I attached the VI with a few modifications I made, the derivative of the unflitered squarewave, and the modulated output. I included no intergration additions in the modifications I made.
The first modulated region works as expected, but subsequent regions have an increased frequency. Since the frequency is constant with the exception of the derivative contribution, it leads me to believe there is an accumulation of derivative values. The peak of the derivative wave form is at 2500 and the valley is around -5000, leaving a net of approximately -2500 per cycle.
