Adaptive control is a control method that has an adaptation mechanism that reacts to model uncertainties. The control method is used to realized synchronization of a new chaotic system in a unidirectional master-slave topology. The master chaotic system and the slave system are adopted as transmitter and receiver, respectively for the purpose of secure communications. Both analog and digital designs are realized. The digital system is a cycle and bit accurate design having a system rate of $450MHz$. The design is targeted at an Artix-7 Nexys 4 FPGA. The transmitters are accordingly modulated by analog signals and fixed-point signals of different resolutions, sampling and frequencies. Although the adaptive controller tends to react on introduction of a modulating signal, we show that with a special detection mechanism or filtering including exponential smoothing, the choice of which depends on the nature of the modulating signal, it is possible to recover the modulating signal at the receiver. Moreover, chaos-based spectrum-spectrum communication systems are realized based on adaptive synchronization of the new chaotic system. Furthermore, in order to ascertain the robustness of the adaptive controller, the modulated signal is transmitted via an awgn channel and the probability of error and bit-error-rate (BER) computed by sweeping across sets of SNR and noise power values in a Monte Carlo simulation. It turns out that the probabilities of error or error rates are reasonably low for effective communication through mediums with certain noise conditions. Hence, the adaptive controlled communication system can be robust.