The ring oscillator comprises a number of inverting transistor amplifier stages connected in series, in which the output of the final stage is connected to the input of the first stage. You have the choice of using three, five, seven or nine stages. The only condition is that the number must be odd, not even. A feature of this circuit is that no capacitors are required. Oscillators of this kind are used widely in integrated circuits, for instance in microcontrollers.
In principle, we are dealing with an amplifier with negative feedback that reaches oscillation as a result of the high total amplification. In the circuit shown in Figure 1, five stages are employed. To avoid affecting the ring a buffer stage is used for uncoupling the oscillator signal. All resistors in the circuit have a value of 2.2 kΩ and all transistors are type BC548A. The oscillator produces a frequency upwards of 1 MHz, which is somewhat dependant on the power supply voltage (see Figure 2). An average maximum frequency of 1650 kHz is produced with an operating voltage of 3 V.
The ring oscillator can be viewed in its broadest sense as a run-time oscillator. The signal run-time of all five stages amounts to half the oscillation period, in other words exactly 300 ns at 1.65 MHz. Every individual stage then has a runtime of 60 ns. At higher operating voltages the delay introduced by each stage increases somewhat because the transistors are driven heavily into saturation.