Why would 2 garmins show the same position but be 25' apart?

[romulox]

This may start out reading like the current "legend cx" accuracy thread, but that's not my intention. This is a general shout out asking for help or advice on "why this is". I'm new to the world of GPSr and would appreciate any clarification.

 

The story starts out with my buddy and I going caching (and this thread has nothing to do with caching).

 

After turning on our units which are both using WAAS, allowing them to sit for about 15 minutes, and then sitting them next to each other, their positions were as follows:

 

Legend cx

==========

N XX°XX.412'

W 0XX°XX.215'

 

60 cx

==========

N XX°XX.412'

W 0XX°XX.214'

 

From doing some math from information found on the internet and realizing I have longitudinal differences around central Florida vs the equator, a difference in .001 minutes of longitude should be about 4 to 6 feet.

 

This seems normal to me and is not the question at all - just laying the groundwork.

 

My buddy and I went caching. The cache, and us, were in a large field - open sky and no trees. The cache was located at the following coordinates.

 

N XX°XX.353'

W 0XX°XX.256'

 

My friend with the 60cx arrived about 9' from the cache. I, with the legend, arrived about 20' from the cache (he's historically closer to all caches we find as referenced in the "Legend cx accuracy" thread on this board). Both of us, walking very slowly, had reached "0 feet to destination", with respect to the cache, according to the GPSr. Standing where we were, we called out our positions. Both of us were EXACTLY on the cache coordinates:

 

N XX°XX.353'

W 0XX°XX.256'

 

... yet, we were standing about 20-25 feet apart. I figured we should be standing somewhere within a few feet of each other since our positions were reported to be identical.

 

 

Finding that this was odd since initially our units were just .001' in difference, I walked towards my friend. Standing next to him my GPSr read:

 

N XX°XX.353'

W 0XX°XX.260'

 

This difference in .007' would account for us standing 25 feet apart.

 

We decided to turn off our units and walk back to work (we were goofing off obviously). A few hours later, we revisited the same cache - this time turning on our units and going as soon as we got a signal - versus letting them sit and acquire data for 15 minutes. We arrived at the cache just like before: standing 20' apart yet reportedly on the same position.

 

As a 3rd test, which is admittedly not fair since it was like setting a cache, my friend walked 200' away, marked a waypoint, gave me the position, and I walked there. When I reached the designated coordinates, I was again 25 feet away from his target (which is the stick in this picture).

 

 

My question is, with 3 tests, why would a legend cx and a 60cx be physically so far apart when our GPSrs are are displaying the exact same positional data? Is this just how GPSrs are? I mean, I could understand maybe Garmin vs. Any_Other_Brand, but Garmin vs Garmin? Is the hardware that different between the two?

 

Seriously, I'm not trying to start another accuracy thread, I'm more interested in learning why this might be. To me it doesn't make sense, but I've only had a GPSr for about 3 months so I'm just trying to understand it all. My current working theory is that the legend's processor is much slower than the one in the 60cx and if I had just stood still for a few minutes, my positional data would have drifted more towards the 60cx's.

 

Thanks!

[+Fhantazm]

Something very similiar happend to a friend and myself tonight and both of us have 60csx's, except we were over 200feet apart showing the same coords. Once we did a re-find all was ok and both were matched up and showing the same location. To answer your question, Im not exactly sure why these things happen sometimes. The only thing I can think of in your case, would be the issues that are occuring in the Legend/Venture Cx series right now. I would also be concerned about that interesting yellow thing that has taken over your head and face.

[Alphawolf]

well...If you had a 12 foot margin of error, and he had the same, you could be compounded to 25'.

[robertlipe]

Alphawolf has the right idea.

 

The fabled "10meter" accuracy (involving perfect sat geometry, a clear view of the sky, the sacrifice of a virgin goat, etc...) is additive. Your GPS decided your were at a set of coords, give or take a bit. Your buddy's GPS did the same.

 

This is why consumer grade GPS (technique and equipment) isn't used for property or political boundary determination.

[+budd-rdc]

This is interesting, especially since the "error" seems consistent. I would have expected the distance to be more random.

 

I've seen different GPSr's that would be about 0.003-0.004 minutes apart but after waiting for them to settle, they'd be within 0.001 minutes, and often match. This even happened under light tree cover at an event, between Magellan eXplorist XL, Garmin GPSMap 60csx, eTrex Summit, eXplorist 500, and a Magellan Sportrak Color. Fizzymagic, forman, and Marky can vouch for this one.

 

The 60cx uses a SiRF-Star III chip, but the Legend Cx doesn't. Not sure if that's the answer, but maybe others can try something similar with the same pair of GPSr's and post their observations here.

 

If you are into benchmarks, you can look for them with the text "location is ADJUSTED" in the description and compare the coordinates against your GPSr's.

[Hertzog]

I wonder if the fence in the background might be generating reflections that effect the two units differently? I'm not sure how good a reflector a mesh fence would be at GPS wavelengths, but there's a good chance someone monitoring this thread would have the expertese to know. As a test you could repeat the experiment and shield your GPSrs from that direction with your bodies (good excuse to take another break from work).

[+apersson850]

I dare say there's nothing suspicious in this at all.

A GPS is like a test instrument. There are inherent inaccuracies in many links in the chain, which eventually leads to the position the GPS shows to the user.

 

In this case, where two units both show the same point (that's as expected, since you were trying to do a nulling operation towards the same target), but end up in different places, the difference can have many combined causes. To mention a few:

 

The number of satellites used for the computation may be different. Due to the collected data, where the two units may have some differences, more or fewer of the available satellites may be included in the PVT computation. As a consecuence, the DOP figures may be different for the two units.

 

Correction tables for ionospheric interference may be different, due to more or less complete reception, in the two units.

 

Correction tables for internal frequency deviations vs. temperature may be less optimal in one of the units.

 

Due to the randomlike nature of correlation of GPS signals, one of the units may be less optimally correlated to one or more satellites. Conditions are never exactly the same, even with two units at almost the same position. No carrier phase, but only code phase correlation is done in units like these.

 

This is just how exact it is, or rather is not.

Talking about accuracy, the figure given by the unit is not to be taken literally. There's a certain confidence in being within the distance given, but it's never 100%. Also, due to the circular nature of these values, they are added quadratically. The combined accuracy is the square root of the sum of the squares of the values given. Hence, for two units, both reporting 10 meters, it combines to 14 meters.

But it's still something like 50% chance that you actually are outside of these 14 meters as well. The exact percentage is up to the manufacturer.

[romulox]

Thanks for the replies. All of the information makes sense. I never thought about the units picking up different satellites either. I remember my buddy saying for WAAS he was showing sat #50-something and I was using #35 (I think). But at the same time, we've been at this cache about 7 times and we're both stading in our same spots 20' away from each other.

 

I think I'm just going to take this as "it is what it is". I know these consumer units are inherently inaccurate. I was thinking maybe the 2 units were using different math or something to calculate position. But the truth is, there seem to be WAY too many variables to determine the hows and whys of a particular unit.

 

My buddy and I are going to do more testing today (all dependant on our workload!). If I come up with more results, I'll let you know. We're going survey marker hunting, be we need to find a modern marker.

 

Thanks to all

-Romulox

Edited August 23, 2006 by romulox

[+Sputnik 57]

Another way to thing about this:

 

Your GPSr knows its exact distance from several satellites, but not its direction from any of them. Therefore your GPSr is solving a math problem trying to find the intersection of 4 (or more) spheres. Due to atmospheric conditions and other anomolies, as it turns out, the spheres don't all intersect in a single point. When you zoom in really close, you find that the spheres get really really close to each other, but a couple don't quite touch or a few of them overlap. The GPS says, "Well, I'm not sure PRECISELY where I am, but I'm somewhere in the little gap between the intersection of the spheres.

 

Your buddy's GPS is doing the same thing, but since his is using a SiRF III processor, his GPS is playing "what if" games a million times a second--"what if the signal from sat 14 is being bent a little more by the atmosphere than the WAAS satellite thinks?" "what if the sat 24 signal is being bounced off of a nearby object?" etc. As a result, it can often give a better guess about where in the "sphere gap" he really is. In addition, as noted above, his unit may be measuring from different satellites than yours, or may be giving more weight to some of the satellites than your.

 

Or another analogy. You pull out a paper map and draw and X on it with a pencil. You buddy does the same thing. You both go to the spot on the map marked with the X yet you are 20 feet apart. You wonder why, so you look at the map. You have to zoom in really really really close to see exactly where you are, and sure enough, you are standing on the graphite mark made by the pencil, but the mark is about 15 feet wide!! You could be anywhere on that mark and still be "at the X". You buddy is also on his graphite mark, which is maybe 10 feet wide. The marks even overlap, but you just don't have enough precision to be on top of each other.

 

The remarkable thing (to me) is that you repeated the test and got the same amount of separation in the same direction. Consumer GPS units are not ususally that consistent.

[+Red90]

Personally, I think there is something wrong. In my experience, in an open area like that you should ne much closer. Also the "error" tends to mainly be atmospheric and time of day errors (not random). When caching in groups almost always everyone zeros out within a 5 foot circle with a clear view of the sky like shown in those photos.

[viamundi]

Do both units in test described above use the same map datum like WGS 84? If they don't that may case this problem.

[+Sputnik 57]

Personally, I think there is something wrong. In my experience, in an open area like that you should ne much closer. Also the "error" tends to mainly be atmospheric and time of day errors (not random). When caching in groups almost always everyone zeros out within a 5 foot circle with a clear view of the sky like shown in those photos.

I hate to disagree, but I saw a photo in the main forum a year or so ago taken at an event. The organizers took a bearing off of a benchmark to fix coordinated in a field. Everyone got a small red flag and were asked to find ground zero. There must have been 30 flags in the field, fairly randomly arranged, in about a 50 foot circle. Closest to ground zero got a prize. As I recall, he or she was about 2 feet off. If your group is always within a 5 foot circle, I would call that extraordinary!

[obfuscater]

There are several major factors affecting coordinate accuracy not mentioned here.

1) Position of satelites -- If satelites are all in a row and not spread out across the sky you will find accuracy much harder to arrive at.

2) Check out the computed accuracy on your GPS in most cases they say 25' to 39' or something like that they are not giving you 10' accuracy. Different models compute to a different level of accuracy.

3) In a clear sky with lots (more than 5) satelite accuracy is much better. You will achieve greater accuracy say down to 14'. But for consumer GPSs that is as good as it gets.

4) Satelite drift -- Satelite information drifts a bit for each satelite. This information is computed and fed over RF. This veries constantly (figure daily). When placing a new site, you should take readings at your hide for at least 4 times spread a couple days a part. This will give you a very good ability to give good co-ordinates about your hide. Other's won't say, " I found that 20' the other side of the trail". :-)

 

I have a pro SRS that I use to place my finds, it has all kinds of radio guided drift information with a sub 1 meter accuracy. With tri-cording I can get surveying accuracy of sub 5cm. Wouldn't that take the fun out of finding things if all GPSs were as accurate?

Edited August 23, 2006 by obfuscater

[+Red90]

Disagree all you like...... I cache in groups ALL OF THE TIME........ We zero out at many, many locations and what I have stated holds true. It is VERY RARE that anyone will be 20 feet away if there is an open view of the sky. 99% of the time everyone zeros out within a good 5 foot circle.

[+hogrod]

Disagree all you like...... I cache in groups ALL OF THE TIME........ We zero out at many, many locations and what I have stated holds true. It is VERY RARE that anyone will be 20 feet away if there is an open view of the sky. 99% of the time everyone zeros out within a good 5 foot circle.

 

Just like you said, you can disagree all you want, but your very wrong. we usually have 2-4 gps with us and yes quite often they are all within 10ft of each other, sometimes they can be much farther off. As stated above by someone any closest to pin contest will prove how wrong you are.

If I mark the same location 100times over 10days I will not get the same coords everytime, acording to you I should. There would be no reason for position averaging to get good coords if gpsr were that accurate.

I can compare my 60cx and legendC and they usually never display the same coords, but are usually within 20ft of each other. I think alot of it has to do with how quick the unit updates, It seems to me that the legendC takes over 5minutes to really settle into knowing where its at, and 60cx doesnt even take half that time.

I was at a cache the other day with some poor coords, I was showing 65ft off(with 60cx) in the area most likly to contain the cache, the legendC was only at about 30ft off. we searched about 5minutes before we found the cache and by then the legendC was also showing about 60ft off.

Edited August 23, 2006 by hogrod

[+Red90]

but your very wrong.

 

Please explain how I can be wrong when stating my personal experience???? I have done hundred of caches with groups and this is MY EXPERIENCE.......

 

Remember, I said "with an open view to the sky". Certainly with cover around, the circle widens, but the photos appear to be in optimal reception conditions.

[+GPSlug]

As stated above by someone any closest to pin contest will prove how wrong you are.

If I mark the same location 100times over 10days I will not get the same coords everytime, acording to you I should. There would be no reason for position averaging to get good coords if Commercial grade gpsr were that accurate.

Measurements taken at different times are quite a different thing from measuremets taken at the same time in open sky. I can certainly see two receivers with the same satellites being almost always within 10 ft. But you also have to make sure neither user is blocking that open sky with their body.

[+Red90]

Measurements taken at different times are quite a different thing from measuremets taken at the same time in open sky. I can certainly see two receivers with the same satellites being almost always within 10 ft. But you also have to make sure neither user is blocking that open sky with their body.

 

Yes, if it was not clear above, that was what I was getting at. When you arrive AT THE SAME TIME, the errors between units in much smaller than the actually error as much of the error is time based do to satellite position and atmospheric conditions. When checking a position at different times, the error circle increases appeciably.

[romulox]

Wow! Thanks for the updates. I did some more testing which is inline with Sputnik57's observation. At the same time - maybe I have a bum unit! So here goes...

 

Wasting more time at work... out we went. This time I was not concerned with finding coordinates, just walking around for about 1/2 mile in the open - stopping a few times to compare my position between the 2 units. While stopped, I would notice the sats being picked up and then move one of the units away until their positions read to be the same.

 

Test 1: units read the same position at 15' apart - both of us picking up a few different sats

Test 2: units read the same position at 12' apart - both of us picking up a few different sats

Test 3: units read the same position at 21' apart - both of us picking up a few different sats

Test 4: units read the same position at 8' apart - both of us picking up a few different sats

 

-if I stop right here, I understand why this is thanks to everyone's posts. I totally do -

 

Tests 3 and 4, however, included a cache. In Test 3, the cache was 7' away from the 60cx and about 20' away from me (the Legend cx) and our positions were reading the same. In Test 4, the cache was 1' away from the 60cx and 9' away from me (the Legend cx) and our positions read the same.

 

This trend in tests 3 and 4 is common with me and my friends 60cx. Hence my post on the current "legend is inaccurate" thread. He's is consistantly, actually always, closer to a waypoint (created by a thrid party) than I am eventhough were both reading the same position.

 

It would seem to me, a dummy who skips out on work to mess with this, that a waypoint might be more "equidistant" between us while we're standing 20' apart.

 

Anyway, in addition to the normal factors of GPSr receiving differences, the SiRF III chipset maybe what's performing all of the magic. The truth of the matter is that between MY legend and HIS 60cx (I'm not saying all legends or 60s) he has the superior unit. And rightfully so, right? He paid more and is using the new SiRF architecture.

 

I'm sure others have different stories of how their legend is dead-on or how their 60cx stinks, but I can just report my case and observations. Plus, I'm tired of testing, tired of dealing with this, and I'm 60 days out of my return policy so I can't really go out and get the 60cx.

 

I think I'll cut down on caching (I'm scared of snakes and aligators anyway) and switch to survey markers - those are pretty obvious from 20' away.

 

Shame for by buddy's accuracy - he has no thrill of the hunt! (See you tomorrow dude )

 

thanks guys/gals - excuse the spelling.

[+Red90]

No there may be something wrong with these units. I have compared many models on our group caching trips and if there are no obstructions they all get the same accuracy (even the cheapest Geko) and all zero out in basically the same location. Under heavier cover the older unit lose accuracy and wander off more.

[+Zinnware]

I did a test the other night with a 60CS and a 60csx. I turned both GPSs for 15 minutes or more and took them out to my patio.

I placed them upright on top of the patio table's umbrella post (hanging from the knob on the back of the GPS units). Both units were 3 inches apart and were back to back. I began marking a waypoint by using the average facility. Each took 26,000 samples (1 per second) I saved the marks after the 26,000 samples.

I thought it was important to test both units at the exact same time because they would (potentially) be recieving the same signals form sats.

 

The GPSMAP 60cs unit showed 1.9 foot accuracy and the GPSMAP 60csx showed 2.1 foot accuracy when the test was completed. The marks were 6.75 feet apart from each other. I guess this could be a result of the units facing different directions.

 

When I compared the marks on Expert GPS's aerial photos, it appeared that the 60csx was closer to the actual location on the map. I had to zoom in a great deal to see the difference, but it could be done.

Edited August 24, 2006 by Zinnware

[yeeoldcacher]

I'll add my own observations about the differences that different gps's show. I work in a building where we have a re-radiating antenna inside the building. I have used a combination of gps units and decided to compare a few of my friends gps's with two of mine. I have placed an Explorist 300, a Lowrance iFinder, a Garmin GPSMAP60cs and an eTrex Legend. At varying periods of the day I would record the location coords. showing on the unit and would typically see a variance of .001-.003 in lat and/or long. This even with the fact that I know that each unit has access to the same data from the one centralized antenna. At that point I realized that any comparison in the field was not worth the time unless it was to determine functionality or just to see how close they can get at a particular moment.

Edited August 24, 2006 by yeeoldcacher

[+Wacka]

Personally, I think there is something wrong. In my experience, in an open area like that you should ne much closer. Also the "error" tends to mainly be atmospheric and time of day errors (not random). When caching in groups almost always everyone zeros out within a 5 foot circle with a clear view of the sky like shown in those photos.

I hate to disagree, but I saw a photo in the main forum a year or so ago taken at an event. The organizers took a bearing off of a benchmark to fix coordinated in a field. Everyone got a small red flag and were asked to find ground zero. There must have been 30 flags in the field, fairly randomly arranged, in about a 50 foot circle. Closest to ground zero got a prize. As I recall, he or she was about 2 feet off. If your group is always within a 5 foot circle, I would call that extraordinary!

 

We did that at an event. The person that won waited until almost everyone had placed their flag, Then without using aGPSr, he placed it where he thought it should be. Was off by only a few feet!

[romulox]

Hi guys... I lied (I said I wouldn't do more testing). I did 1 more test after I remembered I had an old Etrex Yellow. It's not conclusive with only 3 tests, but my Etrex Yellow is apparently more accurate than my Legend

 

In this picture, you can see the positional differences. Something I would suspect to be the case right off, and would assume the Legend is more accurate - since it's a fancier model. Both units were on for 20 minutes (car ride to work):

 

 

 

In this photo I'm at cache #1. All the pics are cell phone pics, so the angles are skewed. Anyway, I think the picture speaks for itself. I'm holding the Legend. The yellow and legend are at "0 feet" to target in all photos.

 

 

Moving on to cache #2. In this one, my Legend and Yellow is equidistant to the cache. But the log for this cache says the coordinates are wrong and it puts you in the ditch in front of the Hooters sign (Yellow wins again):

 

 

Finally, cache #3. This was the most disapointing. The Etrex Yellow is 1 foot away from the cache, my Legend (almost off camera) is 30' away, and I'm standing far back to try to photograph the results.

 

 

==========================

 

My Etrex Yellow is performing the same as my buddy's 60cx. And better than my Legend! (NOTE: I keep saying "mine" and "his" becuase I do want want to conclude that ALL legends are behaving oddly)

 

Any advice? Call Garmin? Email Garmin? Get a replacement unit? Just forget it? My tests are flawed? Sell this and buy a 60cx?

 

Thanks!

Edited August 24, 2006 by romulox

[+budd-rdc]

 

My Etrex Yellow is performing the same as my buddy's 60cx. And better than my Legend! (NOTE: I keep saying "mine" and "his" becuase I do want want to conclude that ALL legends are behaving oddly)

 

Any advice? Call Garmin? Email Garmin? Get a replacement unit? Just forget it? My tests are flawed? Sell this and buy a 60cx?

 

Thanks!

That can't be! Your eTrex Yellow is lying, since it can't have EPE better than 12'!

 

I don't think your tests are flawed. I'd call (but not e-mail) Garmin and ask for a replacement. Not sure if you want to sell the current unit and pass the problem on to the next guy, unless he accepts the fact that GPSr's are ALWAYS 25' off.

 

I suspect you have a bum unit, although the consistency of the error is very strange.

 

Thank you for sharing your observations.

[GeoBobC]

Here is an interesting consideration: I bought a 60cX in May to record tracks of trails that I frequent. I have recorded tracks since 1999 using a variety of Garmin GPS units, starting with a 12XL and progressing through various Legends, eTrex, and several 60c units. Now I have the "highly sensitive" 60cX with SiRF chip. I record the same trails and download them to Mapsource where I've stored six years of tracks. I would expect to find a tight cluster of 60cX tracks near the middle (average) of all the other tracks. In many instances, I do not. I have found on numerous occasions that the tight cluster of 60cX tracks is near the fringe, and in some instances completely outside the cluster of other tracks. Given the large margin of error I encountered with some of those early units, how can that be? Do different units compute location differently?

[+Sputnik 57]

My conclusions (for what they are worth):

 

- Red90 has had exceptionally good fortune based upon my personal experience. I would say that consistant results within 5 feet, even in the best of conditions, is not typical.

-Your repeated and consistant errors are not normal, and as Red90 points you, you definitely a problem

-You're friend's good fortune with the 60CSx is also a bit atypical. The SiRF chip may help, but remember that the cache hiders are marking their coordinates with their Legends and Etrexes (Etrei?) and so the listed coordinates my not be spot on.

-If your test results are right AND YOU ARE ABSOLUTELY TRIPLE SURE that your datum is set to WGS84, as Viamundi aptly pointed out, then I would contact Garmin for a repair/replacement.

[+Sputnik 57]

Here is an interesting consideration: I bought a 60cX in May to record tracks of trails that I frequent. I have recorded tracks since 1999 using a variety of Garmin GPS units, starting with a 12XL and progressing through various Legends, eTrex, and several 60c units. Now I have the "highly sensitive" 60cX with SiRF chip. I record the same trails and download them to Mapsource where I've stored six years of tracks. I would expect to find a tight cluster of 60cX tracks near the middle (average) of all the other tracks. In many instances, I do not. I have found on numerous occasions that the tight cluster of 60cX tracks is near the fringe, and in some instances completely outside the cluster of other tracks. Given the large margin of error I encountered with some of those early units, how can that be? Do different units compute location differently?

 

Have you tried plotting these in USAPhotomaps topos or aerials to see if you can tell which track is closest to on-the-ground routes?

[+Fhantazm]

This just basically re-confirms that there is something drastically wrong with the etrex "x" line. I have the same sort of issues with my Venture Cx, as do more than a few other "x" owners. Its a shame too, cause otherwise they are great lil units.

By the way, VERY thorough tests! Thanks for sharing this with us!

Edited August 24, 2006 by Fhantazm

[GeoBobC]

"Have you tried plotting these in USAPhotomaps topos or aerials to see if you can tell which track is closest to on-the-ground routes?

 

No, I haven't so I don't know what the "right" answer is. It is entirely possible, if not probable, that the 60x tracks are closest to the truth.

[romulox]

Hi Guys.... No one has reported back on the success of the 2 "fixes", so I tried them and here's my report for all of you Legend cx owners.

 

I did both the "reset your unit" with the (page+enter) + power instructions, then I followed that up with the "select new location" instructions.

 

To make sure my GPSr was in the best possible position to start anew, I placed the unit on the apex of my roof - no steel, no trees, no surrounding buildings - just open sky.

 

 

As a side note for anyone just wanting to learn about what happens when units are reset - I don't know what it exactly does internally, but I can tell you what I observered:

1) firmware stayed the same (naturally)

2) All personalization (i.e. day/night mode, lock on road, page sequences, etc...) are reset back to defaults

3) All waypoints are deleted

4) Tracks stored in the unit's internal memory are deleted

5) Things stored on the SD card are not touched.

 

In any event, the 2 "fixes" ended up, for me, being just like the cable company telling you unplug your modem for 5 minutes - it did nothing that I could notice. IN FACT.... it apprears to have made my problem slightly worse (but I really can't prove that 100%)!

 

I revisted all of my test points. Historically I've been going back to the same wrong locations, while good old etrex yellow is putting me to shame. Af ter the "fixes", I appear to be longitudinally off by 4 to 6 feet to my original wrong readings from the Legend - and not in a good way; like, I'm being put even farther away from a waypoint!

 

For me, it feels like my whole problem is that GPSr kind of has me shifted off "center" with respect to longitude. And when I say "center", I don't mean "dead on coordinates"; I mean more like a conceptual mean or average or approximate center. A center that everyone can enjoy.

 

I'll call Garmin today, see what they have to say.

 

I don't know if "cross posting" is frowned upon in these forums, but I'm going to post this entry in http://forums.Groundspeak.com/GC/index.php?showtopic=134219 since they're turning into the same thing.

 

PS: To Sputnik 57 - I'm triple quadrupally sure my datum is set to WGS84

Edited August 25, 2006 by romulox

[calic]

I'm not sure if any one mentioned this or not, didn't see it mentioned but one reason why 2 Garmins show same position and are 25' away could be because one had their WAAS turned on and the other had the WAAS turned off.

[+Train_Man]

Hi Guys.... No one has reported back on the success of the 2 "fixes", so I tried them and here's my report for all of you Legend cx owners.

 

I did both the "reset your unit" with the (page+enter) + power instructions, then I followed that up with the "select new location" instructions.

 

To make sure my GPSr was in the best possible position to start anew, I placed the unit on the apex of my roof - no steel, no trees, no surrounding buildings - just open sky.

 

 

As a side note for anyone just wanting to learn about what happens when units are reset - I don't know what it exactly does internally, but I can tell you what I observered:

1) firmware stayed the same (naturally)

2) All personalization (i.e. day/night mode, lock on road, page sequences, etc...) are reset back to defaults

3) All waypoints are deleted

4) Tracks stored in the unit's internal memory are deleted

5) Things stored on the SD card are not touched.

 

In any event, the 2 "fixes" ended up, for me, being just like the cable company telling you unplug your modem for 5 minutes - it did nothing that I could notice. IN FACT.... it apprears to have made my problem slightly worse (but I really can't prove that 100%)!

 

I revisted all of my test points. Historically I've been going back to the same wrong locations, while good old etrex yellow is putting me to shame. Af ter the "fixes", I appear to be longitudinally off by 4 to 6 feet to my original wrong readings from the Legend - and not in a good way; like, I'm being put even farther away from a waypoint!

 

For me, it feels like my whole problem is that GPSr kind of has me shifted off "center" with respect to longitude. And when I say "center", I don't mean "dead on coordinates"; I mean more like a conceptual mean or average or approximate center. A center that everyone can enjoy.

 

I'll call Garmin today, see what they have to say.

 

I don't know if "cross posting" is frowned upon in these forums, but I'm going to post this entry in http://forums.Groundspeak.com/GC/index.php?showtopic=134219 since they're turning into the same thing.

 

PS: To Sputnik 57 - I'm triple quadrupally sure my datum is set to WGS84

 

Lucky the unit didn't fall off your roof!

 

I once had very bad signal locks after going across the country, so i reset it and left it out to 'download' new almanac data and it was MUCH better within 1/2 hour. I don't know what your issue is, but that might just be common inaccuracies.

[+gpsblake]

The only accurate way to test an accuracy of a GPS unit is to go to a place or mark where the lat/long is known and measured. USGS benchmarks that have the lat/long ADJUSTED are best for this. Geocaching.com has 100,000s of them listed at www.geocaching.com/mark/

 

Then take your GPS units and place them right on top of the mark. See how accurate they are that way. Then you will know for sure.

[NewZealand]

I'm not sure if any one mentioned this or not, didn't see it mentioned but one reason why 2 Garmins show same position and are 25' away could be because one had their WAAS turned on and the other had the WAAS turned off.

I saw in one of the postings above, that both Garmins used different WAAS satellites (posting 8).

In that case it's clear that the position is so different, and that the difference doesn't change.

I would recommend to repeat the test with both units having WAAS turned off.

Edited November 8, 2006 by NewZealand

[+Red90]

It is NOT different WAAS settings. They were getting different WAAS satellites. That makes NO DIFFERENCE. All WAAS satellites send the same data.

[John E Cache]

It is NOT different WAAS settings. They were getting different WAAS satellites. That makes NO DIFFERENCE. All WAAS satellites send the same data.

I thought so too, but then I remembered that long thread about WAAS almanacs. The WAAS sats, even though they have almost twice the altitude, do send position info apparently.

 

This whole thread is fasinating. What the first experiment was to me was a manual DGPS system. DGPS can be in the inches accuracy so I would expect the consistant 25' plus or minus inches. Why they are different is a mystery. Maybe they are picking different sats for the calculations. Don't the xDOP numbers tell you what the geometry is? And EPE too?

[Grasscatcher]

You can go to the Trimble website and download their Trimble Planning software (free).

 

Then you can set it up for a particular specific location, and be able to tell what all the xDOP's are for that particular location at any specific time, and which satellites are visible to your unit. You have to be sure and get the latest almanac (from Trimble site). I can tell when will be the best time to GPS a location and what and how many satellites will be visible, and which specific number birds I will be able to see, even before going out in the field.

 

I don't think it will identify the WAAS birds.

 

Using that software and a little intuition, the user could probably tell which unit is the "more correct" one.

Likely one is using a different almanac and seeing different waas birds.

Going to an "adjusted" BM location would also tell which was more accurate, but would not address the "Why" issue.

[+Timpat]

I just noticed this interesting thread. I have a 60Cx, and at least two caching friends have a LegendCx. After some informal field comparison tests, we see that the LegendCxs are consistently indicating a shift in position of .010' Longitude to the west. In other words, if the 60Cx shows W. XXX° xx.000', the LegendCxs show W. XXX° xx.010'. At our latitude of N43°, .001' of Longitude equals about 4.5 feet, so this shift is about 45 feet between units.

 

This has been observed with all units being in the same location at the same time, and with all other factors being equal. We checked this because caches hidden with our friends LegendCxs were fairly consistently being reported to be 40'-50' off. This is always seen with a longitudinal shift of about 45 feet average, not latitudinal.

 

I can also say with confidence that my 60Cx is giving good positions. I survey and always compare my 60Cx with known control points and rarely see this kind and amount of shift. Very interesting to read all the other observations and comments here on this. We have to conclude (with our very informal tests) that the LegendCxs are the offending units, how ever all the factors come into play.

[John E Cache]

I just noticed this interesting thread. I have a 60Cx, and at least two caching friends have a LegendCx. After some informal field comparison tests, we see that the LegendCxs are consistently indicating a shift in position of .010' Longitude to the west. In other words, if the 60Cx shows W. XXX° xx.000', the LegendCxs show W. XXX° xx.010'. At our latitude of N43°, .001' of Longitude equals about 4.5 feet, so this shift is about 45 feet between units.

Another very interesting point. When you are standing still you are actually moving because of the earth's rotation. Could it be that the LegendCxs isn't compensating for the earth's rotation properly? Maybe by the time taken for the calculation is so long the earth has rotated .01'. How much time does it take to move .01'?

[BacktrailJitters]

I just noticed this interesting thread. I have a 60Cx, and at least two caching friends have a LegendCx. After some informal field comparison tests, we see that the LegendCxs are consistently indicating a shift in position of .010' Longitude to the west. In other words, if the 60Cx shows W. XXX° xx.000', the LegendCxs show W. XXX° xx.010'. At our latitude of N43°, .001' of Longitude equals about 4.5 feet, so this shift is about 45 feet between units.

Another very interesting point. When you are standing still you are actually moving because of the earth's rotation. Could it be that the LegendCxs isn't compensating for the earth's rotation properly? Maybe by the time taken for the calculation is so long the earth has rotated .01'. How much time does it take to move .01'?

[BacktrailJitters]

Sorry...double post. IIRC,at the equator the earth rotates at one nautical mile in four seconds.Now...where is my calculator?

[tdor]

You can go to the Trimble website and download their Trimble Planning software (free).

 

Then you can set it up for a particular specific location, and be able to tell what all the xDOP's are for that particular location at any specific time, and which satellites are visible to your unit. You have to be sure and get the latest almanac (from Trimble site). I can tell when will be the best time to GPS a location and what and how many satellites will be visible, and which specific number birds I will be able to see, even before going out in the field.

 

I don't think it will identify the WAAS birds.

 

Using that software and a little intuition, the user could probably tell which unit is the "more correct" one.

Likely one is using a different almanac and seeing different waas birds.

Going to an "adjusted" BM location would also tell which was more accurate, but would not address the "Why" issue.

 

Actually it does show the WAAS sats, as well as Galileo and Glonass sv's.

[Motorcycle_Mama]

You just bumped a thread that is 2 1/2 years old. So my question would be "Why?".

[+tedwin]

You just bumped a thread that is 2 1/2 years old. So my question would be "Why?".

 

I'm glad you did, I missed it first time round, it makes fascunating reading. I just wonder if the origanal unit complained of was replaced?

Tedwin

[+Renegade Knight]

...My question is, with 3 tests, why would a legend cx and a 60cx be physically so far apart when our GPSrs are are displaying the exact same positional data? ...

 

To us an anology. You pick up two identical darts and throw them at the same spot on the dart board. You hit two different spots. So far as you can tell the darts weighed the same, the board didn't move, you threw using the same style and were aiming for the same spot. The difference is errors in the system. You used a slightly diffent throw speed, your muscles moved a tad differently etc. which resulted in a different spot on the board.

 

A GPS has a map of the world using a coordinate system. That's the target. However it has to calculate where it is using information that's subject to error. That error depends on a lot of factors. The short version is that two GPS's are both "aiming" for the same point, but they only hit it with imperfect precision. Those errors can result in two GPSs being 25' apart but reporting that you 'hit the spot'. It is the spot as far as that GPS is concered but the errors have you standing apart.

 

Whoops. Old thread. Oh well.

Edited March 10, 2009 by Renegade Knight

[+Zinnware]

Here is an example of a dart board with two darts. See the link below. An Oregon 400t (version 2.85) and a 60csx (version 3.70). The 60csx tracks are in red and the Oregon tracks are in green.

 

http://zinnware.com/HighAdv/Geocaching/GPS...7_GPS_Tests.pdf

 

Tracks were laid every second over a period of 2 hours and 40 minutes. This is roughly 9,600 samples. The GPSs were placed in the same location at the same time. Both had their WAAS turned on, but I do not think the Oregon 400t ever got a lock on WAAS. I can't be sure. The 60csx did get a lock. Note: the scale of the map in the lower left of the screen.

 

The 60csx tracks take up (GPS wandered) an area of approximately 50 square feet

The Oregon tracks take up (GPS wandered) an area of approximately 67 square feet.

 

What does this prove? Maybe the 60csx is about 34% more accurate and that Garmin needs to fix the WAAS on the Oregon 400t model.