Timing is everything

Advances in high-performance OCXOs

BY UWE SCHWEICKERT
R&D Manager
IQD
www.iqdfrequencyproducts.com

An accurate clock mechanism is a fundamental element of any modern system design either serving as a reference or for enabling synchronisation to be established. Technological progression in both wireless and wire-line communication sectors is placing huge pressure on manufacturers to produce timing devices capable of keeping pace with ever-higher performance benchmarks. This article looks at the challenges being faced and how more sophisticated devices are emerging as a result.

Fig. 1: Today’s OCXOs must offer higher frequency levels, tighter stabilities, improved phase noise and jitter characteristics, wider operational temperature ranges, andplus more- compact form factors.

In communication and broadcasting, use of a highly stable clock for either reference and synchronisation purposes is advised. This is normally taken care of by a precision crystal oscillator (XO) device — in most cases taking the form of an oven-controlled crystal oscillator (OCXO), typically with a frequency range from 10 to 40 MHz (see Fig. 1 ).

However, as communication infrastructure moves into the new IP-based era, with Long Term Evolution (LTE) mobile networks and 10/40-Gbit Ethernet optical lines being deployed, as well as high-definition (HD) broadcasting becoming increasingly commonplace, far larger quantities of data will be transferred. This will depend on implementation of complex modulation techniques beyond the scope of conventional OCXO technology.

Due to significant increases in the subscriber base and the bandwidth required per individual subscriber, more transfer channels will be needed. However, as the available frequency range for the different kinds of technologies is limited, tighter tolerances are imperative. With tighter tolerances the gap between the channels can be reduced, so the bandwidth for each channel can be expanded or with the same bandwidth, more channels can be squeezed frequency range.

The increase in data transfer rates also calls for a reduction in bit error rates. This means the stability of the clock source must be improved, so that the impact of jitter is reduced and phase noise performance can be lowered. Next-generation communication systems will need to specify higher-performance reference clocks. This can be achieved through a phase-locked loop (PLL), but this has the disadvantage that it simultaneously decreases system performance. So to maintain higher resolution, more advanced OCXO technology is now proving to be the more favored approach. Furthermore, the need to fully use all available board space is leading to greater use of compact, surface mount packaging. At the same time demands are being placed on devices to have more rugged construction, with wider operational temperature ranges.

Crystal stability

The key item for consideration when looking to ensure OCXO stability is the characteristics of the crystal at the heart of its construction (see Fig. 2 ). Crystal stability is defined by:

Fig. 2: The key to ensuring OCXO stability is in the characteristics of the crystal at the heart of the OCXO’s construction.

1. Aging stability: Typically a 10 MHz OCXO will see its stability impinged upon by around 50 ppb/year, with high-end OCXO devices only witnessing a deterioration of perhaps 20/30 ppb/year. This parameter is very important in relation to the overall system stability for a long period of operation.

2. Short-term stability: For periods of 1 s up to 100 s, short-term stability is of prime importance. To reach good short-term stability, a crystal with a high quality factor (Q-factor) is necessary. This depends on the crystal mode, frequency, package, and various other factors associated with its production. A third-overtone crystal reaches higher Q-factors compared with the fundamental mode at the same frequency. For a fifth overtone at the same frequency, the Q-factor is also better, but resistance levels will also increase. It is therefore very challenging to create low-frequency crystals in a fifth overtone. Also the crystal‘s high resistance can hamper the oscillator circuit’s ability to maintain stable oscillation under all operational conditions.

For high-performance OCXOs, an SC-cut (stress compensated-cut) crystal is usually specified. The oscillation-mode third overtone is preferred compared to a fundamental mode, thanks its higher stability in all cases because of the blank thickness. The blank thickness is inversely proportional to the frequency of the crystal, so a highly stable crystal should have thick blank (see Fig. 3 ).

Fig. 3: The blank thickness is inversely proportional to the frequency of the crystal, so a highly stable crystal should have thick blank.

At higher frequencies (above 50 MHz), fifth-overtone SC-cut crystals are the best choice for attaining high stability, due to the fact that, for third-overtone crystals, the Q-factor decreases and also the aging will get worse compared with a fifth overtone. So fifth-overtone crystals generally exhibit a greater degree of optimization for higher frequencies, but this needs innovative crystal design and fabrication to deliver crystals with really tight tolerances.

Use of higher overtones, like seventh or ninth, though theoretically possible, is very hard to realize as it is difficult to fabricate crystals for this. In addition, the oscillator design is very complicated because of the high resistance and low pullability of these crystals. Another important factor is temperature stability. For OCXOs, this is predominantly defined by the heating circuit and heating control of the oscillator circuit. For the crystal, it is very important to have a tight adjustment tolerance at the turnover point because the pullability of the fifth-overtone crystal is less than the third overtone.

Currently OCXO designs are mainly based on a third-overtone SC-cut crystal. To move to higher frequencies using a fifth-overtone SC-cut crystal means that certain, more forward-thinking manufacturers are looking to employ completely new circuit concepts so that the position where oscillation occurs can be moved. Also, because of the higher crystal resistance, the gain has to be improved to guarantee startup under all conditions. For such next generation designs the phase noise performance also has to be improved. IQD Frequency Products has set a goal to reach values near the carrier in the range of the company’s current 10-MHz IQOV-90-series and improve noise floor values (at offset frequencies around 100 Hz away from the carrier). In this frequency range, the phase noise is determined by the crystal, so a very good crystal with strong Q-factor will help to reach similar values. Far away the phase noise is determined by the power supply and output stage, so here filtering is needed. For the temperature stability the main issue is to have efficient thermal coupling between the crystal and heating circuit. For an OCXO the internal heating temperature must be 10 to 20°C higher than the maximum ambient temperature to guarantee a stable operation at high temperatures. This is due to the unregulated part of the power dissipation caused by the oscillator circuit, power supply and output stage.

It is clear that to cope with the exacting demands of next-generation communication systems, OCXOs need to evolve. They must offer higher frequency levels, tighter stabilities, improved phase noise and jitter characteristics, wider operational temperature ranges, and more compact form factors. ■