It was suggested by someone on this forum that I built a phase shift oscillator since I was having trouble with my colpitt. I would like to make the one below, but from a single 9v battery supply. My problem (so far) is I'm not sure how to make that 2.5v at the Vin. I tried a voltage divider, but wasn't sure that was right because if it's a virtual ground it really doesn't go to ground?
Mistake: that op-amp in the picture is the LM358 (I've been making too many searches on the atmega385 )

 

Here is a primitive fixed frequency phase shift oscillator. The reason that it is primitive is because it has rather high output distortion, and it is very sensitive to the value of R7: too high and it won't oscillate; too low and the output is severely distorted.

Maybe if you told us what you are need the oscillator for, its output, stability, distortion, frequency, tuneability, etc, we could suggest a better circuit.

 

Save yourself the headache and use two 9 volt batteries.

My textbook says that phase shift oscillator does not work well at high frequencies, but it does not say how high. Does anyone know?

 

KLillie, the op-amp doesn't know if ground is "real." The voltage divider is fine. Also, can someone else weigh in? I feel like I've built phase shift oscillators that used additional buffers between the reactive stages, and they seemed like they behaved better...

 

KLillie, the op-amp doesn't know if ground is "real." The voltage divider is fine. Also, can someone else weigh in? I feel like I've built phase shift oscillators that used additional buffers between the reactive stages, and they seemed like they behaved better...

It is called a Bubba Oscillator. It is documented in a

TI App Note

 

Creating a Virtual Power Supply Ground

 

Thanks guys! Great answers. @MikeML - I'm currently on an oscillator kick! And I think I'm getting kicked the most! They are hard. I finally made one with a resonator a couple weeks ago, but have yet to have much luck with the other forms (capacitor/inductor pairings).

 

Bill, I had seen this page earlier "Creating a Virtual Power Supply Ground" and was not sure it had my answer. In fig.2 should C3 be pointing that way? Is Vss to be a negative voltage? I don't get what's going on there. Thanks

 

Actually, while the overall theory of that article is great, I have questions about those caps, too. That orientation seems to contradict every other source I've ever seen, including lots of working guitar pedals I've serviced...and built (lots of virtual ground in stompboxes!)

 

Below is the simulation of a 3-stage phase-shift oscillator that uses an opamp buffer for each RC phase-shift stage (the phase shift is performed by the parallel RC's in the feedback network rather than the usual RC to ground).
Back-to-back parallel diodes are used to softly limit the amplitude and avoid hard amp clipping.
The output is taken from the last amp in the chain after the limiting so the signal is filtered by two of the RC stages to minimize the distortion from the limiting. This makes it less sensitive to loop gain as compared to the the single op amp phase-shift oscillator. The output is also low impedance.
The pseudo-ground voltage was selected to be at about 1/2 the op amp maximum voltage swing rather than 1/2 the supply voltage, for maximum symmetrical swing from the op amp output.

 

(also, fixing MikeML's link, cuz I got a 404 back and really wanted to read the article!)

TI Ap Note

 

OK, you got me. I'll update the drawings.

 

That TI app note is great, and has some variations I've never seen before, but I think the assessments of maximum frequency are a liiiiiiittle pessimistic for modern devices.

 

(also, fixing MikeML's link, cuz I got a 404 back and really wanted to read the article!)

TI Ap Note

The last sentence in that article:


May your oscillators always oscillate, and your amplifiers always amplify

 

Found this page-
http://tangentsoft.net/elec/vgrounds.html

 

It was suggested by someone on this forum that I built a phase shift oscillator since I was having trouble with my colpitt. I would like to make the one below, but from a single 9v battery supply. My problem (so far) is I'm not sure how to make that 2.5v at the Vin. I tried a voltage divider, but wasn't sure that was right because if it's a virtual ground it really doesn't go to ground?
Mistake: that op-amp in the picture is the LM358 (I've been making too many searches on the atmega385 )

There are better op-amps than the 741, and a lot of them are dual op-amps, a common way of creating a virtual earth is to use a spare op-amp as a unity gain voltage follower with the + input tied to 2 equal divider resistors splitting the supply voltage in half.

This gives you a low impedance 1/2 Vcc reference point without the high current draw of low value divider resistors.

Don't put a decoupling capacitor directly on the reference op-amp output - it'll probably oscillate! Put about 33r or so from output to bypass cap.

Its generally best to put the bypass cap on the + input - not too large or the virtual earth will take a while to rise up to 1/2 Vcc.

 

Got the single supply, single op-amp phase shift oscillator to work using the LM358 to work! More to come...it's late.

 

Posted my success to projects! Check it out and help me.