comment on this article

OCXOs aren't crumbling under pressure from other approaches

Playing a vital role in everything from mobile phones to military systems, crystal oscillators come in many forms depending on the properties and tolerance required. From basic clock oscillators to temperature compensated devices (tcxos), customers want cheaper products and higher levels of accuracy. Targeting more niche applications, oven controlled oscillators (ocxos) tend to be immune from many of these pressures, but there are still demands being placed on their functionality.

Andy Treble, sales director at Euroquartz, set the scene by describing how the trend amongst oscillator manufacturers in general is to build various forms of complexity into their devices and to then charge a premium. The basic rule of thumb, in his opinion, is the more accurate the oscillator, the more you can charge for it. "The quest for adding more into the oscillator is becoming the driving factor," he said.

While cost pressures may be driving down the price of tcxos, customers are still willing to pay for the performance of ocxos, according to Peter Davenport, managing director of HCD Research. "There is always a push for lower prices, however, you only specify an ocxo if you really need one," he said. This is because an ocxo is, by its nature, application specific. "In terms of volume, they're a niche market compared to other types of crystal oscillator," he added. "They are, if you like, the pinnacle of the triangle."

Indeed, ocxos are used in high end applications such as such as wireless base stations, test equipment and frequency synthesisers. They hold the crystal and oscillator circuitry at a constant temperature, keeping the output stable whatever the ambient temperature. Where a tcxo may offer an accuracy of 5ppm, an ocxo offers between five and ten parts per billion – 1000 times the performance.

Peter Sinclair, applications support manager, and Nick Amey, engineering manager, at IQD further described the capabilities of these devices: "OCXOs are used where you need the ultimate stability from an oscillator," said Sinclair. "They offer the ultimate performance that you can get from a quartz crystal. There is nothing better."

Plus, for the time being, they aren't going to be replaced with confidence by any other technology. According to Sinclair: "There is nothing on the horizon, at this moment, that can replace ocxos as the ultimate frequency source controlled by a crystal." Atomic standards may offer incredibly stable performance, but come at a much higher cost. "You're running into the realms of fantasy for commercial applications with those," he added.

Sinclair described how lower range oscillator products now tend be considered off the shelf items, while ocxos are invariably custom designed. This means a conversation between manufacturer and customer is vital to ensure the product meets the specification and is not overdesigned.

Depending on the requirements, properties such as phase noise or aging can be pushed to the limit, where there is space to deviate in others. Therefore, designing an ocxo is a partnership. "Other oscillators can be put down as commodity products," commented Amey, "but ocxos require a dialogue in order to be produced."

This flexibility is due to the component count. Where a tcxo may have around five components – a chip inside, a crystal on top, perhaps a couple of capacitors and resistors on the side – an ocxo will have closer to 100 pieces, so the bill of materials will be much more complex. "It's quite a complicated piece of design," noted Amey, "but because it's discrete components and analogue design, it's quite easy in some aspects to add more functionality in or take a functionality off the pcb and put it to an output pin."

For example, an oven alarm could be built in to indicate when the circuit has established the right temperature. Amey recalled that, perhaps 20 years ago, customers would have built their own monitor circuit, timing circuit, or alarm systems but over time, it's become easier for manufacturers to include these in the ocxo.

Sinclair and Amey suggest one reason for this is that many younger engineers are digitally trained and less experienced with the analogue nature of oscillator circuits. "Engineers haven't got the capabilities that they had years ago to do this design work themselves," said Sinclair. "So they have to buy a packaged product to realise what they're trying to achieve in their circuitry."

In this respect, an ocxo is an inclusive device that offers related functionalities in one package. "An ocxo won't be used in isolation," described Sinclair. "It'll need other frequency dependent components surrounding it to make it function as part of a wider picture." Rather than being an obscure oscillator, it's an inclusive one. If more features are being added to ocxos, it's a natural progression of the technology within, rather than a quest for more differentiation among ever cheaper devices.
So what are customers looking for in their ocxos?

Garry Thorp is a design engineer at Pascall, which specialises in custom low phase noise ocxos for high end applications such as radar systems, and for customers like BAE Systems and NIST. He said: "Phase noise has been the area we've addressed more than anything else," as exemplified by the Pascall 100MHz OCXOF, an ultra low noise vhf ocxo. The company is also looking for ways to further reduce phase noise and extend the frequency range of their ocxos by integrating low noise frequency multipliers. "This integrated approach can generally achieve better performance than a 'building block' approach involving separate oscillators and multipliers," he added.

According to Sinclair and Amey, their customers are also looking for lower phase noise because of bandwidth issues driven by the proliferation of mobile devices. "The more and more wireless devices you get, the more and more they can interfere with each other," observed Sinclair. "So you've got to control your frequency very accurately to ensure one particular device doesn't 'talk' to another device when it shouldn't."

Looking to the future, ocxo technology is likely to develop very slowly. This is because, when a device has been specified and approved for use, it generally remains there for the lifecycle of the product. "OXCOs don't change that quickly, they evolve with time," said Davenport, "I think it's more evolution, rather than revolution."

Adding more functionality to an oven controlled device is a matter of performance customisation and customer preference, rather than a need to stand out among increasingly cheap devices. Although there will always be power consumption and size pressures with all electronics, ocxos tend not to be affected because, by their nature, neither their current draw nor their size is critical for the application.

This means that future developments for ocxos may involve improvements in phase noise, jitter and aging, but not miniaturisation. "The laws of physics say you can only go to a certain size if you're looking for the ultimate performance," concluded Sinclair. "It might be the laws of physics are broken in the next few years by someone cleverer than us," he quipped, "but don't expect it to happen overnight."

Simon Fogg

Related Downloads

Comment on this article

This material is protected by MA Business copyright See Terms and Conditions. One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.

What you think about this article:

Add your comments


Your comments/feedback may be edited prior to publishing. Not all entries will be published.
Please view our Terms and Conditions before leaving a comment.

Related Articles