Injection locking

Injection locking and injection pulling are the frequency effects that can occur when a

Injection locking has been used in beneficial and clever ways in the design of early

Injection from grandfather clocks to lasers

Injection pulling and injection locking can be observed in numerous physical systems where pairs of oscillators are coupled together. Perhaps the first to document these effects was

In a modern-day voltage-controlled oscillator an injection-locking signal may override its low-frequency control voltage, resulting in loss of control. When intentionally employed, injection locking provides a means to significantly reduce power consumption and possibly reduce phase noise in comparison to other frequency synthesizer and PLL design techniques. In similar fashion, the frequency output of large lasers can be purified by injection locking them with high accuracy reference lasers (see injection seeder).

Injection-locked oscillator

An injection-locked oscillator (ILO) is usually based on cross-coupled LC oscillator. It has been employed for frequency division^[2] or jitter reduction in PLL, with the input of pure sinusoidal waveform. It was employed in continuous mode clock and data recovery (CDR) or clock recovery to perform clock restoration from the aid of either preceding pulse generation circuit to convert non-return-to-zero (NRZ) data to pseudo-return-to-zero (PRZ) format^[3] or nonideal retiming circuit residing at the transmitter side to couple the clock signal into the data.^[4] In the late 2000s, the ILO was employed for a burst-mode clock-recovery scheme.^[5]

The ability to injection-lock is an inherent property of all oscillators (electronic or otherwise). This capability can be fundamentally understood as the combined effect of the oscillator's periodicity with its autonomy. Specifically, consider a periodic injection (i.e., external disturbance) that advances or lags the oscillator's phase by some phase shift every oscillation cycle. Due to the oscillator's periodicity, this phase shift will be the same from cycle to cycle if the oscillator is injection-locked. Moreover, due to the oscillator's autonomy, each phase shift persists indefinitely. Combining these two effects produces a fixed phase shift per oscillation cycle, which results in a constant frequency shift over time. If the resultant, shifted oscillation frequency matches the injection frequency, the oscillator is said to be injection-locked. However, if the maximum frequency shift that the oscillator can experience due to the injection is not enough to cause the oscillation and injection frequencies to coincide (i.e., the injection frequency lies outside the lock range), the oscillator can only be injection pulled (see Injection pulling).^[6]

Unwanted injection locking

High-speed logic signals and their harmonics are potential threats to an oscillator. The leakage of these and other high frequency signals into an oscillator through a substrate concomitant with an unintended lock is unwanted injection locking.

Gain by injection locking

Injection locking can also provide a means of gain at a low power cost in certain applications.

Injection pulling

"Injection Locking and Pulling"

Injection pulling and locking heard alternatively when one oscillator of two sweeps in frequency

---

Problems playing this file? See media help.

Injection (aka frequency) pulling occurs when an interfering frequency source disturbs an oscillator but is unable to injection lock it. The frequency of the oscillator is pulled towards the frequency source as can be seen in the spectrogram. The failure to lock may be due to insufficient coupling, or because the injection source frequency lies outside the locking window (also known as the lock range) of the oscillator. Injection pulling fundamentally corrupts the inherent periodicity of an oscillator.

Entrainment

Entrainment has been used to refer to the process of mode locking of coupled driven oscillators, which is the process whereby two interacting oscillating systems, which have different periods when they function independently, assume a common period. The two oscillators may fall into synchrony, but other phase relationships are also possible. The system with the greater frequency slows down, and the other speeds up.

Dutch physicist

A 2002 study of Huygens' observations show that an antiphase stable oscillation was somewhat fortuitous, and that there are other possible stable solutions, including a "death state" where a clock stops running, depending on the strength of the coupling between the clocks.[7]

Mode locking between driven oscillators can be easily demonstrated using mechanical metronomes on a common, easily movable surface.^[8][9][10] Such mode locking is important for many biological systems including the proper operation of pacemakers.^[11]

The use of the word entrainment in the modern physics literature most often refers to the movement of one fluid, or collection of particulates, by another (see Entrainment (hydrodynamics)). The use of the word to refer to mode locking of non-linear coupled oscillators appears mostly after about 1980, and remains relatively rare in comparison.

A similar coupling phenomenon was characterized in hearing aids when the adaptive feedback cancellation is used. This chaotic artifact (entrainment) is observed when correlated input signals are presented to an adaptive feedback canceller.

In recent years, aperiodic entrainment has been identified as an alternative form of entrainment that is of interest in biological rhythms.^[12][13][14]

See also

• Injection-locked frequency divider
• Phase-locked loop
• LC oscillator
• Electronic oscillator
• Burst mode clock and data recovery
• Entrainment (hydrodynamics)
• Brainwave synchronization
• Synchronization of chaos
• Phase-locked loop range
• Tidal locking

References