From early radios to radar, and now digital computing, every circuit requires a clock or heartbeat that guides its functionality. Timing controls the processing rate, data connectivity and RF transmission bands in a wide variety of applications from low power to high precision. Timing has become a diverse field of engineering. Considering the many ways a clock circuit can be designed as well as the many advances introduced annually into the industry, engineers should revisit their timing considerations on a regular basis. Below is a list of basic timing devices and the best time to use them.
1. LC resonator
The LC resonator is the simplest and commonly used timing circuit, composed of an amplifier, an inductor and a capacitor. The main advantages include low cost and ease of integration, especially at high frequency. However, it is not very accurate and varies widely over temperature. This variability gives one additional property: wide pull range. For this reason, the LC is the resonator of choice when developing small or highly integrated voltage controlled oscillators (VCOs). These oscillators either on a PCB or on-chip are designed to track or lock to other frequencies. The LC is not accurate enough to run on its own since temperature can sway the frequency +/-10,000 ppm or more.
2. Ceramic resonator
The main advantage of the ceramic resonators is its cost. If you are looking for the lowest cost and a somewhat stable solution then this technology can get you there. Don’t count on stability less than +/-1000ppm over temperature. This resonator is low cost but cannot be used for precision or even partially precise timing. Generic applications such as toys, low end appliances and low end MCU applications can get away with this level of imprecision. If you need more precision, other resonators will get you there.
3. Quartz crystals
The quartz crystal reigns as the king of timing for its self-compensated temperature stability, excellent initial accuracy and moderate cost. As a resonator, it features a high Q which enables very low noise in-circuit. Mass production has fine tuned the accuracy and cost of these devices so moderately priced crystals can now achieve +/-20ppm to +/-50ppm overall accuracy. It provides excellent stability and serves as an ideal time base for many of today’s connectivity protocols from Wi-Fi, Zigbee and Bluetooth to automotive LIN/CAN, Ethernet, UART and industrial applications. Timing MCU’s and processors using quartz crystals provide accuracy that can meet the common connectivity protocols. However, there are protocols that require higher performance. Quartz’s precision can be enhanced.
4. Quartz oscillators (XO)
A quartz oscillator integrates an oscillator die along with a quartz crystal. It provides the accuracy and low noise benefits of quartz but decreasing variability induced by board traces. In some cases, the oscillator die also multiplies the base quartz frequency to one desired by the application. It becomes necessary to use a XO instead of a bare quartz crystal in very low noise systems such as high speed communications, optical interconnect, optical modules, test and measurement and advanced RF applications. A XO delivers low noise at a high frequency that is challenging to achieve using common crystals. Top frequencies such as 100MHz, 156.25MHz or 312.5MHz used in high performance systems require the conditioning with differential LVPECL, LVDS, HCSL or CML signal that a XO offers.
5. Temperature compensated quartz crystal oscillators (TCXO’s)
Although XOs provide buffering and often frequency translation, they track the accuracy of the quartz crystal blank. Several communications and telecom applications such as point-to-point RF, GNSS/GPS, mobile phones, LPWAN gateways and other precision RF connectivity systems require frequency stability from +/-0.5ppm to +/-2.5ppm over temperature. Stratum III requires +/-0.28ppm stability. The bare quartz is not stable enough to readily achieve stability levels below 10ppm. TCXOs go through a manufacturing flow that measures and calibrates their frequency deviation over temperature. The obvious disadvantage is cost. Keep in mind that nothing is more costly than an inoperable data link in your end system.
6. Oven controlled crystal oscillators (OCXOs)
OCXO’s can achieve a barely imaginable level of precision of +/-0.1ppm to 0.1ppb or better over temperature. Not only is temperature calibration used in TCXO technology. OCXOs achieve the stability by adding a second order of control—the temperature of the quartz blank. On startup, an OCXO will heat the quartz blank to about 10 degrees higher than ambient and will control the temperature at that level, minimizing temperature perturbations. In many cases, OCXOs are also mechanically guarded against shock and vibration, enabling end systems to achieve the maximum clocking precision for hold over requirements. Many applications related to military and radar, along with base transceiver stations (BTS) for mobile phones need this level of precision. Advanced high precision GPS in fast moving vehicles also require high precision. Only atomic clocks can deliver better timing precision.
7. Micro ElectroMechanical Systems (MEMS)
MEMS technology has evolved in parallel to quartz. Based on silicon instead of quartz crystals, MEMS has the advantages of miniaturization and resistance to shock and vibration. Due to complexities associated with MEMS resonators, the primary disadvantage of MEMS is cost. Although it may be used in a wide variety of applications covered by crystals, XOs and TCXOs, MEMS is optimal when high durability is required. Also, at ultra-small sizes such as 1.6 x 1.2mm footprint, MEMS becomes very competitive with crystals. Applications such as wearables, wireless charging pads, industrial controls, robotics, drones and AR/VR can take advantage of MEMS' durability and size.
In conclusion, knowing which technology to apply is key in developing a successful product. In timing, there is no shortage of technology. To learn more, visit Abracon.