A clock is an essential item in microcontroller operations. Usually, the clock is oscillated internally by connecting a crystal or ceramic resonator and a capacitor to the microcontroller's on-chip oscillator, or a clock that is oscillated externally is used.
The crystal or ceramic oscillation circuit generates the clock by inverting (by 180°) the phase of the output of an inverting amplifier configured by adding the resonator, capacitor, and a feedback resistor to an inverter, then returning that output to the input.
[Oscillation stabilization time]
When the oscillator starts oscillating a clock, clock oscillation is not immediately stable. The amplitude of the clock signal gradually increases and the oscillation stabilizes at the frequency determined by the crystal or ceramic resonator. The time during which the frequency is increasing in this way is known as the oscillation stabilization time. During the oscillation stabilization time, the amplitude is small and the clock includes many high-frequency elements. This clock therefore does not satisfy the required specifications for operating a microcontroller, and must not be used until the oscillation has stabilized.
Coffee Break |
The oscillation stabilization time differs depending on the type and oscillation frequency of the resonator used as well as the circuit matching status. In the case of several MHz oscillation, if the matching is good, oscillation will stabilize in a few milliseconds with a crystal resonator or a few hundred microseconds with a ceramic resonator. When using a low-frequency resonator, such as a 32.768 kHz resonator used in watch applications, it can take several seconds for oscillation to stabilize.
[Securing the oscillation stabilization time]
There are two methods used to secure the oscillation stabilization time, and the method is alternated depending on the microcontroller. With one method, the oscillation stabilization time is automatically secured by the microcontroller, and with the other method, it must be deliberately secured by the user. Most 8-bit microcontrollers, such as the 78K0 and 78K0S, secure the oscillation stabilization time automatically. With V850 microcontrollers on the other hand, this time must be secured using the width of the reset signal when power is applied or when STOP mode is released by the reset.
The method whereby the microcontroller secures the oscillation stabilization time automatically at reset is easy to use, but does not allow the user more freedom. If there are problems with the system startup time at power application or following reset release, it may be necessary to make the oscillation stabilization time as short as possible. Furthermore, if the clock is supplied by an external oscillation, the oscillation stabilization time does not need to be secured. In these cases, it is better to use the method whereby the oscillation stabilization time is secured by the reset signal.
Coffee Break |
Microcontrollers such as the 78K0/Kx2 incorporate an oscillator with a short oscillation stabilization time so that the microcontrollers can be started up quickly. These microcontrollers can start operating by the on-chip oscillator as soon as the oscillator starts up.
The 78K0S/Kx1+ has a function to select the clock. If either external clock supply or use of the on-chip oscillator is selected, it is not necessary to secure an oscillation stabilization time.
In the case of microcontrollers with on-chip PLL functions, such as the V850, both an oscillation stabilization time and the time until the PLL locks must be secured.