Accuracy / variation across brands / models of GPSr

[runner_one]

Don’t know if this question has been asked before or not.

Has anyone ever compared different brands / models of GPS units to see if error variation different / same across different units. What brings this up is there is one hider in my area that all of his caches seem to be about 40 feet due west in every case from where my unit zeros at. I don’t know what brand or model of GPSr this hider uses but the offset seems consistent. If it is one of his caches I can zero out and then look 40 feet west for the cache EVERY time.

Anyone have any comments, observations or experience with such an issue?

[NordicMan]

.. oh yeah this shouldn't cause another huge brand war thread..

 

[+StarBrand]

There are an awful lot of variables to consider here. I'll list just a few?

• brand/model of unit being used
  
• firmware loaded on device
  
• datum settings
  
• user error
  
• map errors
  
• multipath considerations
  
• etc
  

The only cacher I have ever known with that kind of consistent problem was using Google Earth to get thier coordinates instead of a GPS unit. Seems the maps were off by about 30 foot in the area he was hiding caches.

[+yabetterduk]

Big difference in accuracy between Garmin Map 60 and my Oregon 400t.

the 60 has always put me within 3' to 6'of a waypoint. Where the Oregon cant even bring me back to within 20' of any starting point.

I wish the accuracy of the Oregon had to do with one set direction - due West 40' - like you have found the other cachers gpsr has. The Oregon problem has to do alot with drift and its reliablity with or my trust in marked waypoints.

Maybe I should up the difficulty of my caches a star or two when I use my Oregon when placing them because it could be within 40' of the cache.

[+twolpert]

This kind of consistent error smacks of a datum issue. If other cachers report the same consistent error, it is likely that the owner is using a datum other than WGS84.

 

Almost any other source of error, such as multi-path effects, differences in the constellation (DOP and the resulting uncertainty in reported position) is going to be randomly distributed in both direction and distance across multiple caches.

[+fegan]

What brings this up is there is one hider in my area that all of his caches seem to be about 40 feet due west in every case from where my unit zeros at. I don’t know what brand or model of GPSr this hider uses but the offset seems consistent.

We have a hider about an hour away and he uses the exact same brand/model GPS that I have...most of hise caches are 20-25ft away from where my GPS guides me. So even same brand/model doesn't promise to have you standing on the geocache when your GPS reads 0 feet.

 

I think some people expect the GPS to put them on top of the cache. If I hide it with 15 foot accuracy and you find it with 15 foot accuracy...the cache could be 30 feet away. It's not a perfect science.

 

EDIT: fixed a typo (GSP instead of GPS)

Edited August 17, 2009 by fegan

[WanderWomen]

At our last event, one of our veteran cachers had us do an experiment regarding this issue. He gave us a set of coords and a survey flag marker. When we thought we had GZ, we marked the spot. There were all different GPSr's there; Delorme, all flavors of Garmin (including a Nuvi) and a few others I believe. We all hit in about a 10-15 foot circle. Not sure what that means except that there is no real "ground zero", maybe more like a "ground 15"

[+dakboy]

f he is consistently the same distance off in the same direction, I'd be looking into whether one of you is using the correct datum for recording the coordinates.

 

Do you have access to another GPS you can use to cross-check?

[+twolpert]

At our last event, one of our veteran cachers had us do an experiment regarding this issue. He gave us a set of coords and a survey flag marker. When we thought we had GZ, we marked the spot. There were all different GPSr's there; Delorme, all flavors of Garmin (including a Nuvi) and a few others I believe. We all hit in about a 10-15 foot circle. Not sure what that means except that there is no real "ground zero", maybe more like a "ground 15"

This is a graphic illustration of what Fegan said above. Also supports my observation that the error in distance and direction is random.

 

To elaborate a little, there are two sources of error involved in finding GZ. First, there was some error in the coordinates when the cache owner placed the cache. This is probably between 0 and 15 feet or so, but might be larger. Second, when you go looking for the cache, there is an entirely independent error in your reported position -- which is also in the 0 to 15 foot range most of the time. If the two error values happen to be large and in opposite directions, you could end up 30 feet or more from the cache. If both errors are small, or if they happen to be (and it would be entirely coincidence) about the same size and in about the same direction, you will end up on top of the cache.

 

This is why most cachers stop looking at the GPSr and start looking at the environment when they get within 30 feet or so of GZ

[+Xantos10]

And that's what makes this game so interesting. I would not want to step on top of the cache every time. I enjoy using my "geosense" to find what I am looking for. My GPSr just gets me in the proximity of the cache.

[+Okiebryan]

This spring, I set the point for the accuracy skills challenge at our Spring Fling. I went out 2 weeks ahead of time and buried an iron pin in a spot with ideal conditions. I drove it in the ground about 3" deep. I set up over it with a survey grade static GPS, L1 & L2 and Glonass receiver, and observed satellites for 2 hours. Then I post processed the data with OPUS. My results showed peak to peak of 3 cm, and ambiguities were 98% resolved. This is tight as a gnat's a**. Far tighter than the thousandth of a second (3 digits after the decimal) that we put into our handhelds.

 

The day of the event, about 70 people set a pinflag where they thought ground zero was. My handheld came within 3 feet. The best was 10" away and the worst was 11.7 feet.

 

That's with Delorme, Garmin, and Magellan. That was impressive to see this cloud of flags in an ellipsoid shape around the pin when I got out my pinfinder and dug it up.

[+Minimike2]

I would like to add a question to this.

 

Out in the clear, my eTrex Legend can give me a 5-7 feet circle. Under heavy leaf cover this opens up to 30-40 feet. I have WAAS enabled.

 

What difference do the newest SirfIII units have? That is, do they hold a smaller error under the trees? If so, how much?

[+NWArkies]

This spring, I set the point for the accuracy skills challenge at our Spring Fling. I went out 2 weeks ahead of time and buried an iron pin in a spot with ideal conditions. I drove it in the ground about 3" deep. I set up over it with a survey grade static GPS, L1 & L2 and Glonass receiver, and observed satellites for 2 hours. Then I post processed the data with OPUS. My results showed peak to peak of 3 cm, and ambiguities were 98% resolved. This is tight as a gnat's a**. Far tighter than the thousandth of a second (3 digits after the decimal) that we put into our handhelds.

 

The day of the event, about 70 people set a pinflag where they thought ground zero was. My handheld came within 3 feet. The best was 10" away and the worst was 11.7 feet.

 

That's with Delorme, Garmin, and Magellan. That was impressive to see this cloud of flags in an ellipsoid shape around the pin when I got out my pinfinder and dug it up.

 

Were any of the models PN40 or Oregon 400t? If so which performed best?

leaning toward PN40

[+dakboy]

This spring, I set the point for the accuracy skills challenge at our Spring Fling. I went out 2 weeks ahead of time and buried an iron pin in a spot with ideal conditions. I drove it in the ground about 3" deep. I set up over it with a survey grade static GPS, L1 & L2 and Glonass receiver, and observed satellites for 2 hours. Then I post processed the data with OPUS. My results showed peak to peak of 3 cm, and ambiguities were 98% resolved. This is tight as a gnat's a**. Far tighter than the thousandth of a second (3 digits after the decimal) that we put into our handhelds.

 

The day of the event, about 70 people set a pinflag where they thought ground zero was. My handheld came within 3 feet. The best was 10" away and the worst was 11.7 feet.

 

That's with Delorme, Garmin, and Magellan. That was impressive to see this cloud of flags in an ellipsoid shape around the pin when I got out my pinfinder and dug it up.

 

Were any of the models PN40 or Oregon 400t? If so which performed best?

leaning toward PN40

They both use the same Cartesio chipset, so it would come down to which version of the Cartesio firmware each was using. Not even the DeLorme & Garmin firmware, but the GPS chipset OEM firmware.