Coherence between elements of a laser diode array can be established either by coupling the individual oscillators together on-chip or by coupling the array as a whole to an external cavity. Several techniques for on-chip coupling have been demonstrated 1,2,3, however these approaches typically produce an output having low Strehl ratio due to the difficulty of maintaining adequate control of the phase and amplitude of the emitters across the length of the chip. Several authors have reported phase locking of small laser diode arrays by placing the diode array in an external cavity and making use of a spatial filter in a Fourier transform plane of the array to force mutual coherence between the diodes 4,5,6. This technique works well for relatively small arrays, but becomes impractical for large scale arrays. Recently an external cavity technique has been demonstrated that takes advantage of the Talbot self-imaging effect 7 to phase lock an array of laser oscillators 8,9,10,11,12. A significant advantage of the Talbot cavity as opposed to previous external cavity approaches is that the Talbot cavity can be scaled to accommodate arrays containing a large number of lasers if attention is paid to control of the oscillating modes of the array. In this paper we demonstrate how mode control of an array of diode lasers in a Talbot cavity can be achieved through the use of intracavity spatial filters or phase masks. The Talbot (or "Self-Imaging") cavity makes use of the diffractive properties of the Fresnel (near-field) zone of a periodic array of coherent sources, which has a set of characteristic self-image planes associated with it. At distances 2nD2/A from the array an image of the array is formed (D is the separation between elements in the source array and n is an integer). In addition to these planes, there also exists a set of planes at distances D2/mA, (m=1,2,3,...) from the source array in which multiple images of the source are formed, we refer to these planes as "subrimage planes." In the "m-th" sub-image plane, "m" separate images of the source array are formed, the different images having relative phase shifts between them. Figure 1 shows the locations of the first few image planes and the relative phases of the images in the planes.