Advances in design and development of light-weight and low power wearable and mobile devices open up the possibility of lifetime extension of these devices from ambient sources through energy harvesting devices as opposed to periodically recharge the batteries. The most commonly available ambient energy source for mobile devices is Kinetic energy harvesters (KEH). The major drawback of the energy harvesters is limited effectiveness of harvesting mechanism near a fixed resonant frequency. It is difficult to harvest a reliable amount of energy from every forms of device motions with different excitation frequencies. To overcome this drawback, in this paper we propose an adaptive electromagnetic energy harvester which utilises spring characteristics to adapt its resonant frequency to match the ambient excitation frequency. This paper presents a prototype design and analysis of an adaptive electromagnetic energy harvester both in simulation and real. The harvester has tested using a specially designed experimental setup and compared with numerical simulations. The proposed solution generates 3.5 times higher maximum power over the default power output and 2.4 times higher maximum frequency compared to a fixed resonant frequency electromagnetic energy harvester.