# What happens to output of a crystal oscillator if the input it's given is an AC waveform instead of stable DC?

Some things I've read say the crystal acts as a filter and you still get its resonant frequency as the output and something else I read says feeding the crystal oscillator with a sine wave of the crystal's resonant frequency will break the crystal like a singer hitting producing a high pitch tone, breaking a crystal.

I also am curious what happens if you feed the oscillator a non sinusoidal waveform the same as it's resonant frequency. My intuition says it won't damage the oscillator because it's made up of different component sinusoidal frequencies.

• What do you mean by 'crystal oscillator'? That covers a very broad range of things. Are you talking about an oscillator module such as this? Or do you mean the crystal itself without any electronics? Commented Jan 24 at 19:33
• An oscillator doesn't have an input; a crystal has two passive pins so, how can you feed an oscillator a waveform? What do you mean? Commented Jan 24 at 19:35

If you are talking about a crystal oscillator, it is already being fed an AC waveform, via an bi-stable amplifier.

Sometimes they put this circuit into a package (or another circuit that has a different resonator other than a crystal), and those take DC power for operation of the amplifiers. If those circuits were feed different voltages then the chip would not function correctly with indeterminate results. Since EE's like determinate results no reasonable person would do this, but you could always try it and find out.

The crystal works at some resonant frequency when it is loaded properly in a proper environment. It won't break, damage, or degrade more than specified by the aging specs, unless you provide too much excitation. Crystal dataheets state the power it can handle.

A quartz crystal vibrates and flexes when an electric signal is applied in some directions. Resonant response (and electrical impedance changes) occurs at any frequency where the 'echoes' in the crystal are reinforced by the driving signal.

In most (modern) oscillators, low voltages and minuscule drive capacitances are employed; if the drive ever got the crystal close to breaking, the back-EMF from the crystal's piezoelectric generation would be clamped to the power supply rails (assuming the common CMOS driver circuitry), thus limited.

That means an ordinarily linear item (the amplifier that drives the crystal) would have its inputs (the crystal response) driven to the power rails, where diodes in the amplifier would conduct and damp the signal.

Maybe in the old vacuum-tube days, with hundreds of volts available, undamped amplifiers could damage a lump of quartz. These days, 3.3V and fleapower low-output-current drives are unlikely to have that capability. A well-designed oscillator wouldn't damage its rock, of course, but the design elements that prevent it are invisible, intrinsic (and parasitic) elements of those oscillators.