Smoked crystals (Oscillator)

Once upon a time, during a lecture about electronics, more specifically on crystal quartz oscillators, a professor said that you can break a crystal oscillator by dropping it, because the quartz inside can be broken due to the fall, even if it seems solid like a common 1/4W resistor. Let me start with a briefing on this crystal “rock piece”.

What is a crystal oscillator? You can see very well the (main) idea on Wiki. The quartz is piezoelectric material, meaning that if you apply a mechanical force on it, it will produce an electric potential. Based on its shape, the crystal is a resonator in a certain frequency (both mechanical and electrical, since it is a piezo), like a bell. But, differently from a bell, an applied voltage on that quartz, will generate a mechanical distorsion on it. On its turn, if the signal generated by the mechanical movement is amplified, sooner or later, the amplifier will detect the resonator frequency, because the mechanical distorsion will bring also a minimal vibration at that frequency: when you touch a bell, that bell will sound, even if it is a weak sound. This is transduced in voltage and only this signal is introduced in the feedback, bouding the system to oscillate at that frequency, sooner or later. The time to amplify this signal strongly enough, is the time of start-up of the crystal. In other words the crystal behaves like a series LC resonator:


in which is almost a short circuit at its resonator frequency, allowing the signal to pass unattenuated, while rapidly reducing signals outside its resonant one.

If the amplifier is inverting one, with a very high gain and very low cost, an ugly, unaccurated digital NOT port can be used, since the accuracy is instrinsic to the quartz. And is the most adopted circuit for digital electronics (like microcontrollers):


When the system is powered, the inverting port may be in metastability thanks to Rf. But this is not a stable condition and the tiny amount of resonator frequency present in the feedback will lead the amplifier to start the oscillation bounded by the crystal frequency. Since the final signal is a square wave, Rs is used to dissipate the harmonics that would be instead dissipated inside the crystal, avoiding some stress on it.

Where were we?

All this brief introduction had the purpose to let the reader know what I’m talking about. But the fact about breaking up a crystal (these words sounds very close to Breaking Bad stuff) said in a lecture, is similar to what happened in reality and I never thought to say this. I write this to report also how electronics in domestic environment is designed in a “I don’t give a sh*t” way.

We bought recently a vacuum cleaner “centralized”, meaning that you have only one fixed aspiring motor placed outside the home, and inside only one long pipe used to clean every angle of the various rooms. This pipe is connected to the vacuum cleaner by a sort of outlet fixed in the wall: fixing the pipe on the wall, will close a circuit that will turn the motor on. This system can be controlled by a remote transmitter placed in one part of the pipe (the one that you hold), that upon user interaction will close a switch on the other side of the pipe attached to the wall, using a receiver which closes the morot circuit. Here the crap: handling a vacuum cleaner pipe, will result inevitably to banging it on the floor, at least a couple of times.

That being said, one day I heard something dancing around inside the handle of the pipe, I thought it was a piece of plastic that was broken… classic! And then no, opening the handle I found this:


Everything normal? Something is missing here. We can see the exposed crystal of the microcontroller! We can spot the metallic square shaped armor placed on the crystal, almost transparent (it is quartz, probably):

The crystal X1 opened (broken cover)

The armor trasfers the electric potential to the crystal, exploiting the piezoelectricity mentioned before. This was still working, so I searched the cover lost here around trying to fix it. And here the tiny bastard:

Top view of the crystal cover
Bottom view of the same cover

So, just to avoid that something else will damage the quartz, with some hot glue I covered up the crystal with its cover. A first step was searching few marks, trying to understand the original position:

Ready to put on top


Finally, the cover can be put on the top. And it still works. Perfect match, perfect glue!



Things are designed to cost very low, and this is the price often to pay at a second turn: you can repair it or pay to repair it, or you will change it.