high speed op-amp oscillator - sine, square, triangle waveforms

[tonyflair]

I'm trying to design a variable frequency RC-component based oscillator (with a large frequency range) using HIGH SPEED OP-AMPS (no special IC's). I want the output to be symmetric sine, square, as well as triangular waveforms with amplitude control.

Has anyone designed such a circuit in the past or know of a ciruit that has already been created with the above parameters.

Any help would be much appreciated.

Thanks in advance

 

[VEBill]

To make a good & clean sine wave with just an op-amp seems to require a 'Wein bridge' oscillator with the small incandescent lamp used as the amplitude regulator.

For the square and triangle waveforms, look INSIDE the famous 555 circuit. They linearize the RC charging ramp by keeping it within a narrower 1/3 to 2/3 of the power supply. The rest is relatively easy.

Curious - why must you limit your IC selection ??

Re: Shouting "HIGH SPEED" What frequency range?

 

[melone]

For what purpose? Is there nothing available off the shelf that will do the trick? What type of loads are you going to be driving? How much current? ...

 

[VEBill]

Recently, there have been many queries in Eng Tips that seem to insist on using, or not using, a certain type of component. It would be better if they explain at the outset WHY they have such restrictions on the type of parts to be used.

 

[jimkirk]

Max frequency? Ratio of min to max in one range?
How do you want to vary the frequency (V, I, R, other?)
Required rise time of squarewave, linearity of triangle, distortion of sine?
Control function (linear? log?)

Check out National Semiconductor app notes, for example

http://www.national.com/an/AN/AN-263.pdf

has a circuit that does 1 to 10,000 Hz in one range.

Look at the old Intersil 8038 or the newer Maxim MAX038 for ideas.

http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1257

(if you really don't want to use one of these, substitute the various functional blocks with your high speed op amps).

Linear Technology probably has some good info, too.

A popular way to do a function like this, an integrator's input is switched between a plus and minus level by a comparator with hysteresis. This gives a triangle and square wave. The triangle then goes through a non-linear network (diodes, FET or other) to approximate a sine wave. Is this similar to what you're thinking about?

The amplitude control I'd leave to a subsequent stage. The distortion of the sine wave is critically dependent on the ampliture of the triangle wave so is not easily varied.

 

[nbucska]

please read FAQ 240-1032



<nbucska@pcperipherals DOT com> subj: eng-tips