Biomorphic robots have become a valuable tool in understanding neuronal pattern generation, motor control, interaction with environment, and behavior of animals. We propose that low-cost robotic devices may also be useful to teach those concepts. We developed a prototype of an educational tool in a form of a four-legged robot. The servomotors of the legs are controlled by a neuronal pattern generator composed of Integrate-and-Fire neurons connected via integrating synapses, all modeled by RC-circuits. We successfully delivered hands-on lessons for building neural networks on breadboards out of discrete electronic components and subsequently developed a programmable printed circuit board (PCB) that can serve as a useful adjunct to the discrete-components lesson plans. The board is mounted on the back of the robot and allows students to investigate the parameter space of bigger neural networks, and program complicated walking gait patterns. The next implementation of the PCB will bear sensors permitting the robot to interact with the environment. The new PCB also will allow probing of various components of neurons and synapses, mimicking electrophysiological recording with intracellular electrodes, and studying pattern generation. So far these tools have been tried at middle- and high-school level, and trials in college are also underway.