From: Travis T. <tr...@ta...> - 2009-01-23 19:24:06
|
> > Unlikely. 1/16384 gives a resulution of 61 us. It does not explain the big > difference. I'd say: check your crystal and your PIC program. > > Anyway, it's probably not the reason why irrecord does not detect the > protocol. I need to see the raw config file to check this. > I expected it would be fast enough as well. It appears to be about 3 ticks off in the worst case, not much I suppose, except that it adds up to about 180us. I'm using the hardware capture module to grab the times, if anything I would expect it to be missing some ticks, not getting extras. I'll keep checking on things. I suppose my oscillator could be running fast. On the PIC, I'm running a 32.768Khz crystal with caps on the timer 1 oscillator and using a 1:2 prescale to get the clock ticks for 1/16384 sec. Then I just set the capture module to capture the number of ticks on the rising or falling edge. Nothing really complex in the code. If I could convince the oscillator to keep running when connected to the logicport, I could actually check for speed error there and compare the # of pulses to the actual waveform from the IR receiver module. irrecord does detect the RC-6 protocol after the 2 lines of dots at this point. It fails to get any buttons to map, but I think that's because of the readdata() error I'm getting. I haven't had a chance to sit down and debug that yet. I'm not actually using a config file at this point, I just gave it a filename that didn't exist and let it try to create a config file. I did try using remotes/generic/RC-6.conf, and it gives me the same issue there. If you need any files or capture output, let me know and I'll post it up. Any help would be great. |