Coded excitation for low SNR pulse-echo systems: Enabling quasi-real-time low-power EMATs

There are many scenarios in industrial ultrasound where the signals lie below the noise threshold, i.e. SNR <; 0 dB. Examples include electromagnetic-acoustic transducers (EMATs). EMATs are contactless and can penetrate through non-conductive coating layers requiring minimum surface preparation. However, they are also inefficient and hence power inputs in the order of a few kW are needed. They would be ideal for wall thickness monitoring in oil and gas installations but intrinsic safety regulations require low power inputs so they have been disregarded to date. A potential solution is to use averaging over many repetitions but this takes a long time, and any changes, such as temperature drifts, ultimately limit their performance. An alternative is to use coded sequences, they require less acquisition time but in the past their length has been limited by the time it takes the wave to travel to the closest reflector and back. The presented approach removes this limitation by introducing receive intervals into long coded sequences, so that reception can take place intermittently and on average signals from nearby reflectors are not masked. The optimal synthesis of these sequences is discussed; sequences that have the same number of receive and transmit intervals of equal length are found to perform optimally and are an order of magnitude faster than averaging when SNR <; 0 dB. As a result, EMATs can be driven with 4.5 Vpp (less than 0.4 W) obtaining a clear signal in quasi-real-time (e.g. ~9 Hz refresh rate) using the proposed sequences.