400t Questions with Google Earth

[+rjmlakota]

Here is an image that is have from this weekend. I was wondering if some could explain my questions on the image.

 

Thanks - RJ

 

 

[+g-o-cashers]

Have you upgraded to GPS Firmware version 2.8? The points to the east and west look like they might be the location error issue or they might just be normal drift. Search for the Colorado Location error thread to read more about these problems.

 

The line to the south is a track segment that probably connects this track the previous place you had the Colorado powered on. The Colorado doesn't do a great job of terminating track segments when you power up in a new location and you are still navigating. What you'll see is one long line from the place where you powered off to where you powered back on. Could that explain this line?

 

GO$Rs

[+rjmlakota]

Yup, everything is up to date. I just spoke with Garmin and they said the "drift" error is do to the high sensitivity receiver? Does that sound right??

 

I'm getting very annoyed!

 

Thanks,

 

RJ

[segler999]

No. I sometimes get the same weird points with my "Ole-Blue" Legend, that has a low-sensitive receiver.

 

I think it is an artifact of some numerical error that occurs just at the instant that the fix is taken by the receiver. You know, an interruption in a register read or write. The error gets propagated through the position calculation. This can occur with any kind of receiver. It is a function of the software and/or electronics design. Just a theory.

 

Good software can be designed to detect these things and stop the calculation until these things are completed. One hopes that Garmin has programmed such safeguards into their software.

 

There is one pattern I notice, however. The weird points always occur near land but never occur out on open water (I use Ole-Blue for windsurfing). So, they could really be the result of reflections, which can occur with any receiver, sensitive or otherwise. Also, the points always occur only at the beginning of a session, never later. Well, the beginning of my sessions are always on land, doh!

 

You can mitigate this by turning on your gpsr and letting it fix and stabilize under a clear sky for many minutes before embarking on your trip/hike/geocache/voyage. I always look for mine to show at least 6 good solid satellites before I reset the trip computer and start off.

[+twolpert]

I don't know what the scale of the image is. However, some features in the image could be characteristic of high-sensitivity receivers in multi-path conditions. In particular, momentary "spikes" in your reported location are typical. The spikes are characterized by very fast (almost instantaneous) changes in reported position which almost immediately revert to (very near) the original location. It is not unusual for such spikes to cover 50 to 75 feet. I suppose larger ones are possible, but I can't say I've seen such myself. Another characteristic (not apparent in the image) is a sort of spider-web track surrounding your true location when you are standing still.

 

The explanation here is that in challenging positions, the receiver "sees" the same satellite more than once because the signal reflects off the foliage canopy, a nearby hill, a nearby building, parts of your car, or whatever. The unit periodically re-evaluates the relative quality of all the signals from the same satellite and picks what it believes to the be the "real" signal. Each signal travels a different distance to get to the receiver, so changes in the choice of signal result in very fast changes in your reported position. This is normal for a high-sensitivity receiver. Such behavior may also be exhibited by older, lower-sensitivity receivers but it is much less noticeable because the older receivers simply do not see nearly as many low-power reflected signals.

 

This behavior is different than the "drift" discussed in the forums. The behavior characteristic of the drift problem is a gradual divergence from your true position as you move along. The accumulated error can be sizable -- 600 or 700 feet in the space of half a mile or so. This is usually accompanied by higher than normal EPE (on the order of 70 to 100 feet). With the firmware that originally exhibited the problem, the error took a long time (20 minutes or more, if ever) to correct itself when returning to good reception conditions. On the other hand, power-cycling the unit resulted in a correct fix immediately. The jury is still out on newer versions of the firmware. People (like g-o-cashers) who saw the problem frequently on earlier releases see it less often now. And it appears to correct itself within a much shorter timespan.

 

The concensus in the forums seems to be that the problem is caused by bad "position smoothing" in the chipset firmware. This results in a gradually accumulating error which the unit does not recognize. In a general way, this probably is related to attempts to correct some of the characteristic multi-path behaviors. But it's not an inevitable consequence of high sensitivity. The drift is much worse than the typical multi-path errors because it results in a large, persistent error in position which is generally difficult for the user to detect.

[+rjmlakota]

Thanks so much for the detailed explanation! I guess that's why I don't see this type of errors with my GPSv. I really have not seen the drift then. There are times that my have an EPE of 43-50'. But, it normally comes back to the norm with in a few seconds.

 

RJ