??? 05/31/06 02:52 Read: times |
#117327 - Of course, but... Responding to: ???'s previous message |
Thomas said:
I'd be violating the companies wishes to say anything more than that. Of course, I fully understand. But without more facts I cannot successfully think about a strategy to solve your problem... My first idea was to use a shift register, in which the micro first stores the whole pulse train, parallely or serially, and which is read-out afterwards at a high speed shift rate. If the shift rate must not be varried continuously, but only in a few steps, then the programming of read-out clock speed would not present a bigger problem. With a sufficient high read-out shift rate, though, and a sufficient high depth of shift register (number of bits) even a constant read-out shift rate could do the job: You could store "000000111111000000111111" or "00000111110000011111", as an example, and by this simulate slightly different periods of your square wave signal. I think you get what I mean? Kai |