In ring laser gyros (RLGs), dynamic lock-in, which results from information loss in the lock-in region, occurs when a constant sine bias is introduced. However, sampling some signals in the lock-in region for a particular duration allows the retrieval of lost information. It is demonstrated how dynamic lock-in and the flat region in the input-output curve near the zero angle rate can be eliminated after compensation.OCIS codes: 140.3370, 140.3430, 230.5160. doi: 10.3788/COL201210.061403. Ring laser gyros (RLGs) have more advantages than traditional spinning-mass gyros. The operation principle of a RLG involves the Sagnac effect [1,2] . The laser cavity supports two independent waves traveling toward opposite directions. When an angle rate is introduced, the oscillation frequencies of the two beams differ. Thus, the angle rate can be measured by detecting the frequency difference. The RLG is, in fact, a ring laser whose gain media is He-Ne gas [3] . The lock-in phenomenon can significantly constrain the performance of the RLG [4−7] . Lock-in is caused by the mutual coupling between the opposite traveling waves [8−11] . With a low angle rate input, both waves lock to a common frequency, and the gyro becomes unresponsive to the input.The lock-in effect can be avoided by introducing a mechanical dither and placing the gyro far from the lock-in region most of the time [12] . The gyro that works in this manner is often called a "mechanically dithered ring laser gyro" [13] . However, if a pure sine wave dither is introduced, dynamic lock-in may occur. In other words, if the angle rate is below the dynamic lock-in threshold, the output also becomes zero [14] and a flat region near the zero input angle rate appears. Dynamic lock-in is the same as regular lock-in, except that the value is much smaller.Previous studies have indicated that dynamic lockin is mainly caused by information loss during lock-in traversal [15,16] . Information loss is related to the phase of the beat frequency signal and angle acceleration in the zero rate point [17] . This letter successfully compensates dynamic lock-in by sampling the signals in the zero rate point.The phase equation of a RLG can be written aswhere ψ is the phase of the beat frequency signal, Ω i is the input angle rate, and Ω L is the lock-in threshold, which must be positive.(1) has the following stationary solution:where the mean output is zero., the average frequency of ψ can be obtained. Changing the form of Eq. (1) yieldsTake integration of ψ over [0, 2π]; the mean period T can be obtained asThe average frequency of the beat frequency signal isIn Fig. 1, the response curve with the lock-in threshold is represented by the green line, whereas the ideal response curve without lock-in is denoted by the red line.When a sine dither is introduced, Eq. (1) can be written asψwhere Ω d is the peak dither angle rate and ω d is the dither frequency. Although the gyro works outside the lock-in region most of the time, it must traverse the lock-in region twice every dither cycle, ...