Beamforming techniques utilized either at the transmitter or the receiver terminals have achieved superior quality-of-service performances from both the multi-antenna wireless communications systems, communications intelligence and radar target detection perspectives. Despite the overwhelming advantages in ideal operating conditions, beamforming approaches have been shown to face substantial performance degradations due to unknown mutual coupling effects and miscalibrated array elements. As a promising solution, blind beamformers have been proposed as a class of receiver beamformers that do not require a reference signal to operate. In this paper, a novel gradient-based blind beamformer is introduced with the aim of mitigating the deteriorating effects of unknown mutual coupling or miscalibration effects. The proposed approach is shown to find the optimal weights in different antenna array configurations in the presence of several unknown imperfections (e.g., mutual coupling effects, miscalibration effects due to gain and phase variations, inaccurate antenna positions). By providing numerical results related to the proposed algorithm for different array configurations, and bench-marking with the other existing approaches, the proposed scheme has been shown to achieve superior performance in many aspects. Additionally, a measurement-based analysis has been included with validation purposes.