Author Topic: Quartz Crystals  (Read 3529 times)

0 Members and 1 Guest are viewing this topic.

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Quartz Crystals
« on: January 24, 2018, 01:30:15 pm »
Hi Guys :)
I want to know more about unbuffered (two pin) crystal packages.

The first aspect is how they are driven.
What I think happens, is the micro (or whatever it is) outputs a pulse to one side of the crystal, and receives a pulse from the crystal from it’s opposite pin.
Then the micro (or whatever it is) outputs a pulse to the first side of the crystal again, and waits for an input, and the cycle continues, but the micro
(or whatever it is) is playing a part in the oscillation by providing power, and a feedback loop that I guess must be faster than the crystal.
I have arrived at this “gut feeling” because a crystal can be driven with one single inverter,
and another inverter in the same chip package such as 7404 might be used to buffer the output.

Or I’m totally wrong, and please correct me :D

The second question I have is about how the electrical connection is made from the crystal package’s two pins to the natural quartz crystal itself.
I have no idea about this at all.

Any insight would be appreciated :)



 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19527
  • Country: gb
  • 0999
Re: Quartz Crystals
« Reply #1 on: January 24, 2018, 02:19:19 pm »
It's not that simple. The MCU uses a linear amplifier for the crystal oscillator circuit. It's normally an unbuffered CMOS gate.

There are other ways of making a crystal oscillator, than just using the built-in amplifier inside a microcontroller. A single transistor will work.

https://www.google.co.uk/search?safe=active&client=firefox-b&dcr=0&ei=fZVoWoTaOanLgAbIlIuQCQ&q=crystal+oscillator+circuit&oq=crystal+oscillator+circuit&gs_l=psy-ab.3..0i67k1j0l9.2729.3643.0.3891.8.3.0.5.5.0.100.272.2j1.3.0....0...1c.1.64.psy-ab..0.8.373....0.1t7-BWuKeFU
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #2 on: January 24, 2018, 02:49:53 pm »
Is that not essentially the same thing? although perhaps inherently sinusoidal?

In fact, right in the results you posted :D
http://www.mwrf.com/sites/mwrf.com/files/uploads/2016/01/Figure_10_3.gif
« Last Edit: January 24, 2018, 02:51:45 pm by @rt »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19527
  • Country: gb
  • 0999
Re: Quartz Crystals
« Reply #3 on: January 24, 2018, 03:49:38 pm »
Is that not essentially the same thing? although perhaps inherently sinusoidal?

In fact, right in the results you posted :D
http://www.mwrf.com/sites/mwrf.com/files/uploads/2016/01/Figure_10_3.gif
Any crystal oscillator is inherently sinusoidal. The reason why it's not, is because the amplifier is non-linear, often deliberately so, in order to generate a nice square wave.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Quartz Crystals
« Reply #4 on: January 24, 2018, 06:56:58 pm »
As a filter.

A filter is a three terminal network, one pin ground, one input and one output.  The input and output are interchangeable for purposes of signal flow (a filter always works the same in both directions), but their impedances can differ; a symmetrical filter of course has identical impedances and it doesn't matter at all.

The kind of filter used is a bandpass, so that a signal applied on one side is strongly attenuated except for a peak at the desired frequency.  Crystals give a very narrow, sharp, well defined frequency, making them great for precision timing and filtering.

The equivalent circuit of a crystal is a series resonant network, with some capacitance in parallel.  A series resonant circuit has minimum impedance at resonance.  If we put the crystal in series (from input to output), we get transmission through the filter, at resonance.  That's good.  But we also need some impedance to work against, otherwise plenty of signal will be let through away from resonance.  It's not enough to have an impedance that varies: we need an impedance divider.  So we add loading capacitors, from input to ground, and output to ground.  This transforms the crystal into an equivalent parallel resonant circuit (as if input and output were tied and the crystal were parallel resonant to ground), giving a typical bandpass characteristic.

What impedance is the crystal filter?  Typically low kohms (for ~MHz range crystals).

If the oscillator has a logic level output, there will be a series resistor connected to the filter, which terminates the filter and drives it with adequate power (you don't want to overexcite the crystal -- it can shake itself to bits!).

If the oscillator has a linear or current limited output, it will connect directly, and it doesn't look like anything special is going on (you have two pins, OSC1 and OSC2, and a crystal and two caps between them), but it's still the same input-output filtering scheme.

The impedance is basically the reactance of the loading capacitors, give or take an error factor.  Thus a crystal that takes 20pF at 50MHz is a very low impedance, while one that takes 4pF at 32.768kHz is a very high impedance indeed (100s kohm).  The latter is very useful for low power applications (it consumes very little drive current, and power), but be careful that it's also sensitive to contamination (say if there's conductive flux residue on the board).

The one difference is if there's a single pin oscillator, which internally is a negative resistance circuit, which resonates with a crystal and one or no capacitors (in its parallel resonant mode, slightly off from rated frequency).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: abraxa

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Quartz Crystals
« Reply #5 on: January 24, 2018, 07:37:55 pm »
I don't see any answers to the second part of the question which is the electrical connection. In the crystals I've seen inside, there is a disc of deposited metal on each side of the quartz wafer and thin support wires are soldered or welded to these deposited plates.
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #6 on: January 24, 2018, 11:48:30 pm »
There’s a video on YouTube about the manufacture of radio crystals for the war. For those times at least, the crystal wafer, with polished sides, simply clips into an assembly that clamps a pair of electrodes either side of the wafer.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8275
Re: Quartz Crystals
« Reply #7 on: January 25, 2018, 03:21:31 am »
The crystal itself is not a conductor; it's piezoelectric, and the electrodes only serve to provide the electric field in which it operates.
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #8 on: January 25, 2018, 03:46:07 am »
There was another consideration from the video.
The crystal is first sliced up either aligned, or perpendicular to it’s alignment (polarity).
I’m not sure the correct term to use there.
The electrodes face the quartz at a particular orientation.

I might as well come clean.
I want to try making a crystal oscillator with a relatively large double terminated natural quartz,
which can be obtained from any new age hocus pocus retail shop.
The piece that I have is resonant at an audio frequency, making it about 9.5cm length.
Am i barking up the wrong tree completely? Otherwise what considerations have to be made.

ps. That appears right, no matter how I mate any malleable copper with it, it’s not remotely conductive.

I assume it’s one of the lower peaks that’s fundamental.



« Last Edit: January 25, 2018, 04:02:12 am by @rt »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Quartz Crystals
« Reply #9 on: January 25, 2018, 05:50:17 am »
Right, it's a capacitor with a kind of electro-acoustic coupling added in.

A solid crystal will need incredibly high voltages (many kV) to move perceptibly.  Consider a piezo ignitor is a crystal maybe 1cm long, struck with a spring loaded hammer, and makes 5-10kV peak.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6912
  • Country: ca
Re: Quartz Crystals
« Reply #10 on: January 25, 2018, 06:41:09 am »
What impedance is the crystal filter?  Typically low kohms (for ~MHz range crystals).

I measured single resonator impedance in 10s of Ohms at the fundamental and increases to hundreds Ohms at high overtones. They can be measured with regular 50 Ohm VNAs. The larger the crystal, the smaller its impedance at the fundamental frequency, can be less than 10 Ohm.
Facebook-free life and Rigol-free shack.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Quartz Crystals
« Reply #11 on: January 25, 2018, 07:01:38 pm »
Yes, the series resonant impedance is the ESR, which is listed on the datasheet. :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #12 on: January 26, 2018, 03:02:44 am »
I could do the high voltage including the capacitors, but then wouldn’t have the transistor for anything like a Pierce oscillator.
The result i was after wasn’t something you could see though, maybe hearing the crystal itself, or a frequency output,
which I suppose could still be achieved with some coupling to the high voltage circuit.
Slicing the crystal is going a bit further than I was prepared for though.
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #13 on: February 09, 2018, 10:24:02 am »
I’m not sure that I’m out yet.

If the crystal has an audible resonant frequency (it does) The New Agers call that a singing crystal.

Say it was suspended from a thin tether, and had a pair of piezos (or one piezo and an electrodynamic mic) attached to it.
If the piezo sweeps it’s range, shouldn’t the mic send the highest amplitude signal when the output piezo is at the crystal’s resonant frequency?
That way, you’d have a feedback to control the source of the sweeping audio signal.
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1059
Re: Quartz Crystals
« Reply #14 on: February 09, 2018, 11:56:19 am »
I’m pretty sure this is not a resonant f or harmonic of the piezo itself, and it does sound close to the crystal “clink” f.
But then I’m not sure it’s anything special, because doesn’t any mass have a resonant f?


 

Online TimFox

  • Super Contributor
  • ***
  • Posts: 7951
  • Country: us
  • Retired, now restoring antique test equipment
Re: Quartz Crystals
« Reply #15 on: February 10, 2018, 08:02:59 pm »
The simplest mechanical resonance needs a mass and a spring-like object with compliance.  A swinging pendulum has the same mathematics for small-amplitude motion.  See any freshman physics textbook about "simple harmonic motion".
 

Online CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5235
  • Country: us
Re: Quartz Crystals
« Reply #16 on: February 11, 2018, 12:25:27 am »
I have no contact with new agers so don't have any idea how a singing crystal resonates, but it almost has to be some form of bending mode instead of the compression modes used by typical of crystals.  If you can figure out the mode by tapping the crystal in various places while supporting it in others you will have the first step in n driving this crystal electrically.  That would be done by arranging electrodes to cause piezoconstriction to excite that mode.  No guarantee that it is possible, but in some cases it can be done.  Look at the tuning fork configuration used in quartz rate sensors for an example of one near audio configuration that works. 

Most crystal resonators I have seen used plated metal regions for the electrodes.
 

Offline JugglingElectrons

  • Contributor
  • Posts: 15
  • Country: us
Re: Quartz Crystals
« Reply #17 on: February 15, 2018, 04:13:49 am »
The simplest mechanical resonance needs a mass and a spring-like object with compliance.  A swinging pendulum has the same mathematics for small-amplitude motion.  See any freshman physics textbook about "simple harmonic motion".

Ahh, the beauty of multiplying decaying exponentials (damping) with oscillating sinusoids. Damped harmonic oscillators are one of my favorite topics in electronics and physics theory.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf