A feasible neuron model can be effective to estimate the mode transition in neural activities in a complex electromagnetic environment. When neurons are exposed to electromagnetic field, the continuous magnetization and polarization can generate nonlinear effect on the exchange and propagation of ions in the cell, and then the firing patterns can be regulated completely. The conductivity of ion channels can be affected by the temperature and the channel current is adjusted for regulating the excitability of neurons. In this paper, a phototube and a thermistor are used to the functions of neural circuit. The phototube is used to capture external illumination for energy injection, and a continuous signal source is obtained. The thermistor is used to percept the changes of temperature, and the channel current is changed to adjust the excitability of neuron. This functional neural circuit can encode the external heat (temperature) and illumination excitation, and the dynamics of neural activities is investigated in detail. The photocurrent generated in the phototube can be used as a signal source for the neural circuit, and the thermistor is used to estimate the conduction dependence on the temperature for neurons under heat effect. Bifurcation analysis and Hamilton energy are calculated to explore the mode selection. It is found that complete dynamical properties of biological neurons can be reproduced in spiking, bursting, and chaotic firing when the phototube is activated as voltage source. The functional neural circuit mainly presents spiking states when the photocurrent is handled as a stable current source. Gaussian white noise is imposed to detect the occurrence of coherence resonance. This neural circuit can provide possible guidance for investigating dynamics of neural networks and potential application in designing sensitive sensors. © 2020 Chinese Physical Society and IOP Publishing Ltd.