If there is no requirement on phase noise, oscillation frequency etc, 3 inverters connect as ring will do.
If oscillation freq is high, you still able to get it by inverters ring. If the oscillation freq is low, you may need to think about 555 timer topology, controlling charge and discharge currents.
The attached circuit is a pretty simple one (so is a little bit unflexible) but it generates a triangle and square wave with variable frequency and amplitude of ±Vcc). Let me know if you need something else.
Sure... The lower opamp is an schmitt trigger with the attached transfer function. The upper opamp is a simple integrator. Now... suppose that when you turn on your circuit your Schmitt trigger ouput (hence integrator input) is Vcc and that the capacitor of the integrator is completly discharged, so the capacitor's voltage will be given by: Vcap=(1/C)*∫ ic(t)*dt, where ic(t) is Vcc/R5, so Vcap=(Vcc*t)/(R5*C) (ramp function). Since U2A is an inverter, it's ouput will be 0-Vcap, so your schimitt trigger input will be a ramp with negative slope given by -Vcc/(R5*C). As soon as this ramp reaches -Vcc (rembember the capacitor started with 0 V) the schimitt trigger will be force to change it's output from Vcc to -Vcc, now making the capacitors current negative which makes it discharge, which implies that the integrator's output will now start rising with a positive slope given by Vcc/(R5*C)
I hope I was clear enough... sorry but my english is not that good
try the simple circuit i have given...using 74hc/hct14 ...u can also use 40106..4093...any schmitt trigger inverter ....the best part is that it is very cheap...
or search the 2 transistor oscillator on google....
Do you ramp up your supply voltage or directly place a DC voltage?
For oscillation to begin, you need an excitation. It could be supply ramp or power on reset signal.
I simulated that circuit using Pspice and it worked fine. If all your circuit components (especially the opamps) in your circuit are idealized models, it could happen that the schimitt trigger input doesn't have the offset voltage necesary to give an Vcc or -Vcc output, instead it would be 0 V, so the integrator would never start to work, but in real life that would never happen. Your DC voltage doesn't need to be a ramp, the key feature is to have a little DC offset on the opamps output.