Knowledge of the fine-grained location and orientation of devices on a surface can be used to enhance the surface-based computing tasks of mobile users in the home and workplace. However, existing systems which provide surface-based positioning information are often not a practical solution for mobile users, since the systems all rely upon pre-installed and calibrated environmental infrastructure. In this paper, we present prototype positioning devices for surfaces which do not rely on such infrastructure. We show that inexpensive infrared transducers can be used to effectively sense relative location and orientation of surface devices. We evaluate the novel approach of using intensity of light pulses for fine-grained location measurements.