Solid State Attitude and Heading Reference Systems (AHRS) are ubiquitous in small unmanned aerial vehicles (UAVs). Calibration of these systems is challenging, particularly as the aircraft and sensors become smaller and harder to assemble with precision. Parameter instabilities due to component aging, impact and air frame magnetic field variations were a primary motivating factor for developing a field calibration technique. We present an optimisation algorithm and calibration methodology and its results. The algorithm simultaneously solves for biases and scale factors in magnetic and acceleration sensors. No ancillary equipment is required to obtain the dataset used for calibration which differentiates the system from factory calibration methods used for commercial AHRS. It also computes the misalignment between the coordinate systems of sensor triplets. Soft iron effects and non-orthogonality in the coordinate systems of sensor triplets are shown to be a challenge for statistics based sensor calibration techniques. Refereed/Peer-reviewed