In many sensor protection applications, it is useful to selectively attenuate a strong coherent beam in the presence of a collinear, weak white light image. We call a device which can perform this function a "coherent beam excisor." One convenient approach for sensing coherence is to divide the input beam and recombine it in a pho torefractive material (see Figure la). The portion of the input beam which is coherent will write a refractive index grating, resulting in energy exchange between the two beams. Inparticular, the sample can be oriented so that a large fraction of the energy in the coherent beam is diverted away from the direction of the incoherent white light. Put differently, we wish to optimize the depletion of the signal beam, accompanied by strong amplification of the probe beam. Several recent studiesl.2 have shown that photorefractive two-wave mixing is a viable means of implementing a coherent beam excisor. The only disadvantage of this approach is that portions of the optical train are not coaxial; a beamsplitter and at least one turning mirror are required to form the auxiliary probe beam. One obvious simplification is to use a single beam (Figure 1b), with the role of the second beam now taken by scattered light inside the crystal. This beam fanning approach was actually the first photorefractive limiter reported in the literature.3 Recently, Salamo, et al.4 have proposed an intermediate solution consisting of a single inci dent beam, with a grating in contact with the entrance face of the crystal (see Figure le). The grating can be blazed so as to produce a single diffracted order, with a controlled diffraction efficiency. The diffracted beam can be looked upon as the probe beam in a two-wave mixing approach, or an auxiliary seed beam in the fan ning approach. Inany case, the diffracted beam will be amplified, leading to depletion of the signal beam. Until recently,5 no systematic study of excisor performance using a grating for probe generation has been undertaken. In particular, the effect of walkoff of the probe beam has not been considered. The ideal beam crossing posi tion is at the center of the sample, but the use of a separate grating places the effective beam cross ing position at the entrance face of the crystal. In this paper, we have simulated the operation of a grating by using a separately generated probe beam with a variable intensity, which can be set to cross the signal beam at a number of positions. We measured both the rejection and the activation energy. We will show that a factor of -10 improvement in activation energy can be obtained (compared to a pure fanning geometry), even when the beams cross at the entrance face.