
However, in most cases, frequency tuning needs to change two parts in parallel (no single-element tuning).
But there are some circuits which can be tuned by a single element, but the WIEN - not WEIN - oscillator does NOT belong to this class.
So - you have two basic problems:
(1) Select a topology that allows single-element tuning,
(2) Tuning capability over a range of 3 decades. THIS IS PRACTICALLY IMPOSSIBLE !

Note that for commercially available oscillator devices we normally have to select (via switch) different frequency ranges.
 
LvW said:
(1) Select a topology that allows single-element tuning,
(2) Tuning capability over a range of 3 decades. THIS IS PRACTICALLY IMPOSSIBLE !
Well, Motorola once had an application note discussing that. It ended up with an integrator-comparator oscillator (triangle wave and square wave) and created a sine wave using the triangle wave and a couple of silicon diodes. Intersil used to have an IC that did just that (https://www.intersil.com/content/dam/Intersil/documents/icl8/icl8038.pdf).

Or you could start with a crystal oscillator, use a programmable divider and create a sine wave using a lookup table and a D/A converter.

Or - use a synthesizer/keyboard tone generator (see http://www.armory.com/~rstevew/Public/SoundSynth/TopOctave/topdividers.html)

Think outside the box - there are a lot of options.
 
  • Like
Likes   Reactions: Averagesupernova
Svein said:
Well, Motorola once had an application note discussing that. It ended up with an integrator-comparator oscillator (triangle wave and square wave) and created a sine wave using the triangle wave and a couple of silicon diodes. Intersil used to have an IC that did just that (https://www.intersil.com/content/dam/Intersil/documents/icl8/icl8038.pdf).

Or you could start with a crystal oscillator, use a programmable divider and create a sine wave using a lookup table and a D/A converter.

Or - use a synthesizer/keyboard tone generator (see http://www.armory.com/~rstevew/Public/SoundSynth/TopOctave/topdividers.html)

Think outside the box - there are a lot of options.
There truly are a lot of options. I was thinking along the lines of the integrator-comparator as well to get this kind of range out of a single oscillator. Also, no one says there absolutely needs to be a single oscillator. It is implied in the OP but I think this individual is probably interested in a signal generator for audio more than just proving that a single oscillator needs to be able to cover this kind of range. I doubt the OP even realizes that it is any more difficult than covering just one octave. Before any formal electronics training I sure wouldn't have known.
-
The following is not bad advice either. Would be willing to bet that an inexpensive function generator uses several different oscillators throughout it's entire range.
if you are wanting reasonably calibrated precision, then buy a signal generator/function generator
They will give you a referenced frequency and a controllable output