Hallo all,

sometimes, question arises of how reliable such an oscillator circuit will work. It's the absolute nightmare of a hardware developer, when one fine day oscillator will not start at power-up. Well, this moment, namely start-up, is the most critical condition for any oscillator. When it will fail one fine day (or earlier!), then it will fail at power-up, very very probably. The reason for this is, that oscillation condition at this very moment is hardest to fullfill, mainly because for lower supply voltage small signal gain and bandwidth of gate his drastically decreased.
Also, oscillator needs the sharp Vcc edge of power-up for being able to start.

It's important to note, that many relevant parameters are highly dependent on temperature. Having an oscillator working nicely at room temperature, does not mean at all that it will run correctly at much higher or lower temperatures!

There is a very easy methode during development phase to get an idea of how reliable oscillator will work: Put your oscillator circuit at limits of temperature range and rise power supply voltage very slowly from 0V to Vcc. Do this many times, and wait some seconds after every remove of Vcc.

This methode is also helpful for testing whether a certain quartz crystal from certain manufacturer is well matched to your mcu. By this test rise time of Vcc should not be longer than some tens of miliseconds. Otherwise mcu could hang-up. When rather high rise times of up to 100msec shall be simulated, an increased filter capacitor in power supply unit, directly behind rectifier diodes, can be used: Resistance of output winding of transformer in combination with filter capacitor makes the time constant.

Bye,
Kai