On Sun, Jan 19, 2014 at 6:32 AM, Charles Steinmetz <[email protected]>wrote:
> > The question I haven't seen answered is what error band is acceptable to > the OP. Mark has posted that it is not terribly difficult to get within > small fractional minutes if you start with GPS time and position data, as > he has done for a future release of LH. But getting to the millisecond > level or better is likely much more difficult. What would you do with millisecond level timing? One can only draw to a computer monitor about 60 to 120 times per second at most. If the computer were driving a costom build LED numeric display it could write milliseconds but the human eye can not detect chances about about 100Hz Don't worry about fitting this inside a PIC, modern uPs have 32K of storage, can be programmed in C++, have floating point math libraries and cost $5 shipped for one mounted to a PCB with supporting electronics. Pro-Mini-atmega328-5V-16M<http://www.ebay.com/itm/New-Pro-Mini-atmega328-5V-16M-Replace-ATmega128-Arduino-Compatible-Nano-/200957063666?pt=LH_DefaultDomain_0&hash=item2ec9f971f2> I'm using something like this for robot inverse kinematics, basically some linear algebra using 4x4 matrixes and it does it in real time with 3/4 of the memory unused. It is pretty easy to connect one of these to a small cell phone sized graphic screen. -- Chris Albertson Redondo Beach, California _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
