Jump to content

Talk:Crystal oscillator

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Fundamental workings

[edit]

tl;dr I think the article should specify that a quartz oscillator works by means of the cyrstal deforming slightly, which then opens the circuit as the physical connection is lost, and so then the crystl deforms back into original shape, whereupon it touches the wire again and is again subjected to a voltage which again causes it to deform, re-starting the cycle. If indeed that is how a crystal oscillator works.

I've never fully understood an explanation of how a crystal/quartz oscillator works. I'd always heard the explanation that "a quartz crustal is piezoelectric, which means it will slightly deform/get stresses when a voltage is applied to it." OK.... so if you glued a wire to a quartz crystl, and aplied a steady DC voltage, it would deform, but it would just stay deformed, no? I have always GUESSED that what they meant was that in deformation the contact with the voltage source is disconnected, which removes the voltage and then the deformation goes back to original shape, creating a cycle of voltage - no-voltage accompanied by stress - no-stress.

I just figured to look it up on reddit, ELI5, and it seems my hunch about how it works was correct. One poster said it essentially as I said it.

So if that's correct, should we not somehow convey that? I'm sick and tired of half-explanations and wikipedia is supposed to be the one source that's relatively good at stuff like this. — Preceding unsigned comment added by EdwinAmi (talkcontribs) 22:27, 2 February 2017 (UTC)[reply]

OK, the same night that I made this comment I finally managed to research enough to understand what happens. The crystal's oscillations is maintained SPECIFICALLY by wiring it through an amplifier/oscillator circuit that outputs back to the crystal itself. That is, without this special set up, nothing would happen. A DC current would create a momentary oscillation as the crystal deformed, and that oscialltion would die out as the crystal maintains its new shape. An AC power source alone also wouldn't do anything special. It specifically has to be this amplifier circuit set-up — Preceding unsigned comment added by EdwinAmi (talkcontribs) 21:49, 19 February 2017 (UTC)[reply]


Split article?

[edit]

Shouldn,t this topic be Crystal instead of Crystal oscillator? Thanks.

PhysicistQuery 20:50, 21 July 2007 (UTC)No, this is about the impedance presented by the oscillator circuit to the crystal. If that is now clear, please remove this note. Otherwise, please ask detailed questions.[reply]

If someone would start a section on crystal resonators I would be happy to contribute.

I agree. There should be separate articles on "Crystal oscillator" and "Piezoelectric resonator". "Crystal Resonator" should be a redirect to "Piezoelectric resonator". An oscillator is a circuit, while a resonator is a component. The oscillator article can talk about the various circuits, their development and so on. The resonator article can talk about various electrical models, manufacturing, history, various crystal cuts and alternative materials. John Dalton 02:04, 7 August 2007 (UTC)[reply]

I was going to suggest the same thing. This article could cover crystal resonators and treat oscillators and filters as two separate cases. -Roger (talk) 23:10, 26 March 2008 (UTC)[reply]
There is a fundamental misconception involved in the presentation of this article, which is clearly exemplified by the questions asked on this talk page, and the points made immediately above. That is, that the atricle is entitled "Crystal Oscillator", and that term is correctly defined as meaning "an electronic circuit" in the first sentence, yet the remainder of the article concentrates on only one component of the circuit - namely the crystal itself, and no examples of oscillator circuits are given. You must turn to articles such as this if you want to find out about actual oscillator circuits: http://ww1.microchip.com/downloads/en/AppNotes/00826a.pdf As a professional electrical engineer I find this wikki article to be highly deficient, and I can understand the confusion expressed in the questions asked in the various contributions to the talk page. The article gives very little help, and a lot of misleading information to a newcomer to this topic, and in its present form is not satisfactory as an encyclopedia entry. Conflating "crystal oscillator" with "quartz crystal" has probably been the cause of the problem. A. J. Smith — Preceding unsigned comment added by 77.96.59.93 (talk) 09:29, 17 January 2014 (UTC)[reply]
I agree, also see "Rename this page Piezoelectric resonator?" below. --ChetvornoTALK 04:08, 18 January 2014 (UTC)[reply]

Hi Chetvorno. I have not had the patience to read fully the discussion you refer to, but it seems as though the consensus is against us. This does not help newcomers who visit this site wanting to know what crystal oscillators are. You are correct to point out that the crystal is better described as a resonator, rather than as an "oscillator", particularly as it seems that it has been agreed that the latter is a "circuit". I propose that you drop this argument as intractible, and write an entirely new page entitled "Crystal-Controlled Oscillators", which could also refer to circuits which use ceramic resonators. I will cooperate with you on technical content if you like - I have no idea how to write wikki articles however. A. J. Smith — Preceding unsigned comment added by 77.96.59.93 (talk) 09:55, 18 January 2014 (UTC)[reply]

I don't think starting a rival article is the best way forward (see WP:POVFORK). A better approach is to split the material we already have while it is on this page into two sections. We could even use a level one heading like = Crystal resonator = or whatever title is desired and write as two entirely separate articles. That way everyone can see what is happening and easily take part and when the split is complete moving the material to a new title becomes a simple cut and paste. SpinningSpark 14:16, 18 January 2014 (UTC)[reply]
In the case "split the article" we all are not free in our decision of the correct heading. All components and devices are industrial products. And industrial products follow strict rules, the national (EN, DIN, BS...) or international (IEC) standards. Here are the terms, the international standards are using:
  • IEC 60122, IEC 60444, IEC 61080, and IEC 61178: Quartz crystal units (resonators)
  • IEC 60679: Quartz crystal controlled oscillators
The standardization split very clearly between resonators and oscillators. May be someone find the time to spend some minutes to have a look at the German Wiki page "Schwigquarz". There the splitting into two articles is already done.--Elcap (talk) 14:07, 19 January 2014 (UTC)[reply]
Wikipedia is not bound to follow the terminology of the IEC, or anyone else. We have our own criteria for determining article titles. The most relevant one here is WP:COMMON. We use the name the subject is most commonly known by, not the name the IEC would like it to be known by. SpinningSpark 17:39, 19 January 2014 (UTC)[reply]
I agree with what Spinningspark said; the article should not be POVFORKed, and IEC industrial classifications are not relevant. I still think the best solution is to spin off the resonator content into a new Piezoelectric resonator or Quartz crystal (electronic) article. But if the consensus is against this, there are ways to improve the article organization to prevent misunderstanding. The article could be divided into two top-level sections: "Crystal oscillator circuits" with a description of how the circuit works, and "Piezoelectric resonators". There actually are a lot of important things to say about the circuits themselves: merits of the different oscillator circuits, drive level, avoiding "loading" the crystal, series and parallel crystal resonance, "tuning" crystals with capacitance, etc. The existing article is pushing size maximums, but maybe some of the prose bloat (for example in the "Stability and aging" section) could be edited down to make room. --ChetvornoTALK 12:40, 5 February 2014 (UTC)[reply]
I don't think that consensus is against a split. It is against just abandoning this article as hopeless and starting a new one (or at least that's my position). Why don't you follow my suggestion above and start a division of the material in situ. A physical split can follow later when we have knocked it into shape on this page. If nothing else, it is a good excuse to invoke WP:IAR and have a level one sub-heading on the page. SpinningSpark 13:14, 5 February 2014 (UTC)[reply]
Sorry I missed your above suggestion and repeated it. I'll do that. That's an interesting idea to use the level one heading markup. --ChetvornoTALK 18:23, 5 February 2014 (UTC)[reply]

Crystals and frequency

[edit]

The entire second paragraph sounds funny to me. If nobody objects, I would like to rewrite the entire thing. There are no frequencies going into our coming out of the crystal. Depending on the implementation of a oscillator circuit (if it uses series or parallel resonance), one usually builds a feedback system with the oscillator in the feedback loop. At the frequency of interest, the impedance of the oscillator becomes very small (series resonance) or very large (parallel resonance). This is in turn used to create a positive feedback, ONLY at those frequencies. The positive feedback amplification will thus build up and sustain an oscillation. Nonlinear effects or amplitude control are usually employed to sustain the oscillation at a certain amplitude and not let it grow too large. If there are no objections during the next week or so, I will rewrite this section. Agunther (talk) 21:43, 7 March 2008 (UTC)[reply]

Well, that is true, but you can also consider the crystal as a very narrow band pass filter, such that the gain of the system is greater than one over a narrow frequency range. An initial transient starts the system, the frequencies with gain less than one die away, and the output is at the resonant frequency. This might also make it easier to discuss the noise in crystal oscillator circuits. Gah4 (talk) 06:33, 1 October 2013 (UTC)[reply]

Frequency list

[edit]

I've taken 10.7 MHZ off the list. The last thing you want in a Superheterodyne receiver is an oscillator right on the IF frequency - this would essentialy block the desired IF signal.I don't think it's a common frequency at all. It certainly isn't in the Digi-Key on-line catalog. You will find ceramic filters at 10.7 MHZ but that's not a crystal oscillator. The whole list of frequencies is pretty well useless; you can get a crystal ground to any frequency you like and I can't see what value the list has. The general reader will be baffled and wonder why only certain frequencies are on the list, and the technical reader already knows taht some frequencies are stock and some are custom order. --Wtshymanski (talk) 14:14, 19 March 2009 (UTC)[reply]

Well, there are some frequencies with pretty specific origins. The 3579545Hz crystal used for NTSC color TVs, for example. Now, being so plentiful and cheap, they were then used for other things, such as Touch-Tone (R) generators. The original IBM PC had a crystal at 4x3579545Hz to make generating an NTSC signal easier, and also used it, divided by three, for the microprocessor clock. (4.77MHz is close to the 5MHz of the original 8088, and using one crystal for two uses saved a little on costs.) There isn't much point, though, in listing the nice round numbers. 20MHz can be divided by 2 or 4 or 5 to generate 10MHz, 5MHz, or 4MHz when needed. Gah4 (talk) 20:28, 10 January 2013 (UTC)[reply]

Should 28.322 MHz be on the list?

[edit]

28.322 MHz is also common frequency used on VGA cards for video, to get same horizontal and vertical timings for 360 and 720 pixel wide modes than 25.175 MHz generates for 320 and 640 pixel wide modes (see VGA article). Does anyone else know any other uses for 28.322 MHz? —Preceding unsigned comment added by 91.153.27.229 (talk) 13:16, 19 August 2009 (UTC)[reply]

28.322 MHz is also used in Software Defined Radio(SDR) receivers as the part of front end circuit which generates I-Q signals for input towards PC where the decoding algorithms run. This frequency seems ideal for the reception of 40m amateur band (around 7Khz). — Preceding unsigned comment added by 14.140.178.34 (talk) 10:45, 23 July 2013 (UTC)[reply]

What the web thinks of this article

[edit]

The site Explain that Stuff! has an article on quartz clocks and watches that has a link to this article at the bottom. It describes this article as: "This is one of those slightly baffling Wikipedia articles likely to make sense only to people who know enough about the subject to write the article in the first place." Just thought we should take note of this feedback (although the article may have been substantially improved since it was written). I think the description of how a quartz oscillator works could use a rewrite to make it more accessible to nontechnical readers. --ChetvornoTALK 22:15, 28 April 2010 (UTC)[reply]

OK, what is it that actually makes them work? There are a few things that all come together in quartz. First it is piezoelectric. Second, it has a very high Q. One or the other isn't enough. The piezoelectric effect requires a certain asymmetry in the crystal structure. That allows one to couple the electric field with the mechanical motion of the atoms. High Q means low friction, but there isn't an easy way to explain the friction internal to a crystal. Otherwise, it is a resonant system just like a pendulum clock, with a system to keep it oscillating. Gah4 (talk) 07:07, 1 October 2013 (UTC)[reply]

Exactly. The article has been improved quite a bit since that 2010 post, but I think the explanation could still use a rewrite to emphasize that those two points are the requirements for a piezoelectric resonator. Another point that should be mentioned is low CTE. --ChetvornoTALK 08:00, 1 October 2013 (UTC)[reply]

Well, it is more than low CTE, it is that with the appropriate cut that the effects of expansion mostly cancel. I would have to look it up to give a better explanation, but now that you mention it, should it be part of the article? Gah4 (talk) 22:06, 11 March 2014 (UTC)[reply]

32 kHz or 32,768 Hz?

[edit]

In "Electrical model" we can read:

"... a 6 pF 32 kHz crystal has a parallel resonance frequency of 32,768 Hz when a 6.0 pF capacitor is placed across the crystal"

I think that is not accurate. 1 kHz (kilo-Hertz, not kibi-Hertz) = 1 000 Hz (not 1 024 Hz.) Besides, it is recommended to use blank spaces instead of commas for separating groups of three digits. I suggest the following:

"... a 6 pF 32 kHz crystal has a parallel resonance frequency of 32 000 Hz when a 6.0 pF capacitor is placed across the crystal"

Or perhaps:

"... a 6 pF 32.767 kHz crystal has a parallel resonance frequency of 32 767 Hz when a 6.0 pF capacitor is placed across the crystal"

Ignacio González (talk) 11:15, 3 September 2010 (UTC)[reply]

The current drain from most CMOS circuits is proportional to the rate at which transistors switch. The 32768 Hz crystal is commonly, if not universally, used for digital watches. It is almost amazing that the frequency is that low, yet it is easy to divide down to the 1Hz needed for the timing circuits. There would be zero demand for a 32000 Hz crystal, so everyone knows that 32kHz means 32768Hz. Gah4 (talk) 21:46, 8 January 2013 (UTC)[reply]

Reminds me, I used to wonder in the days of IBM S/360, if they were priced in $K where K=1000 or K=1024. Gah4 (talk) 22:08, 11 March 2014 (UTC)[reply]

Temperature compensation

[edit]

We need to add discussion of temperature compensation techniques and point TCXO to this discussion. --Kvng (talk) 14:19, 10 August 2012 (UTC)[reply]

Rename this page Piezoelectric resonator?

[edit]

There is an unresolved problem with this article. In engineering usage, the term "crystal oscillator" refers to the circuit (see citations supporting lead sentence), and the resonator crystal itself is referred to as a "crystal" or "quartz crystal", which of course have more general meanings in ordinary usage. This has been brought up before (see Talk:Crystal oscillator/Archive 1#Only the quartz crystal or more, Talk:Crystal oscillator/Archive 1#What does it do exactly?, #Split article above). The content of this article covers the resonator itself; crystal oscillator circuits are discussed under Electronic oscillator. Should this page be moved to Piezoelectric crystal or Piezoelectric resonator, more technical terms for the resonator, and Crystal oscillator be redirected to Electronic oscillator? --ChetvornoTALK 16:22, 1 January 2013 (UTC)[reply]

Opppose No. "Crystal oscillator" is a distinct class of oscillators; if it's good enough for the index of "Art of Electronics", it should be good enough for a topic heading here. --Wtshymanski (talk) 01:25, 2 January 2013 (UTC)[reply]
First, the content of the article describes a component, not an electronic circuit. It is not an oscillator.
Second, I have made a little research in official documents. The British standard named this component: Quartz crystal units, see f.e. BS EN 60122-3:2011-02-28. I am against Piezoelectric resonator, because the ceramic material for Ceramic resonator is piezoelectric, too, but have a lot of different features and an own article. And additional, what's about a piezoelectric MEMS resonator? I prefer Quartz crystal resonator or Quartz crystal unit, because "quartz crystal" is in use since more than 40 years (I am 70 now) .--Elcap (talk) 08:38, 2 January 2013 (UTC)[reply]
The above comment by Elcap illustrates why I suggested the name change. Since the entire content of the article is about the crystal, not the circuit, people are getting the idea that the term "crystal oscillator" refers to the component, the crystal, not the circuit. --ChetvornoTALK 10:36, 2 January 2013 (UTC)[reply]
  • Oppose. As Elcap indicates, piezoelectric resonators are not necessarily crystals, and they are not necessarily being used in an oscillator. Lead zirconate titanate, for instance, is a favourite ceramic material for filter resonators. Changing the title would radically change the scope of the article leaving the current text confused and off-topic. The possible confusion can easily be clarified in the text. I don't really see why there is a problem with this article discussing circuits as well as the crystals themselves. It already has a lot more to say on circuits than the electronic oscillator article that Chetvorno points to, and in any case that is a general article on oscillators where it would not be appropriate to include the level of detail that it would here. SpinningSpark 18:18, 2 January 2013 (UTC)[reply]
You're right. Piezoelectric resonators are not necessarily crystals, and are not necessarily used in oscillators. So why do we have them in an article named Crystal oscillator? As far as I can see, there is no content in this article on crystal oscillators (the circuits) at all, except the acronyms under Circuit notations and abbreviations. Its all about the resonators. So changing the name to Piezoelectric resonator wouldn't make it off-topic, it would put it on-topic. And the article is already too bloated; if we want to put in anything about crystal oscillators, we'll have to split it anyway - keep Crystal oscillator for the circuits, and create Piezoelectric resonator for the resonators. --ChetvornoTALK 22:43, 2 January 2013 (UTC)[reply]
That being said, I'm not committed to a name change. I agree the confusion is minor and can be clarified in the text. I just brought it up because readers coming here are confusing "crystal oscillator" with "piezoelectric resonator". --ChetvornoTALK 22:43, 2 January 2013 (UTC)[reply]
I certainly think there is room for both articles to exist. Should the other article be created there might then be cause for some rationalisation here. SpinningSpark 23:31, 2 January 2013 (UTC)[reply]

Well, all of those are crystalline, some are used in single crystal form, others polycrystalline. Polycrystalline materials are made up of many tiny crystals stuck together in some way. usually in the form of a ceramic. Seems to me, then, that crystal is still the right word. Gah4 (talk) 08:04, 3 January 2013 (UTC)[reply]

Yeah, that's true, and "crystal" is the common term used for piezoelectric resonators in general. --ChetvornoTALK 11:13, 3 January 2013 (UTC)[reply]
Polycrystalline objects are not crystals in the generally accepted meaning of the word. That is really stretching the terminology too far. Some people might well call all piezo resonators "crystals" but this is only in the generic sense that all vacuum-cleaners are called "Hoovers" - we all know they are not. SpinningSpark 19:18, 3 January 2013 (UTC)[reply]
Hi Wikis, please let me explain, why the German Wiki Elcap is involved in this discussion. Since weeks some bots are deleting consequently the link between German “Schwingquarz” and English “Quartz oscillator”. Because the contents of both articles are nearly identical describing an electro-mechanical component I every 2 days I recover this links on both languages. But the bots seeing “oscillator” and are very busy in deleting once more the links not noting that the English text describe a component, not an oscillator.
I found out that the standardized English name for this component is “Quartz crystal” since decades. (BS EN 60122-1, Quartz crystal units of assessed quality. Generic specification, BS EN 60122-3, Quartz crystal units of assessed quality. Standard outlines and lead connections) It would be very helpful for all Wiki users searching for correct translation if now the English text could be moved to “Quartz crystal resonator” or “Quartz crystal unit”.
Quartz oscillators are listed in the article under paragraph “Circuit notations and abbreviations”. These types and their description would be titled correct with the term “Quartz oscillators” or could be placed in an extra paragraph under Electronic oscillator--Elcap (talk) 10:31, 3 January 2013 (UTC)[reply]
Sounds like this might be another reason to move the content. Sorry, Elcap, I thought you misunderstood the term "crystal oscillator". Thanks for clarifying that. --ChetvornoTALK 11:13, 3 January 2013 (UTC)[reply]
Since the characteristics of crystal oscillators are critically dependent on the (electrical) characteristics of quartz crystals, and since these characteristics are otherwise of little significance, it makes sense to describe the electrical properties of crystals, the different cuts, etc. in this article under the title "Crystal oscillator". --Wtshymanski (talk) 19:08, 3 January 2013 (UTC)[reply]
(ec) I don't think problems with interwiki links on the German Wikipedia is an especially good reason for a rename here. If bots are causing a problem over this the issue should be taken up with the bot owners. SpinningSpark 19:18, 3 January 2013 (UTC)[reply]
There's an entire separate application for piezoelectric resonators besides in oscillators, Wtshymanski; as crystal filters 1, 2, 3. They're used as IF bandpass filters in radio receivers and other high Q circuits. It is not true that they are "of little significance" outside oscillators. That's another reason for moving the "piezoelectric resonator" content to a separate article. --ChetvornoTALK 20:54, 3 January 2013 (UTC)[reply]


Hi SpinningSpark, you wrote above I don't really see why there is a problem with this article discussing circuits as well as the crystals themselves.. For me it is a problem, because I see very clearly the difference between a component and a circuit. But otherwise, if the circuit is encapsulated it seems to be a component, too. May be the article was written and titled by a large company want to sell oscillators and not crystals, see: Hewlett Packard, Fundamentals of Quartz Oscillators Application Note 200-2 [[1]]. Nevertheless, to split up the excisting text in a part called Quartz crystal units/resonators and a second part Quartz oscillators makes a lot of work. May be I try it in some month on a user page. --Elcap (talk) 11:06, 5 January 2013 (UTC)[reply]

  • Oppose the proposed narrow renames. The proposed names are too broad; the article is about the application of quartz crystals to oscillator circuits. The existing article addresses issues that involve both the component and the oscillator circuit. The common name crystal oscillator refers to quartz crystal oscillators. More generally, the article is long, so separating some ideas out may be appropriate. Crystals are used as resonators in electronic filters. For the bot issue, a workaround might be making en:Quartz crystal (electronic component) a soft redirect here. Glrx (talk) 16:23, 5 January 2013 (UTC)[reply]

The difference between the crystal component and crystal oscillator is pretty small. In SSI CMOS many years ago, we put a crystal across one (or three) inverters to make an oscillator. With LSI, those inverters went inside the IC, at least back to the 6502. Another use for oscillator crystals is the extraction of the reference for decoding analog color television signals. Given a few cycles of the reference, the circuit has to generate a few thousand cycles. That can either be described as a high-Q filter or a high-Q driven oscillator. If I understand it right, polycrystalline materials are used either to reduce cost (though quartz crystals are pretty cheap) or reduce Q. For IF filters, the Q affects the bandpass, so you don't want it to be too high. Otherwise, the difference between high-Q filter and high-Q oscillator is pretty small. Even so, much could be done to improve the Ceramic resonator page. Gah4 (talk) 14:43, 8 January 2013 (UTC)[reply]

One needs to distinguish between the Q of a resonator element (which is always desired to be as high as possible) and the equivalent Q of a filter working into its resistive load (simply inverse fractional bandwidth). A filter composed of ideal resonators with infinite Q will still have a finite bandwidth. The real reason ceramics are preferred to quartz in filters is that, with the right choice of material, a higher mechanical coupling efficiency is achieved. SpinningSpark 19:00, 8 January 2013 (UTC)[reply]
OK, that explains quartz vs. PZT, but not single vs polycrystalline. Also, it is still the crystal properties in the ceramic that make it piezoelectric. I believe that amorphous (fused) quartz is not piezoelectric. I didn't see it described anywhere, but I assume that piezoelectric ceramics are fired with an electric field to align the microcrystals. So, I still don't see any reason to split the article based on single vs. polycrystalline materials. And again, much could be done to improve the Ceramic resonator page if one was interested. Gah4 (talk) 21:55, 8 January 2013 (UTC)[reply]
A polycristalline quartz crystal could not resonate, not oscillate. It must be crystalline.
In ceramic resonators it is the ferro-electric material gives the properties, the cristalline particles are sintered together and behaves like crystalline.
By the way, to improve Ceramic resonator, please have a look at the German version [[2]]. May be it gives some ideas and f.e. some additional pictures you can use. Besides, additional graphs and pictures, f.e. the picture of the quartz crystal, can be used for the English article Quartz oscillor you can find in the German articel Schwingquarz [[3]] --Elcap (talk) 08:31, 9 January 2013 (UTC)[reply]

If you go to a toy store and buy a bouncing ball, then take it home, put it on the floor, and find out that it doesn't bounce. You need to supply energy, drop it or throw it, to make it bounce. Yes crystals don't oscillate without something supplying some gain. It doesn't take much, a CMOS inverter will do it. Now, an inductor or capacitor is not an oscillator, but in combination and with a little gain, they can form an oscillator. The name seems fine to me. Gah4 (talk) 02:53, 21 January 2013 (UTC)[reply]

This discussion seems to me to be a side issue, but here goes. Elcap, polycrystalline materials can resonate, they may have a lower Q than crystals, but as Gah4 said that is acceptable in many applications. The problem is that without processing they are not piezoelectric, so they can't convert mechanical vibrations into electrical oscillations. As Gah4 said, polycrystalline ceramic resonators are subjected to an electric field during manufacturing which aligns the crystal grains. They are pretty clearly not "crystals". The issue is whether there should be an article called Piezoelectric resonator in which they are described along with the other kind of piezoelectric resonator, the quartz crystal. It seems to me there should, because they both operate in exactly the same way (as a piezoelectric resonator) 1 and are used in similar applications. --ChetvornoTALK 05:54, 21 January 2013 (UTC)[reply]

In solid-state physics, materials are either crystalline (single crystal or polycrystalling) or amorphous (non-crystalline). In X-ray crystallography, polycrystalline materials will have sharp peaks, where amorphous materials will have much less structure. (There is still some short-range order visible in the spectrum.) The tiny crystals in polycrystalline materials are still very large on an atomic scale. They are pretty different if you are interested in surface effects, as is much of semiconductor physics. Gah4 (talk) 02:34, 24 January 2013 (UTC)[reply]

Similarly, piezoelectric resonators (both crystal and ceramic) 1 have applications far beyond oscillators. The most rapidly expanding part of the field is in bandpass filters, driven by the scarcity of bandwidth and the consequent need for closer frequency channel spacing. Gah4, the difference between a crystal component and a crystal oscillator is not small; crystals are used in different ways in filters than in oscillators. In an oscillator you are only concerned with one frequency, the crystal's oscillation frequency; in a filter you are concerned with the crystal's entire bandpass function. Often several crystals are used together to make a crystal lattice filter to synthesize a sharp-skirted bandpass function.1 In fact, crystals used for filters are not even built the same way as crystals used in oscillators 2. Modern monolithic lattice filters have several coupled resonators fabricated on a single quartz wafer 3; there are many types of devices. 4. Crystal filters are widely used in many devices: radio receivers, spectrum analyzers, modems, and telephone FDM trunkline equipment. 5 In summary I think it is misleading to try to shoehorn the entire expanding field of piezoelectric resonators into Crystal oscillator. They're not your daddy's crystal oscillators any more. --ChetvornoTALK 05:54, 21 January 2013 (UTC)[reply]

Yes crystal filters are different from crystal oscillators, including differences in the way they are constructed. This page is about oscillators, and the crystals that are used for them. As I previously noted, there is a fine page on Ceramic resonators, though it could be expanded some. Gah4 (talk) 02:34, 24 January 2013 (UTC)[reply]

But "crystals" are not the only resonators used in crystal oscillators; ceramic resonators are too. And the voluminous sections in this article on quartz cuts, temperature coefficient, equivalent circuit, overtones, etc. also apply to filter crystals. Why should it be here rather than in Crystal filter, considering that a crystal oscillator is a circuit, not a resonator? Both quartz crystals and ceramic resonators are used in both crystal filters and oscillators. How do we organize these 4 overlapping topics so there is minimum redundancy and readers can find information with a minimum of clicking back and forth? The common factor is they are all piezoelectric resonators. We can move information common to several articles to a Piezoelectric resonator article, allowing the other articles to focus on their specific topics. --ChetvornoTALK 11:20, 24 January 2013 (UTC)[reply]
I have inserted a new disambiguation page: Piezoelectric resonator (disambiguation) because to merge quartz crystal resonators, ceramic resonators and MEMS resonators is not a good idea. The components are too much different. For the existing article “Crystal oscillator” I think someone have to split up it into one article about the component “quartz crystal resonator” and a second article about the electronic circuit “Quartz crystal oscillator”. But this costs a lot of time. In the meantime I propose to insert a info-box at the top of the article, saying:
This article describes the electronic component “quartz crystal resonator” as well as the electronic circuit “Quartz crystal oscillator”. Please help to split up the text
This would make the international comparison of the terms a little bit easier. --Elcap (talk) 10:42, 27 January 2013 (UTC)[reply]
The disambiguation page is a good idea but I oppose the infobox. The consensus on this page seems to be against our position that the article should be split up. We should respect the consensus opinion. --ChetvornoTALK 14:02, 27 January 2013 (UTC)[reply]

Hi Chetvorno, thanks, but I am not sure if I chose the correct way for that, what I name "disambiguation page". It is more or less a "Please see" page. I don't know the rules of the English Wiki in this case. Please check and correct it, if I did it wrong.

But in the case "split the article" we all are not free in our decision. Both the component “quartz crystal resonator” as well as the electronic circuit “Quartz crystal oscillator” are industrial products. And industrial products follow strict rules, the national (DIN, BS...) or international (IEC) standards. In some years surely someone will come and say, the articles has to follow the international harmonized standards and has to be splitted. So far I know the international harmonizing process (IEC) is running (The speed here is more a "creeping") so we can wait some years. Regards --Elcap (talk) 08:51, 28 January 2013 (UTC)[reply]

The electronic circuit symbol shown at the top of the page is for a crystal, not an oscillator circuit. 107.13.240.126 (talk) 04:03, 4 June 2018 (UTC)[reply]
Inside the box is a picture of a crystal, not an oscillator, and the symbol is for the crystal. The heading outside the box matches the heading for the page. If you want to change the heading for the box, I won't revert, though others might. As above, the difference between crystal, and crystal oscillator is often small, as enough gain is available in existing circuits. There are some complications, in some cases. Overtone crystals need a filter, to get to the right resonance. An article on crystal oscillators, minus any discussion of crystals, would be pretty useless. (As noted, a CMOS inverter will do it.) Gah4 (talk) 05:12, 4 June 2018 (UTC)[reply]

What is this a Graph of ?

[edit]

The vertical axis "spurious response" graph is not labelled, except as "dB", which we can assume represents a power ratio. But by what conceptual chicanery can a power ratio be seen as a characteristic of a crystal ? Unless this is explained the graph is meaningless. — Preceding unsigned comment added by 77.96.59.93 (talk) 10:08, 18 January 2014 (UTC)[reply]

Suggest Regrade

[edit]

Who rated this article C ? 'Start' seems more appropriate : "Provides some meaningful content, but most readers will need more". Or even 'Stub': "The provision of meaningful content should be a priority". The reason I am suggesting this is that I understand a 'crystal oscillator' to be a certain class of electronic oscillator circuit, as defined in the first sentence of the article, yet the article contains almost no information on that topic. AJS 77.96.59.93 (talk)

Further reading

[edit]

A long list of general references for papers by Poddar and Rohde have just been added by user:Drakpoddar who would seem to be connected judging by the user name. Do we really need all of these? It smacks of promoting ones own stuff to me. SpinningSpark 00:36, 1 March 2014 (UTC)[reply]

Absolutely. This is strongly discouraged in WP:Further reading. In addition to COI issues, most should be removed just on the issue of balance: "Further reading" should offer a range of different sources. In addition they look like the subject matter is somewhat redundant. --ChetvornoTALK 01:02, 1 March 2014 (UTC)[reply]
Any view on which ones we should keep? SpinningSpark 00:44, 2 March 2014 (UTC)[reply]
Take your pick. I tend to think only one of them at most should be kept, there are plenty of other good sources. The multiple titles are basically BOOKSPAM. That was a good catch. --ChetvornoTALK 01:55, 2 March 2014 (UTC)[reply]

Crystal Brand

[edit]

Hey, as a Precision Devices, Inc employee, I recognize the crystals in the images. Perhaps this is relevant information. — Preceding unsigned comment added by 71.13.246.85 (talk) 04:31, 8 March 2014 (UTC)[reply]

I don't think so. Numerous manufacturers make crystals in the standard HC49/S package. There is nothing useful to be gained for the reader by promoting the Precision Devices product which is entirely unremarkable against competitors. Also, to make that statement would require a reliable source saying that they are Precision Devices product. There is no marking (other than the frequency) on either of the HC49/Ss in the lede photos identifying the brand so we are not going to say that even if it was worth saying. The crystal oscillator in the second picture does seem to have a brand name. I cannot read it very well, but it does not seem to be Precision Devices. SpinningSpark 15:55, 10 March 2014 (UTC)[reply]

Crystalline or not?

[edit]

The most common type of piezoelectric resonator used is the quartz crystal, so oscillator circuits incorporating them became known as crystal oscillators,[1] but other piezoelectric materials including polycrystalline ceramics are used in similar circuits.

Note that polycrystalline is still crystalline, so that doesn't disqualify it. But yes, the usual high-Q oscillators are, single crystal. As far as I know, there are no amorphous quartz oscillators. Gah4 (talk) 17:32, 10 August 2015 (UTC)[reply]

There are amorphous ceramic resonators, although I can't think of amorphous quartz ones. Andy Dingley (talk) 17:34, 10 August 2015 (UTC)[reply]
But are they piezoelectric? I would think you would need atoms arranged in a crystal lattice to have piezoelectricity. Polycrystalline ceramic resonators have their crystal grains aligned during manufacture to give them piezoelectric properties. In an amorphous material, without a crystal lattice I don't see how you could have piezoelectricity. --ChetvornoTALK 23:40, 25 August 2015 (UTC)[reply]
They don't need to be. The transducer can be deposited as a separate surface layer. Andy Dingley (talk) 00:36, 26 August 2015 (UTC)[reply]

Questions and comments

[edit]

I read sections 1 through 7 of the article and I did not enjoy this article at all. I feel it's too heavy for a casual reader and, honestly, even to a technically inclined reader.

1. The first image on the top right shows a quartz crystal resonator and quartz crystal oscillator.

What is the difference between the two? Why is the quartz crystal resonator shown?

2. "The first crystal-controlled oscillator, using a crystal of Rochelle salt, was built in 1917 and patented[4] in 1918 by Alexander M. Nicholson at Bell Telephone Laboratories, although his priority was disputed by Walter Guyton Cady."

How was it disputed? What was the outcome?

3. "Prior to crystals, radio stations controlled their frequency with tuned circuits, which could easily drift off frequency by 3-4 kHz."

Was the drift due to temperature or something else?

4. "Both left and right-handed crystals can be used for oscillators, if the cut angle is correct."

This sentence is somewhat ambiguous.

5. "Care must be taken during manufacturing and processing to avoid phase transformation. Other phases, e.g. the higher-temperature phases tridymite and cristobalite, are not significant for oscillators."

It would help to introduce the idea of phase transformation before mentioning types of phases.

6. The "Crystal structures and materials" and the "Stability and aging" sections have too much information. They should be simplified, divided in sections or they should be spun off into several articles.

7. "Details were published in "Radio" magazine (3/1978) by UB5LEV."

Does it really need to be part of the text?

8. What is an "AT cut"? Unlike the "SC cut", it's not explained what "AT" stands for.

ICE77 (talk) 03:31, 26 August 2015 (UTC)[reply]

1. This was the subject of a long discussion on this page; see Split article? and Rename this page Piezoelectric resonator? above.
  • A "crystal oscillator" is an electronic oscillator circuit that uses a piezoelectric resonator, a crystal as its frequency-determining element.
  • A "crystal" is the common term used in electronics for the frequency-determining component, a crystal of quartz or a ceramic wafer with electrodes connected to it. A more accurate term for it is piezoelectric resonator.
Piezoelectric resonators, "crystals" are sold as separate components for use in crystal oscillator circuits, which is what the left image shows, but they are also incorporated in a single package with the crystal oscillator circuit, which is what the right image shows.
This article is named Crystal oscillator, so its subject should be the circuit itself, not the crystal. However the article has almost no content on crystal oscillator circuits; virtually all of this article is about the crystals, piezoelectric resonators, not the circuit. So about 95% of the article is off-topic. This was causing confusion among general readers, who didn't understand whether the term "crystal oscillator" referred to the circuit or the crystal. I suggested spinning off the content on the crystal itself into a new article, Piezoelectric resonator, leaving this article to cover the circuit, but the general opinion was that the two topics are too closely related and should be covered in one article. I can see this point of view.
6. I agree, these sections are bloated. This could be another reason to spin off the crystal content into a Piezoelectric resonator article.
--ChetvornoTALK 12:30, 26 August 2015 (UTC)[reply]
Anyone remember when Transistor meant Transistor radio, when Microwave didn't mean Microwave oven, before App was short for Application, and Application didn't yet mean Application program? When crystal was a leaded glass (and not crystalline) that you put on the dining room table? (But don't drink out of it.) If no other reason than to distinguish from tableware, it seems that crystal oscillator is a fine name to me. Gah4 (talk) 19:21, 26 August 2015 (UTC)[reply]
That's what the circuit that uses the crystal is called. Take a look at the article. The term "crystal" and "crystals" appears many times, referring to the resonator itself. It is kind of unavoidable, that is what the electronics industry calls a piezoelectric resonator. --ChetvornoTALK 20:28, 26 August 2015 (UTC)[reply]

1s/30 year = 10^-7 ?

[edit]

>> With accuracies of up to 1 sec in 30 years (30 ms/year or 10−7)

One second in 30 years makes 10^-9 - a hundred time better accuracy. 10^-7 (3 seconds per year) seems to better reflect early quartz oscillators accuracy. — Preceding unsigned comment added by 87.205.153.50 (talk) 19:45, 8 August 2016 (UTC)[reply]

Worldwide Crystal Oscillator Production

[edit]

The article currently states that "More than two billion crystals are manufactured annually", but this is uncited and seems extremely low. ARM claims 10 billion chips were shipped with their IP in 2013 (http://www.anandtech.com/show/7909/arm-partners-ship-50-billion-chips-since-1991-where-did-they-go) and it seems likely that almost all of those used a crystal oscillator in their eventual application. Do any of the contributors to this page have a good source for crystal production numbers? — Preceding unsigned comment added by 76.102.118.240 (talk) 21:21, 22 January 2017 (UTC)[reply]

Since 10 billion is more than 2 billion, the statement is not wrong. Gah4 (talk) 22:56, 13 May 2021 (UTC)[reply]

Resonant

[edit]

There is a change from "resonance" to "resonant". Seems to me that "resonance frequency", meaning "frequency of resonance" was fine, but "resonant frequency" should be fine, too. Though sometimes "resonant frequency" sounds too specific to me. For one, there are higher order frequencies of resonance, not quite integer multiples, of the fundamental. Also, the frequency is different for series and parallel resonance. To me, "resonance frequency" allows for the other possibilities more than "resonant frequency". Gah4 (talk) 19:06, 2 February 2017 (UTC)[reply]

"Resonant" is an adjective, so it should be used for "resonant frequency".
"Resonance" is a noun that describes the state. A noun can be in a prepositional phrase, so "frequency of resonance" is fine but sounds a bit stilted.
The noun-noun combination "resonance frequency" is understandable, but it sounds a bit odd to me, and I would avoid it given the established adjective "resonant". English has noun-noun combinations such as "water tank" and "boy friend", but I don't think they have adjective forms. I cannot think of an adjective that means "water" ("hydro" is a noun), and "moist tank" does not convey the right meaning. Similarly, "boyish friend" means something different than "boy friend". Glrx (talk) 19:29, 2 February 2017 (UTC)[reply]
There is hydrated, but that doesn't really explain a water tank. Gah4 (talk) 08:45, 3 February 2017 (UTC)[reply]
I agree. I have my own pet theory why "resonant frequency" has been the dominant form for 100 years. In science the term "resonant frequency" is always applied to a specific object, a "resonator" such as a quartz crystal, tuned circuit, or speaker cone, not to the abstract concept of "resonance". So the term "resonant frequency" is short for "frequency at which this object is resonant", explaining why the adjective form is favored. "Resonance frequency" would translate as "frequency at which resonance (in the abstract) occurs" which doesn't make much sense to a scientifically-educated person, since the abstract concept of "resonance" doesn't have a frequency. That's why this form sounds a bit odd to us. --ChetvornoTALK 20:24, 2 February 2017 (UTC)[reply]
Except when you do want the more abstract form. For low-Q resonators, that might make sense. Or the cases I indicated above. Gah4 (talk) 08:45, 3 February 2017 (UTC)[reply]
Thinking about this one again, it seems that resonant frequency suggests that there is only one, where resonance frequency works when you are using one of the many resonant modes. As well as I know, crystals always have more than one possible mode. Gah4 (talk) 23:02, 13 May 2021 (UTC)[reply]
[edit]

Hello fellow Wikipedians,

I have just modified 14 external links on Crystal oscillator. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 03:56, 15 August 2017 (UTC)[reply]

assemblers

[edit]

Not so long ago, I started a discussion in the assembly language article on the possibility of splitting off a separate article on assemblers (the programs that process assembly language programs). More specifically, the abstract language, and the actual implementation of language processors. As far as I know, it hasn't been done yet, but the discussion about crystal vs. crystal oscillator seems similar. There is a fair amount of art to writing assemblers, such as hash tables for symbols, multiple passes to resolve backward references, and such. All of which are independent from the abstract language. It might be that there is enough here to write about crystal oscillator circuit implementations, separate from the crystals themselves. I suspect not, but then I haven't tried. There is a whole Theory of Oscillators[1] which might help in such a page. (I have the book, but haven't finished it.) Gah4 (talk) 05:42, 4 June 2018 (UTC)[reply]

  • Oppose I'm not seeing any advantage to that. There are two topics, but one really needs the other to make a readable article, and the oscillators content wouldn't be huge. They'd still work fine as a single article. Andy Dingley (talk) 10:47, 4 June 2018 (UTC)[reply]

References

  1. ^ Andronov (November 2, 2011). Theory of Oscillators. Dover. ISBN 0486655083.

AT crystal frequency characteristics

[edit]

The AT xtal's frequency-temperature charactersitics is described as sinusoidal. This is incorrect. It is approximates a third order curve with the "true AT" curve being close to a cubic with an inflection temperature close to 26degC. Crystal cuts close to AT have additional linear coefficientsPhysicistQuery (talk) 23:01, 29 June 2021 (UTC)[reply]

The first two terms of the Taylor series for sine are linear and cubic. You might say cubic is a good approximation for sine, or sine is (over the same domain) a good approximation to a cubic. I don't know what the next term is for an AT crystal. One way to minimize temperature effect is to use a region with a derivative of zero. It seems that in this case, they use a region where the variation in minimized over the appropriate range, so a little over one cycle of a sine, if you approximate it as sine. So, what is the fifth order term? Gah4 (talk) 01:37, 30 June 2021 (UTC)[reply]
It seems that the favorite reference is here.[1]

References

  1. ^ "Some improvements in quartz crystal circuit elements". Bell System Technical Journal. July 1934. doi:10.1002/j.1538-7305.1934.tb00674.x. Retrieved 30 June 2021.

Schematic?

[edit]

There is not a single practical circuit in this voluminous article. Schematics are the coin of the realm. This page has nothing to do with practical applications. How much more basic can a topic be? 71.8.170.195 (talk) 11:51, 26 July 2022 (UTC)[reply]

Because this is an encyclopedia, not an electronics textbook? We hardly need *this* article, let alone decorating it with graphics that will be fiercely contested for their wrong-ness. Put your favorite schematic here and watch the teapot lid rattle. --Wtshymanski (talk) 22:48, 13 August 2022 (UTC)[reply]
I haven't thought about this for a while, but we used to do it with three CMOS inverters in series. That avoids the need for detail (the insides of the inverters) but gets the idea across. It might be that one is enough, but I mostly remember three. Gah4 (talk) 12:23, 15 August 2022 (UTC)[reply]
Do you think that will satisfy our commentator above? And you'll have to explain why three, and not five or seven or one or two. And crystal oscillators can be made with everything from ICs to vacuum tubes and tunnel diodes, so it will be several dozen pages of schematics that you'd need so as not to give the misleading idea that this is the ONLY way to build a crystal oscillator. International Crystal used to put out a 60-page handbook that only talked abou the models *they* sold - a subset of all crystal oscillator designs. Like I said, this is an encyclopedia article, not a textbook of electronics. --Wtshymanski (talk) 16:17, 22 August 2022 (UTC)[reply]
@Wtshymanski: And yet the article has 15 pictures of crystals, crystal packages, and everything to do with them, and sections on every other possible peripheral topic including an 8 paragraph bloviation on radiation damage, all of which you have no problem with? Sounds like you are only opposed to adding content which is WP:ONTOPIC. --ChetvornoTALK 18:14, 25 September 2022 (UTC)[reply]
What a spirit of collegiality is represented here. It wasn't 8 paragraphs on radiation damage, by the way. The article is greatly overgrown and needs editing. We're not writing textbooks here. Have at it, edit out things you find redundant or off topic. After all, it is the encyclopedia anyone can edit. --Wtshymanski (talk) 02:36, 29 September 2022 (UTC)[reply]

I absolutely agree, 71.8.170.195 and Gah4, great point. This has been brought up repeatedly on this page, see Talk:Crystal oscillator/Archive 1#Only the quartz crystal or more, Talk:Crystal oscillator/Archive 1#What does it do exactly?, #Split article? and #Rename this page Piezoelectric resonator? above. This article is named "Crystal oscillator" but there is only one small section about oscillators in it; the rest is about crystals. And that section not only does not have a single schematic, but only devotes a paragraph to how it works. It doesn't mention important points like the Barkhausen stability criterion, poles, startup requirements, and typical circuits used. This section desperately needs some schematics, and a better explanation of the basics. Anyone want to work on it? --ChetvornoTALK 04:03, 25 September 2022 (UTC)[reply]

Frequency range inconsistency

[edit]

The article says "Depending on the manufacturer, the highest available fundamental frequency may be 25 MHz to 66 MHz". The Crystal oscillator frequencies article shows frequencies up to 1075.804MHz so the statement seems to be inconsistent. ICE77 (talk) 00:21, 25 September 2022 (UTC)[reply]

Higher frequencies are usually overtone crystals. Though the traditional way, and maybe still, is to put a frequency multiplier after the crystal. The thickness is inverse proportional to frequency, and some would just be too thin. Gah4 (talk) 05:54, 25 September 2022 (UTC)[reply]
OK, it is about 5km/s, so thickness = (5km/s)/(2f) at 1GHz it would be 2.5 microns. That includes the electrodes. Gah4 (talk) 05:59, 25 September 2022 (UTC)[reply]

Gah4, thanks for the information. That clarifies. ICE77 (talk) 19:37, 22 October 2022 (UTC)[reply]

axes

[edit]

While crystallography is commonly done with a, b, and c axes, this article uses the x, y, and z axes. Is there a reason for that? Gah4 (talk) 10:20, 29 September 2022 (UTC)[reply]

Because they are the terms used by the manufacturers of crystal resonators? 86.162.147.159 (talk) 11:46, 29 September 2022 (UTC)[reply]
The current reference 25, this one:[1] calls them a, b, and c, though the paragraph that the reference is used in calls them x, y, and z. Gah4 (talk) 20:28, 29 September 2022 (UTC)[reply]
Nowhere in that reference does it discuss crystal resonators, so my point still stands unchallenged. 86.162.147.159 (talk) 12:09, 30 September 2022 (UTC)[reply]

References

  1. ^ The Quartz Page: Quartz Structure. Quartzpage.de (2010-10-23). Retrieved on 2012-06-21.