harmonic responses. For m s 0.25 and a fortiori for lower values of m, good agreement can be noted between results obtained with the use of the HB technique and the perturbation method. Only a slight difference appears around the resonance frequency. As m increases, HB responses become less smooth. Secondary resonances appear near submultiples of f . The value of the resonance frequency f decreases with r r
the respect to f . The peak value at f is lower than that at r 0 r f , and the difference increases with m. r 0
The analysis presented here is a useful tool for predicting harmonic distortion. As for the perturbation analysis, various parameter values can be easily tested. The HB technique applies for both small-and large-signal modulation.
VI. CONCLUSION
This study presents IM harmonic distortion in semiconductor laser diodes. Results are obtained with the use of an equivalent circuit model derived from the rate equations and a harmonic balance technique implemented on a commercially available simulator. This technique is valid for small-and large-signal modulation. In the first case, good agreement is achieved in comparison with the perturbation method. The method used is interesting because electrical elements, including parasitic elements, whose models are available in MDS, can be easily added to the basic equivalent circuit in order to simulate the performance of more complete emitter modules.