Oznog
Active Member
I need to put in a new section of fence on my property. I have my survey from when I bought the property. I had to do a piece of fence earlier and had the luxury of having it right beside the house and the survery already said how far the foundation was from the property line, ok that's great because it conversely allows me to determine how far the prop line is from the foundation.
Now this section I need to do is more than 100 ft from the foundation or the prop line pin. There are plenty of plants and obstructions, no line-of-sight. Even if I did have an LOS I don't really have the instruments to proper accuracy over this distanve. I called about a site survey again, they said over $300 and that it might be more with the obstructions and wanting these mid-points plotted not just corners. This is way too expensive.
So, here's my question. I have a Garmin GPS III+ and could probably get another consumer-priced model if need be. Is there any way to get sub-foot accuracy with "differential GPS"? Differential cancels out most errors, you could place one receiver at a known position and another receiver at the measured location and take the difference in readings. Or take just one GPS, zero it at a known location and quickly move the GPS to the new location (error usually takes awhile to change).
I saw where you can hook up a Starlink Differential Beacon receiver that get you 1-5 meters of accuracy, impressive but not enough for a fenceline. Those rely on a tower somewhere in the basic region broadcasting info on the current error for the general area. Hundreds of feet away from a known location should be much better.
I had to note that the Garmin serial data itself has longitude taken to 4 decimal places which is 430 inches at the equator (somewhat less than that in Texas here since a sec of longitude is less), even if I was able to achieve full accuracy (which would be a difficult task itself). I noticed that Wikipedia's entry for the NMEA format had an example to 5 decimal places, I saw an example from a SiRF receiver that went to 5 places, so perhaps I could find a unit with much better resolution than this Garmin. Even trying to average this may be tricky, the GPS only puts out one sample every sec so it'll take awhile to try to get an averaged reading with a high resolution for a single location.
If needed, maybe I could make a radio link between two receivers. Got plenty of microcontroller experience though the radio link is a bunch of extra hardware. Heck, at this distance maybe I could just find enough wire to make a wired connection? Esp if I could stab a stake in the ground on each side and run a single unshielded wire over there. NMEA is only 4800 baud serial anyways, not exactly challenging in the transmission dept.
Now this section I need to do is more than 100 ft from the foundation or the prop line pin. There are plenty of plants and obstructions, no line-of-sight. Even if I did have an LOS I don't really have the instruments to proper accuracy over this distanve. I called about a site survey again, they said over $300 and that it might be more with the obstructions and wanting these mid-points plotted not just corners. This is way too expensive.
So, here's my question. I have a Garmin GPS III+ and could probably get another consumer-priced model if need be. Is there any way to get sub-foot accuracy with "differential GPS"? Differential cancels out most errors, you could place one receiver at a known position and another receiver at the measured location and take the difference in readings. Or take just one GPS, zero it at a known location and quickly move the GPS to the new location (error usually takes awhile to change).
I saw where you can hook up a Starlink Differential Beacon receiver that get you 1-5 meters of accuracy, impressive but not enough for a fenceline. Those rely on a tower somewhere in the basic region broadcasting info on the current error for the general area. Hundreds of feet away from a known location should be much better.
I had to note that the Garmin serial data itself has longitude taken to 4 decimal places which is 430 inches at the equator (somewhat less than that in Texas here since a sec of longitude is less), even if I was able to achieve full accuracy (which would be a difficult task itself). I noticed that Wikipedia's entry for the NMEA format had an example to 5 decimal places, I saw an example from a SiRF receiver that went to 5 places, so perhaps I could find a unit with much better resolution than this Garmin. Even trying to average this may be tricky, the GPS only puts out one sample every sec so it'll take awhile to try to get an averaged reading with a high resolution for a single location.
If needed, maybe I could make a radio link between two receivers. Got plenty of microcontroller experience though the radio link is a bunch of extra hardware. Heck, at this distance maybe I could just find enough wire to make a wired connection? Esp if I could stab a stake in the ground on each side and run a single unshielded wire over there. NMEA is only 4800 baud serial anyways, not exactly challenging in the transmission dept.
