±50ppm silicon oscillators herald the beginning of the end for crystal timing devices, writes Michael McCorquodale, general manager, silicon frequency control, Integrated Device Technology
Fuelled by consumer and industry demand for smaller, faster designs with greater levels of innovative functionality, semiconductor technology is moving forward at a rapid pace. But one thing that has remained unchanged for over 50 years is the use of quartz crystals as the primary frequency source for electronic circuits.
Until quite recently, semiconductor-based frequency control products have not had the accuracy needed to challenge the incumbency of quartz.
The status quo is now changing as timing solutions that use proven semiconductor processes are becoming available that not only offer a range of important benefits, but also expose the inherent weaknesses and frailties of quartz. Until recently, the specification that limited the ability of so-called ‘crystal-free’ oscillators to replace quartz crystal-based devices in mainstream applications was frequency accuracy.
The high accuracies increasingly demanded by the latest product designs were simply out of reach of what could be achieved through a silicon process such as CMOS.
However, all that has now changed with the emergence last year of extremely cost-effective ±100ppm accuracy parts, and very recently, similarly low-cost ±50ppm devices. Importantly, these higher accuracy parts are able to address the latest networking and telecom applications that demand greater frequency accuracy.
Now there really is a great opportunity for engineers to replace crystals with silicon based timing devices in the majority of applications and bring their designs into the 21st century.
Almost every piece of electronic equipment requires one or more timing sources, whether it is a consumer, industrial, medical computing, or telecom product. Estimates put the total value of the global market for timing devices at greater than $5bn US. This is a market that is currently almost completely dominated by quartz, but it seems, that this will not be the case for much longer.
But why are quartz timing solutions so limiting and why do they cause circuit designers challenges that they would rather not have to deal with? In order to understand the answers to these questions, we should consider the demands and trends in current and future electronic product design.
The first consideration is power; as the world tackles the challenge of reducing global demand for power whilst simultaneously embracing and developing renewable energy generation, there is also significant pressure on engineers working within the electronics industry to make their designs as power efficient as possible. CMOS oscillators consume around 75% less power than quartz crystal devices; a saving of this magnitude is good reason alone to switch approaches to timing.
For frequencies over 50 MHz – which accounts for most current and next generation high speed applications – an approach that uses a crystal often requires and additional dedicated circuitry to multiply the output frequency to the required level; extra circuitry means more power is consumed.
An additional reason for designers to adopt low power timing solutions is that more and more products nowadays are battery powered and designed to be portable. To a lesser degree, battery technology advancements have been somewhat stalled in a similar way to crystal-based timing solutions.
This means that increases in the operating time between recharges for portable battery powered products have mainly been achieved through the power efficiency of the electronic circuits contained within them, rather than through batteries that deliver greater capacity relative to their physical size.
Staying with the theme of portable electronics, by definition portable products are moved around and used in environments that are often uncontrolled or harsh. Away from a relatively cosseted ‘fixed’ environment, electronics equipment can be exposed to problems such as mechanical shock and vibration, sudden temperature changes, humidity and other factors.
CMOS oscillators that use established, well-proven fabrication processes and are housed in industry-standard sealed plastic packages, are far more resilient to such phenomenon compared to quartz crystals that are fragile – a problem accentuated when they are sliced very thinly to achieve the higher operating frequencies demanded by modern-day applications.
Another inevitable consequence of designing electronic products for portability or just compactness is that the space available for the electronic circuitry at the heart of the product has become more limited. This is compounded by the need to include greater levels of functionality in a reduced area, and items such as batteries and man-machine interface (MMI) features – for example capacative or resistive touch screens – that compete for the available space.
Whilst the potential to shrink the size of crystal-based oscillators has long since hit the limits of the technology, silicon oscillators are really only at the start of that journey of evolution. Furthermore, silicon timing devices can integrate various features and functions and are designed to operate as standalone devices; they do not have a requirement for additional, PCB real-estate consuming external circuitry.
In these early stages of gaining industry acceptance for crystal-free approaches to timing and encouraging designers to change their historical tendency to use quartz oscillators, it makes sense to offer silicon oscillators in pin compatible packages. This means that they can be offered almost as a ‘plug-and-play’ solution and be easily adopted in existing product designs as well as those in the initial stages of development. However, in the longer term there is plenty of scope to shrink package sizes to save even more space on the PCB.
The arrival of cost-effective, highly accurate silicon-based timing devices heralds a mini revolution and a new era for engineers designing almost every type of circuit. It should make their task easier and result in more reliable, robust and compact end products.
www.idt.com
