What determines the amplitude of a Crystal Oscillator?

[dicky96]

Hi, so the question is quite simple really.

When I stick my scope probe on the legs of a 2 pin Crystal Oscillator in circuit (usually during diagnosis of a faulty device such as motherboard or embedded microcontroller) what determines the amplitude of the oscillation I should see?

Sometimes it is quite large, greater than 2.5V, other times it can be small, say 60mV.  One leg usually has a much larger amplitude than the other.  Is this a function of the device the crystal is attached to or is there such a thing as a faulty crystal that does oscillate at the correct frequency but with too low amplitude to drive the device?

Thanks

[Whales]

I believe it entirely depends on the drive circuit.  Crystals are not useful on their own.

Sidenote: MCUs with the ability to directly-attach crystals still have drive circuitry, just internal.

[ataradov]

The "large" leg is Xout - the driving output.
2.5 V is a very large amplitude. Are you sure it is a crystal and not a CMOS generator?

MCUs often have automatic gain control. Visually pins should have undistorted sine waveform.

And crystal manufacturers specify crystal drive power, and it should be followed for optimal performance. Overdriving the crystal may shorten its life.

[RoGeorge]

The amplitude seen on the oscilloscope is irrelevant.

The parasitic capacitance of a passive oscilloscope probe is very big for a quartz oscillator.  It is so big that can stop perfectly stable oscillators, or can make defective ones to start oscillate.

[dicky96]

The amplitude seen on the oscilloscope is irrelevant.

The parasitic capacitance of a passive oscilloscope probe is very big for a quartz oscillator.  It is so big that can stop perfectly stable oscillators, or can make defective ones to start oscillate.

Then how to reliably test if the Quartz crystal is actually any good?

[RoGeorge]

First to say, Quartz crystals are very reliable when compared with other components.  They almost never brake.  The only time I've seen a defective one was in a drone that felt from a height of about 20-30 meters, and the crystal shattered at impact.

Most probably the Quartz will be the last thing to check on a motherboard.  The most common failure for a motherboard is bulged or high ESR electrolytic capacitors.

Anyway, if a quartz is bad, it usually won't oscillate at all, so if you see other digital signals changing, then the quartz is good.  There is no such problems as a quartz is oscillating, but the amplitude is not enough.

However, if you still want to check a quartz oscillator, the best way is to measure this indirectly on another digital output, not by directly probing the quartz.  If you don't have such a digital output available, or you insist to probe the quartz directly, you can use an active probe.  Alternatively, to see if it oscillate without disturbing it much, try a small wire loop on the tip of the probe, acting like an antenna, and put it near the oscillator without touching the circuit.  Another technique could be to measure through a very small capacitor in series with the tip of the probe, about 1pF or less.

In any case, the amplitude is irrelevant, and it is not something to worry about.  If you have, let's say a 20MHz quartz and you see some oscillations at the expected 20MHz frequency, no matter the amplitude, the quartz is good.

[dicky96]

OK got it thanks :-)

Actually  I don't generally find Crystals failing either.  Sometimes I have seen corrosion causing the legs to fall off.