In a non-critical low budget, for example, a 10-US mass market electronic thermometer, this clock can be made of a simple resistance / capacitor (RC) oscillator. However, for the vast majority of more critical, the oscillator is based on quartz crystals (Fig. 1). This is a kind of maturity (more than 80 years) and efficient technology that supports various frequencies from KHz to hundreds of MHz, with performance from excellent to excellence, depending on crystal cutting, manufacturing, packaging, and other considerations.
Figure 1: Ancient quartz crystal (but not the entire oscillator) is represented by a standard schematic symbol; b) The equivalent circuit starts from the simplified model shown, but as the operating frequency increases can become more complicated.
However, the progress of the crystal has reached a stable level, and the performance, size and cost of timing functions are summed, and the integration is increasing. In order to meet these needs, a new destructive method begins eroding quartz devices based on silicon MEMS (microcircular system) technology, which provides quartz grade performance, and performance and cost levels apply to many applications. MEMS devices have been highly developed, a large number of inductive pressure, motion and acceleration, now they are extending to new roles.
The requirements for timing functions in RF applications are particularly challenging, and the oscillator is more than just the clock of the processor, but also tolerate a little jitter. In RF, it establishes basic carrier / channel tuning within hundreds of MHz and GHz to ensure the correct clock of the A / D and D / A converters. For converters, any jitter will be converted to converter noise and distortion, so it is a key specification in the RF design.
Oscillator operation
The structure and operation of the crystal timing device is based on well known piezoelectric principles, that is, the electrical signal generates stress in the crystal, and vice versa: the applied stress causes the crystal to produce a small voltage. By using a tiny slate or quartz blank and a suitable circuit, quartz provides an accurate interval clock signal for the entire electronic system as a tuning resonator.
In MEMS-based devices, use a completely different way. Etch silicon at the core of the chip is like a tuning fork, which is resonant at the desired frequency, while the additional electronic circuit on the chip manages and enlarges this clock signal (Figure 2).
Figure 2: MEMS oscillator technology uses a type of tuning fork in a silicon, and support circuitry. (Provided by SitiMe)
There are many first, second, and even the third layer parameters for assessing any oscillator, whether crystals or other oscillators. Of course, the minimum or maximum value required depends on the application, but the relative weight of these parameters varies with the design.
Key parameters include nominal operating frequency, absolute precision, aging-related stability, short circuit and long-term drift (temperature coefficient and compensation), jitter, operating temperature range, package, size, operating voltage, power sensitivity, power consumption, anti-impact / vibration , Start time, supplier changes, and cost, reference A minority. Based on application needs and any historical background, most of these are legally measured in different ways and different conditions.
MEMS advantage and reality
Based on the quartz crystal oscillator, it is assembled by a plurality of components, including a precision cut and polished quartz blank, which is fixed to the package in the package, and also provides electrical contact (and some anti-impact / vibration), and the housing package itself ( For more backgrounds, see the CTS Product Training Module "Crystal Clock Oscillators). In contrast, the MEMS oscillator is an IC that is manufactured using a standard process CMOS production line, and an 8-inch wafer is used in most cases. After the detection, trimming and testing, the device is packaged; again, like any IC. Therefore, MEMS devices benefit from batch production techniques and processes for traditional ICs. (For other contexts about the MEMS oscillator, see the AbraCon product training module on the ASFLM1 series). Other advantages for MEMS-based devices include:
The final device is smaller than the quartz version. This can not only save valuable PC board space, but also allow timing devices close to the device it supports to achieve better signal integrity and reduce EMI.
The MEMS oscillator can build an active circuit on the chip, which can be used to compensate circuitry, improve performance and temperature or power rail. It can also be used to provide complete oscillator functions because quartz crystals and MEMS resonators are not a complete oscillator (although the term usually uses this manner); each requires some related circuits to drive core timing components And regulate / zoom output. Many oscillators also require PLL to multiply the basic oscillator frequency by the required carrier frequency, which can also become part of the IC.
Complete MEMS oscillator kernel, oscillator circuit and interface power consumption
In addition, it is working to allow the MEMS device chip to co-encapsulate in the same manner as the memory IC. Now with their microcontroller or microprocessor stacking and common packaging. This will bring a variety of benefits: there is a need for less board space, simplified BOM, improved single integrity, and the performance of the oscillator and converter, without considering the PC layout problem (these issues are GHz's RF range is challenging and often frustrating in all these advantages, and there are several reasons for the MEMS device without replacing crystal oscillators:
The performance of available MEMS devices may not be like this.
RF designers are very cautious because timing functions are critical to system performance.
Although the crystals have shortcomings and artifacts, these are quite good to understand. In contrast, the subtleties of MEMS devices and the unpredictableness of the MEMS device are just beginning to be known for the frontier design accepted by the RF designer.
The new RF design typically contains one or several new product listing components such as high performance LNA or A / D and D / A converters, but designers are reluctant to try too many new components. This is about risk management and how many unfamiliar uses of designers, even if each device has potential benefits.
Cost, of course: As a mature technique, crystal suppliers have tried to reduce costs through empirical and quantity. Although MEMS devices have potential to reduce costs, this must be evaluated according to specific situations.
MEMS oscillator becomes ready-made standard parts
Some available MEMS timing devices illustrate the functions of these components. For example, Sitime Sit8209 high frequency, ultra-high performance oscillator (Figure 3) can order any frequency between 80.000001 and 220 MHz, accurate to six in the decimal point. In order to transition, it is packaged for the pin to the pin to the pin, which has only 0.5 psec ultra-low phase jitter, and as low as ± 10 ppm frequency stability. In addition, Sitime also provides a number of MEMS device families applicable to different application design requirements.
Figure 3: Sitime sit8209 provides extremely low jitter, vital to many communication applications; shown in the figure, when the LVCMOS output is employed in 3.3 V, the phase noise is 156.25 MHz.
Silicon Labs offers four series (Si501, Si502, Si503, Si504), and its members have different features in additional function, and the performance guarantees 10 years of frequency stability, including solder offset, load traction, power changes, operating temperature range, Vibration and impact; the supplier claims that this is a 10-fold guarantee than quartz equipment. These units provide any frequency between 32 kHz to 100 MHz, and the frequency stability option includes ± 20, ± 20, ± in the temperature range of commercial (-20 ° C to 70 ° C) and industrial (-40 ° C to 85 ° C). 30 and ± 50 ppm. The quadratic device (Figure 4) can work on any power rail between +171 V to +3.63 V.
Figure 4: Silicon Labs of Si501 Silicon Labs have the same basic performance specifications, but the availability of additional functions such as output enabling and frequency selection is different.
Micro's Micrel MEMS oscillator unit (Fig. 5) can operate at frequencies of 2.3 to 460 MHz (e.g., the frequency of DSC 1123 is 156.25 MHz). Typical RMS phase jitter is less than 1 Psec, while stability can order ± 10, ± 25 or ± 50 ppm rating. The LVDS output device is packaged in 2.5 × 3.2 and 7.0 × 5.0 mm package, suitable for existing packages, require 2.25 to 3.6 V power supply. The supplier claims that MTF (average fault time) is 20 times better than quartz equipment.
Figure 5: The DSC unit of Micrel is a "direct" alternative to standard 6-pin LVDS quartz crystal oscillator; the different devices are only enabled to control the pin.
Summarize
It is difficult to predict that MEMS-based timing devices will replace units in historic quartz-based crystal RF design, and how long is this transformation takes. There is no doubt that the advantages of MEMS units and their future potential advantages in performance, size, costs, and packages make them very attractive competition at lower frequencies, and increasingly entering higher RF spectrum. . The market research company recently predicted that more than 1 billion MEMS timing devices will be shipped in 2016, mainly for mobile phones and consumer electronics. The supplier sees the opportunity, MEMS technology has been used in the mass market, it is promoting and mature, as long as any weighing - obviously change due to application - users can accept, users will benefit.
Be
Reprinted from -Wiku Electronic Market Network
Our other product:
