Contents:Home PageFAQDr. John Vig's TutorialsRecommended ReadingIndustry LinksGlossaryHome Page Of:Florida Technical Writing |
Technical Information Piezoelectric Quartz Crystals And Crystal Controlled Oscillators Frequently Asked Questions About Quartz Crystals
Please note that while most of these questions and answers are applicable to a variety of crystal unit types, they are intended to be specific to the "thickness-shear" "A-T" and "B-T" crystal units. What is a Piezoelectric Quartz Crystal Unit? What is a Thickness-shear A-T or B-T cut Crystal Unit? What is the difference between a fundamental and an overtone crystal? How do I know if I need a fundamental or an overtone crystal? What is the difference between a "crystal" and a "strip resonator"? If a "strip resonator" is called a "strip resonator," why isn't a "crystal" called a "crystal resonator?" What about "ceramic resonators?" What is the difference between a series and a parallel crystal? What is load capacitance? Will a series crystal work in a parallel circuit and vice versa? What is pullability? Why can't I get a crystal with symmetrical pulling characteristics? What is trim sensitivity? Do I need to be concerned about trim sensitivity? What are the motional and shunt capacitances of a crystal unit? What are spurs? Do I need to be concerned about spurs? What are "coupled modes" and do I need to worry about them? What are "activity dips" and do I need to worry about them? Can I safely change from a conventional crystal to a "surface mount device"? Why does it take so long to get a two piece sample? I waited for my sample crystals quite a while. When I got them, I found that they first oscillated at some frequency not at all close to what I specified. If I touched a leg of the crystal with a scope probe or even with my finger, they began oscillating at the correct frequency. What's up with that? What is a Piezoelectric Quartz Crystal Unit? A Piezoelectric Quartz Crystal Unit is an electronic component used in frequency control, filtering, and clocking applications. It consists of a quartz resonator equipped with electrodes, housed in a hermetically sealed package that provides some means of connection to a circuit. Piezoelectric Quartz Crystal Units are usually called "crystals." What is a Thickness-shear A-T or B-T cut Crystal Unit? "Thickness-shear" describes the motion of the quartz plate as it oscillates. "A-T" and "B-T" describe resonator plates cut from the original stone at precise angular orientations that are well known to quartz crystal unit manufacturers. What is the difference between a fundamental and an overtone crystal? A fundamental mode crystal oscillates at a frequency determined by the physical dimensions of the quartz plate. The fundamental frequency is the lowest frequency at which a given resonator plate will oscillate. Overtones are frequencies that are approximately odd integer multiples of the fundamental. How do I know if I need a fundamental or an overtone crystal? Consult with your crystal unit vendor as early in your design process as possible. Remember that at some frequency which varies from manufacturer to manufacturer, the use of an overtone is mandated. Remember also that within a certain frequency range, the fundamental frequency and the overtone frequencies may overlap. In such a case, your selection may be determined by your application. If you are developing an oscillator that is to be extremely stable and accurate, with a high "Q" value, you probably need an overtone. If you require a lot of pullability, you probably need a fundamental. Again, consult with your crystal vendor as early in the design process as possible. What is the difference between a "crystal" and a "strip resonator"? The difference is primarily one of geometry. A "crystal" is usually thought of as a device using a disk shaped quartz plate while a "strip resonator" uses a rectangular quartz plate. The operating characteristics of the two may differ significantly. If a "strip resonator" is called a "strip resonator," why isn't a "crystal" called a "crystal resonator?" Actually, many people do call a "crystal" a "crystal resonator." In the case of the "strip resonator" I believe that when they were first appearing in the market place, some shrewd marketing person started referring to them that way as if they were in some way superior to a "crystal."
What about "ceramic resonators?" A "ceramic resonator" is a resonator fabricated from a piezoelectric ceramic material, of which there are several. These materials are not naturally piezoelectric, the property is induced in them during manufacture. Some of the materials used for this purpose are quite remarkable and represent major engineering accomplishments. To my knowlege, however, none of these devices can compete with quartz in terms of frequency stability, particularly over a range of temperatures, nor can their operating frequencies be held to any real accuracy. These types of resonators exhibit "Q" values much lower than those manufactured from quartz. These devices, I believe, tend to age far more than do quartz crystals. Still, if a very low-cost, loosely specified device is suitable for an application, piezoelectric ceramics have much to commend them. What is the difference between a series and a parallel crystal? Physically the two are identical. The difference lies in the fact that a parallel resonant crystal is adjusted to frequency with the specified value of load capacitance in place. A series resonant crystal requires no load capacitance and therefore none is used. Load capacitance is defined as being the total capacitance present in an oscillator circuit as measured or calculated across the pins of the crystal socket. Load capacitance has the effect of increasing the frequency of a crystal unit. Please see the tutorial section for a complete explanation. Will a series crystal work in a parallel circuit and vice versa? Yes to both, but a crystal unit manufactured as a series resonant crystal will operate at a frequency higher than expected if used in a parallel resonant circuit. A crystal unit manufactured as a parallel resonant crystal will operate at a frequency lower than expected if used in a series resonant circuit. Pullability is the amount by which the frequency of a crystal will change when the circuit condition is switched from series to parallel resonance. Pullability is also used to describe the frequency change that occurs when the load capacitance is switched from one value to another. Please see the Technical Note on Frequency Pullability for a more complete explanation. Why can't I get a crystal with symmetrical pulling characteristics? Quartz crystals are not linear devices and they do not behave in a linear way. But, it is theoretically possible to achieve symmetrical pulling. Please see the Technical Note on Frequency Pullability for an explanation. Trim sensitivity is the amount by which the frequency of a crystal oscillating with a specific value of load capacitance will vary if that load capacitance is varied slightly about its nominal value. Please see the Technical Note on Frequency Pullability for a more complete explanation. Do I need to be concerned about trim sensitivity? A good question and the answer is an unequivocal "maybe." If the crystal unit in question has a significant value of trim sensitivity and if a fairly stringent value of frequency tolerance is imposed, it is likely that the manufacturers tolerance on the load capacitors in use will result in actual values of load capacitance that will "pull" the crystal frequency beyond the desired tolerance. As an example, an 18.432000 MHz crystal might easily have a trim sensitivity value at 20.0pF of +/-20.0ppm/pF. Let us suppose that a frequency tolerance of +/-10ppm is specified. If the load capacitor has a tolerance of +/-10%, or 2.0pf, it is entirely possible that an acceptable load capacitor will pull the frequency by +/-40ppm. In such a case, the user would never consider the tolerance on the load capacitor but would instead unfairly blame my crystal unit. What are the motional and shunt capacitances of a crystal unit? The shunt capacitance (Co) is the capacitance resulting from the presence of the electrodes on the quartz plate plus the capacitance inherent in the crystal holder. The motional capacitance is a parameter of the equivalent circuit. It is used as a means of describing the elasticity or "stiffness" of the quartz resonator. Please see the Technical Note on the Equivalent Circuit for a more complete explanation. Spurs are frequency responses higher in frequency than the main response but not as high in frequency as the next regular overtone. The word "spur" is used as an abbreviation of the word "spurious" but the frequencies described by either of the two are not "spurious" at all. They are regularly occurring natural frequency responses, the amplitude of which is subject to some degree of control through crystal unit design. Please see the Tech Note on Spurious Responses for a more complete explanation. Do I need to be concerned about spurs? Another good question and another unequivocal "maybe." Usually, spurs are not a problem with crystal units intended for use in oscillator applications. Crystal units intended for use as filters are another matter. The control and suppression of spurious responses in filter crystals is critical. If you must specify some value of spurious response suppression, PLEASE specify the test fixture to be used (IEC 60444 Pi Network is a good choice for oscillator crystals) and a reasonable frequency range over which the test is to be conducted. What are "coupled modes" and do I need to worry about them? This is an excellent question. Guess what? Please see the Technical Note on coupled modes in quartz crystals. The short answer is that you as a user should not have to be concerned with coupled modes. What are "activity dips" and do I need to worry about them? "Activity dips" are sudden increases in the resistance of a crystal that may perturb the frequency. They are caused by moisture within the crystal package and are not generally seen until the crystal is heated from cold to hot while oscillating. In general, the end user of a crystal unit should not have to worry about activity dips. Please see the Technical Note on Activity Dips. Can I safely change from a conventional crystal to a "surface mount device"? I would not make the change without a thorough investigation. Some of the SMD (surface mount device) crystal units are comparable to conventional crystal units, others are not. In general, SMD's have higher resistance, differing values of shunt and motional capacitances, and are more sensitive to drive level. The pullability characteristics of these devices may differ significantly from a conventional crystal. I would recommend a fairly exhaustive qualification sequence before making the switch. Please see the Technical Note on Surface Mount Devices. Why does it take so long to get a two piece sample? Even a relatively simple, straightforward crystal unit requires ten to fifteen separate and distinct manufacturing stages. If you specified the full range of your unique requirements, chances are that your samples had to be manufactured from scratch. Had you been willing to accept a "standard" part at your specific frequency in order to evaluate the vendor's quality, you probably could have gotten your samples much faster. I waited for my sample crystals quite a while. When I got them, I found that they first oscillated at some frequency not at all close to what I specified. If I touched a leg of the crystal with a scope probe or even with my finger, they began oscillating at the correct frequency. What's up with that? Without knowing the details, my supposition would be that you ordered a series resonant crystal on the assumption that you had a series resonant circuit. The frequency you see at first is not the crystal frequency. It is some frequency resulting from an "L-R-C" configuration. In other words, your circuit does not generate enough noise to start the crystal and the circuit is "free-wheeling." By touching the crystal with a probe or with your finger, you are introducing noise into the circuit and therefore the crystal starts. This is one of the drawbacks to a series resonant circuit. | Home | FAQ | Tutorial | Tech Notes | Reading List | Industry Links |Glossary | Email |Created and maintained by: Florida Technical Writing |