You'll get some drift as different satellites come into or out of "view". I did some testing of the timebase accuracy (which affects everything else) and it drifts and changes over time, can trend one way for half and hour then trend another way for two hours etc.
10m sounds a bit much but it depends on which module and it's spec and how many sats you had. It also depends how good the thing you are comparing it to is.
What Georacer is seeing as "drift" over an hour could be in some part from an accumulation of stored averages, which SHOULD get more accurate over an hour of being run. If that is causing it to get less accurate as the averages accumulate it might mean the datum is wrong. Normally I'd run it for a couple hours and see if it stabilises, then take that as a "good" measuement.
10m sounds a bit much but it depends on which module and it's spec and how many sats you had. It also depends how good the thing you are comparing it to is.
Yeah that all sounds logical and is probably the way I would expect it to work too. But you don't know exactly how they sequence the math in processing the coords or what averages etc they store in flash, those averages help a lot to give better accuracy than can be gained from a few minutes of sat signals.WBahn said:
What Georacer is seeing as "drift" over an hour could be in some part from an accumulation of stored averages, which SHOULD get more accurate over an hour of being run. If that is causing it to get less accurate as the averages accumulate it might mean the datum is wrong. Normally I'd run it for a couple hours and see if it stabilises, then take that as a "good" measuement.
