This work describes a method to estimate the mean collapse and snapback voltages of CMUT elements and their statistical distribution by extracting the voltage-dependent capacitance from fast dynamic excitation of the device, enabling fast analysis of fabrication-related intra-element and inter-element variability. While both voltage excitation and current excitation of the CMUT element provide comparable array-level results in terms of mean and deviation of positive and negative collapse and snapback voltages, it is demonstrated that improved detection of isolated collapse and snapback events is achieved through current-driven excitation, implementing charge-controlled actuation of the membranes.

Estimation of the collapse (V_col) and snapback (V_sb) voltages of Capacitive Micromachined Ultrasonic Transducers (CMUTs) is usually performed by extracting C-V curves from low frequency impedance measurements at different bias points. However, impedance analysis in several bias conditions is time-consuming, making this technique unsuitable for wafer-level testing. Additionally, prolonged exposure to high electric fields may lead to charge injection and trapping phenomena in the CMUT in-cavity insulation layers. This paper proposes an adjustment to the conventional impedance analysis technique aimed at enhancing estimation accuracy and introduces a novel technique for fast C-V assessment enabling rapid wafer-level characterization. Results from both techniques are compared, demonstrating the validity of the proposed approaches.