From: Stephen D. <st...@da...> - 2002-06-01 08:13:52
|
On Fri, 31 May 2002, Bill Paxton wrote: > What I need to know is the highest software carrier > someone has transmitted? Are there any limitations on > the software clock or can it be taken up and over any > known CIR carrier freq? Can we "invent" our own > computer-to-computer IR protocols at 101kHz? Does > higher freq carrier cause more or less clock slip? > Should this be a private email to Christoph? :-) On CPUs with an rdtsc instruction then the carrier frequency can be generated rather accurately up to a decent fraction of the cpu clock. For sure 455kHz (B&O) will be no problem. Reckon a GHz processor could make something like 10MHz carrier. Be interesting to get someone with a 'scope to view the results and report back. CPUs without rdtsc time the pulse generation in microseconds. So max carrier frequency is much more limited - 500kHz is theoretically possible, but the "evenness" of the generated carrier will suffer for higher frequencies. 101kHz would have a 10% frequency jitter. Hope that helps, Steve Davies |