[time-nuts] Computing GPS Distance Error in Time

Mark Sims holrum at hotmail.com
Mon Nov 4 20:13:54 EST 2013


I was using the Adafruit GPS receivers to calculate the bearing and distance between two boxes that had radio tranceivers in them.  The control box polls the remote box position and does simple trig on the difference between the GPS coordinates.   I noticed that the Adafruit GPS seems to go into some kind of funky "position hold" mode if thinks that it is not moving (or is actually moving very slowly)...  the GPS coordinates stop updating until the box moves some distance.
I got much better results using a CN06 GPS module that is based upon the Ublox Neo-6M.  You need to solder a wire to the Neo module to access its PPS output.  HobbyKing sells them for cheap...  also lots of sellers on Ebay.   I could put one box on the ground, and move the other one around it only a couple of feet away.  It gave remarkably good results considering the boxes were much closer than the average GPS error...  it is basically a differential GPS application.  Also,  the processor is an AVR running single precision floats.
I laid out a breakout board that can take the Adafruit, CN06, or Trimble Resolution T GPS boards and convert the TTL/CMOS levels to true RS-232 levels via a MAX3232 chip to a DE-9 connector.  PPS polarity is jumperable. 		 	   		  


More information about the time-nuts mailing list