Elia:
You are implying some fairly accurate timing resolution for the sampling rate for the sine wave generator. For the moment lets ignore the low end of your frequency range and just look at the range from .1 Hz to 250 Hz at the 0.1 Hz step size.

If one presumes you use a 100 point resolution at this range then you are looking at a sample frequency range from 10 Hz to 25000 Hz. This seems quite reasonable to implement but you need to look carefully at the actual sample rate that can be achieved at the frequency steps you propose. The sample rate is generated via integer divides off whatever crystal that you have attached to the 80C320 processor. The timers on this processor can clock at a rate that is a divide by four of the oscillator. If one were to guess at a crystal frequency of 22.1184 MHz (a common choice to support accurate baud rates) then the counters can clock at 5.5296 MHz.

This excerpt from a spreadsheet will show that there will be errors in the actual frequencies that you will be able to achieve.




This link will access a ZIP file that contains the complete analysis in an EXCEL spreadsheet.

Michael Karas