Model Based Drive Signal Optimization Of Cmuts In Non-Collapse Operation And Its Experimental Validation

A model based transmit optimization method for Capacitive Micromachined Ultrasonic Transducer (CMUT) arrays operated in conventional imaging mode is presented. For the large signal analysis we used the transient CMUT array model presented in [1]. As an example, a 15 MHz, 60% FBW CMUT array element with 16 membranes and 70 nm gap is modeled, fabricated and tested. Through, fast iterative simulations, the transmit behavior of the transducer for the full range of different drive signals is explored as a function of DC bias, pulse width, and amplitude for unipolar and bipolar pulsed operation. It is shown that for a CMUT element operating in the non-collapse mode, the absolute maximum possible pressure output is generated when the transducer is driven with no DC bias and a very short unipolar pulse that results in full gap swing. By adjusting pulse width and amplitude, full gap swing can be achieved with small peak pulse voltages, maximizing the transmit sensitivity in Pa/V. In all cases, maximum achievable pressure is about 2 dB of the absolute maximum. The case where the TX and RX elements are the same is also investigated where the 95% of the collapse voltage is applied to the transmitting element for optimal receive sensitivity. In this case, a bipolar pulse is applied. The first part of the pulse adds to the DC bias force to pull the membrane down near the full gap and the second part effectively removes DC bias, releasing the membrane to achieve largest swing past its rest position. With this strategy, absolute maximum pressure can also be achieved with large DC bias levels.