3D ultrasound (US) requires expensive transducers comprising thousands of elements and complicated hardware. This complexity originates from the classical idea on spatial sampling requirements for US imaging. The discovery of compressive sensing allows to ease this sampling constraint, enabling smarter ways of recording the required information. Inspired by this work we introduce a US imager that can perform 3D imaging using one acoustic sensor. Our device sends and detects US waves through a random coding mask that enables unique signals at every voxel. Rotation of the mask allows for several compressed measurements. By knowing the voxel signals, a full 3D reconstruction of the object can be obtained, as we demonstrate in this work.