Gps60sx Wandering

[+rdb3141]

My 60csx is wandering badly when stationary, so much that the odometer shows that distance is being traveled when the GPS is stationary. Satellite page shows an adequate lock. Any answers appreciate!

 

 

PS - I searched for an answer before posting.

[Neo_Geo]

My 60Cx wanders a lot too... when it has an "adequate lock" indoors in my living room sitting on my coffee table. I expect it to wander a little in that environment as the signals are not being received directly from the satellites. This is called multipath error and it's caused by the additional time it takes the signal to bounce off of objects and be deflected to the GPSr.

 

By adequate lock, I mean that my GPSr has a lock on the satellites and is providing me with navigation data. Please define your term "adequate lock" and describe the environment in which your GPSr wanders.

[+GOT GPS?]

This is why I bought the 60Cx, instead of the 60CSx, because a GPS of this type works best when held vertically, and does poor when held flat, and because the electronic compass has to be held flat, I got the 60Cx.

 

My 60Cx, when indoors, will rack up a ton of miles when laying down, but when it is in it's holder standing vertically, it does not rack up the miles, and the same thing outdoors, when im on a run, and waiting at a light, it will stay at zero speed, but if I hold the GPS so that it is NOT vertical, it will indicate speed at the red light.

 

The older GPS units would lose siganl, when NOT held vertical, or enough exposure to the sky, but the new X-series GPS units will amplify the lower quality signal to the point, that it will track the satellites, even when it is in the glove compartment in the car.

[+rdb3141]

I thought it might be multipath. By adequate, I mean a clear lock (not indicating a weak signal) and without satellites going in and out after the lock. More than adequate is outdoors when I'm getting nine bars with strong signal stregnths.

 

I will do some comparisons and maybe post some scatter plots for the group - inside lying vertical and horizontal and outside lying vertical and horizontal. I also noticed on a bike ride (with the GPS lying flat inside a pouch behind the seat) the mileage was short when compared with the track log.

 

This is troubling and to me is a significant drawback to the new chip - trading better reception for unreliable results when there are frequent stops. I will appreciate any more technical analysis of the question.

[+zahadoom]

This is why I bought the 60Cx, instead of the 60CSx, because a GPS of this type works best when held vertically, and does poor when held flat, and because the electronic compass has to be held flat, I got the 60Cx.

 

My 60Cx, when indoors, will rack up a ton of miles when laying down, but when it is in it's holder standing vertically, it does not rack up the miles, and the same thing outdoors, when im on a run, and waiting at a light, it will stay at zero speed, but if I hold the GPS so that it is NOT vertical, it will indicate speed at the red light.

 

The older GPS units would lose siganl, when NOT held vertical, or enough exposure to the sky, but the new X-series GPS units will amplify the lower quality signal to the point, that it will track the satellites, even when it is in the glove compartment in the car.

 

Ok I have a question, i need to replace my old etrex and have been looking that the 60cx

 

This wandering worrys me..... But you all end up saying INDOORS..... Does it wander OUTDOORS when standing still??

It it not able to realize that you have stopped moving now? And never starts averaging your location??

[+YuccaPatrol]

The wandering is a result of the very high sensitivity of these units. They are capable of receiving signals and maintaining lock in situations where other units would not be able to calculate any position.

 

So when I am inside my brick house with my CS and CSX side by side, My 60CS tells me that it doesn't know where the heck I am because it can't see any satellites, but my 60CSX knows I am inside my house but isn't always sure if I am in my office or in the adjacent bedroom.

 

The fact that I can get a useful signal under adverse conditions is awesome, and I don't see the wandering as any sort of real problem since I am getting useful location information when I would otherwise have none.

 

When the CSX is outside with direct view of at least a few satellites, this "problem" goes away.

Edited April 19, 2006 by YuccaPatrol

[dfred]

Indeed... And I'm somewhat worried the cure would be worse than the disease.

[planewood]

With my 60cx I have found that the quality of the recorded data is directly proportional to the quality of the signals received at the time the position was marked. The more maxed out solid bars you have, the better the recorded data. With 7 or 8 maxed out bars you should be getting 3-5 meter accuracy at least 95% of the time. But, you gotta have something to judge the quality against if your trying to test the unit. A differentially corrected Benchmark with a known position is the best.

 

Also, my experience has shown that the stated accuracy of the unit is also a good indicator of how well the unit is performing. If it's showing 7'-8' accuracy, then your probably getting a lot better results than if it's showing 20'-25' accuracy.

 

But, it can throw in a 'flyer' at any time. That's why they say "95%" of the time.

 

I have no other GPS to compare to but the 'x' series seems to be pretty stinkin good! Yeah, it's got software issues, but the positioning part of it is darn good. An external antenna will definitely improve signal reception.

 

Also, get the Trimble Planning program (version 2.7). It will tell you the time of day when you can expect best results. I use the # of satellites bar chart and the horizontal DOP chart. It's free on the Trimble web site.

 

Like most highly technical machines, GPS units have a very steep learning curve that one has to negotiate. I think I may now be half way up that slippery curve (thanks in most part to this forum). I started at the bottom of the curve about 2 months ago.

[+rdb3141]

Great info, thanks. I just ate lunch outside with my GPS horizontal on the table:

 

Over ~40 minutes

Six to seven satellites

Urban environment but facing an open park with a buidling behind me.

Reported accuracy ~20 feet

 

Odometer 7 feet

Clocked 39 minutes 44 seconds stopped and 1 minute 31 seconds moving

maximum speed 1.5 mph

Tracklog showed furthest outlier about 50 feet.

 

Not bad. And the reception of the unit is outstanding. Not ready to return it!

[+zahadoom]

Great info, thanks. I just ate lunch outside with my GPS horizontal on the table:

 

Over ~40 minutes

Six to seven satellites

Urban environment but facing an open park with a buidling behind me.

Reported accuracy ~20 feet

 

Odometer 7 feet

Clocked 39 minutes 44 seconds stopped and 1 minute 31 seconds moving

maximum speed 1.5 mph

Tracklog showed furthest outlier about 50 feet.

 

Not bad. And the reception of the unit is outstanding. Not ready to return it!

 

Ok thats good, but it does seem the unit does NOT average your location when your are not moving.

 

Would be nice to be able to tell the unit your are Stopped and let it start averaging your location over time to give you a ever more percise location no?

 

I know with my old etrex every now and again I would get a tracklog that was way off, but usally when I stood still it would NOT keep leaving more and more tracklogs.

 

Edit: Just looked at the manual online now... There is a Average waypoint fuction, which is awsome BTW!! For placing a geocache, but would be nice if you could switch to that mode quickly while walking, and then back when you start to move again....

Im glad to see the average though.... Im happy with that..

Edited April 19, 2006 by zahadoom

[planewood]

 

Ok thats good, but it does seem the unit does NOT average your location when your are not moving.

 

Would be nice to be able to tell the unit your are Stopped and let it start averaging your location over time to give you a ever more percise location no?

 

I know with my old etrex every now and again I would get a tracklog that was way off, but usally when I stood still it would NOT keep leaving more and more tracklogs.

 

Edit: Just looked at the manual online now... There is a Average waypoint fuction, which is awsome BTW!! For placing a geocache, but would be nice if you could switch to that mode quickly while walking, and then back when you start to move again....

Im glad to see the average though.... Im happy with that..

 

When I hit the average button for a waypoint I always watch the lat/lon window to see if the values change. If they change significantly while preforming the average, I stop and mark the position again and restart the averaging. I wonder if the averaging is doing any kind of 'weighted' averaging??? Like throwing out the most anomalous? Since it's doing it on the fly, I kinda doubt it.

 

Also, due to satellite configuration at the time, all averaged waypoints or tracks can still be 'off the mark'. I've never seen mine track around a known point. It's always been on one side or the other of the known point. The largest 'flyers' I've observed so far has been 4 or 5 points thrown in that were about 100' away. All the rest of the points were within 10'. Also, if you go back to a known point the same time each day to mark the position, your probably getting the same satellite configuration and a good chance of getting the same errors.

[robertlipe]

This is analyzed in some detail at http://www.mtgc.org/robertlipe/showdown/

 

The X models definitely return your position in a "cloud" - they don't do a static position average when they deduce you're standing still. For some uses that's good and others it's bad, but that's what they do. Explorist definitely does this averaging; that's why Explorist returns many points with the exact same coords as shown in the tall green spikes in the 3D graph under "My Ghost Likes to Travel".

 

That said, the "wander" is actually more of a "jitter" - I've not calculated the edge-to-edge distance of that red flat cloud (indicating the jitter on the CSX) but it's small. It's just a few feet a second over the course of a few hours does start ticking into real distance and it does make a messy tracklog for those having the unit powered on in one spot.

 

Now that I've used a 60csx for a few hundred finds, I find this to be a total non-issue while caching.

Edited April 19, 2006 by robertlipe

[+zahadoom]

This is analyzed in some detail at http://www.mtgc.org/robertlipe/showdown/

 

The X models definitely return your position in a "cloud" - they don't do a static position average when they deduce you're standing still. For some uses that's good and others it's bad, but that's what they do. Explorist definitely does this averaging; that's why Explorist returns many points with the exact same coords as shown in the tall green spikes in the 3D graph under "My Ghost Likes to Travel".

 

That said, the "wander" is actually more of a "jitter" - I've not calculated the edge-to-edge distance of that red flat cloud (indicating the jitter on the CSX) but it's small. It's just a few feet a second over the course of a few hours does start ticking into real distance and it does make a messy tracklog for those having the unit powered on in one spot.

 

Now that I've used a 60csx for a few hundred finds, I find this to be a total non-issue while caching.

 

Thats good to know, because Geocaching is one of the things I want to use it for... I was just worried about if I wanted to place one....

 

EDIT: Ive read that article now, very interesting, but Im noteing that he was using a raw data dump. And he never used the AVERAGE waypoint built into the unit, I wonder how it preforms, but it would be nice to be able to quickly turn it on or off from the map screen.

 

So anyone know how good the averaging is on the 60Cx? Like was mentioned is it a weighted average, I saw that it highlights how many points were observed to give it, and its approx range. But how low can it go based on how long you leave it on?? Anyone tested this?

Edited April 19, 2006 by zahadoom

[Hertzog]

When I hit the average button for a waypoint I always watch the lat/lon window to see if the values change. If they change significantly while preforming the average, I stop and mark the position again and restart the averaging. I wonder if the averaging is doing any kind of 'weighted' averaging??? Like throwing out the most anomalous? Since it's doing it on the fly, I kinda doubt it.

 

I've compared the waypoint averages with track values during the same time period, and it appears to be a straight average over the waypoint average time. But the periods I checked were "good" (no anomolies, and little change in EOD), so its possible they do some sort of weighting and rejection of anomolies. Interesting point; when I have some time I'll try some tests where I "disturb" the tracking and see what happens.

Edited April 19, 2006 by Hertzog

[+zahadoom]

When I hit the average button for a waypoint I always watch the lat/lon window to see if the values change. If they change significantly while preforming the average, I stop and mark the position again and restart the averaging. I wonder if the averaging is doing any kind of 'weighted' averaging??? Like throwing out the most anomalous? Since it's doing it on the fly, I kinda doubt it.

 

I've compared the waypoint averages with track values during the same time period, and it appears to be a straight average over the waypoint average time. But the periods I checked were "good" (no anomolies, and little change in EOD), so its possible they do some sort of weighting and rejection of anomolies. Interesting point; when I have some time I'll try some tests where I "disturb" the tracking and see what happens.

 

Heres the quick question, how long can you leave the average screen up vs. how accuarate can the location get on the this screen?

 

And if it gets really really close, we really need to get Garmin to add this to the map screen to quickly select the option that we have stopped/rest/camp/geomap. Something along this line.

[robertlipe]

 

EDIT: Ive read that article now, very interesting, but Im noteing that he was using a raw data dump. And he never used the AVERAGE waypoint built into the unit, I wonder how it preforms, but it would be nice to be able to quickly turn it on or off from the map screen.

 

The author of that work finds it weird that we're discussing him in the third person. :-)

 

Continuous position information was used becuase that's the most practical way to get data out of the unit for study. Anything involving manually pushing buttons is impractical when you're working with several receivers at once and trying to not block the reception of any of them.

 

Though I've not done any math to prove it, I suspect 'waypoint averaging' is just taking the mean of successive position fixes and then examining mean deviation or a standard deviation (this is tradionally where Fizzymagic steps in and delivers a wedgie) to assign a "goodness" factor to the numbers based on HDOP/VDOP and time stability of the values. When you have reasonable reception, I find the difference between just letting the receiver sit for a few moments until the numbers stabilize and messing with the averaging function to be a distraction.

 

In short, if you place a cache with one of these while you have a good view of the sky and a low EPE, you'll be fine. Just learn to ignore the logs that go:

 

"GPS took me to a spot, but I could not find the ammo box. I twisted my ankle while examining that 2 foot circle becuase I kept tripping over a hollow stump 3 feet away that had 4,000 parallel sticks on it..."

[planewood]

 

Heres the quick question, how long can you leave the average screen up vs. how accuarate can the location get on the this screen?

 

And if it gets really really close, we really need to get Garmin to add this to the map screen to quickly select the option that we have stopped/rest/camp/geomap. Something along this line.

 

Zaha -

 

It will keep averaging until you click the save button. But, even if you let it go for an hour, there is no guarantee that the final solution will be any better than the first reading. Statistically, it will probably be a better answer. In times of poor reception I've seen the averaged answer as much as 20' - 30' off.

 

With mine hooked up to an antenna, my SOP is to let it average 100 times. Most of the time, as I watch the lat/lon, the values do not change from the first solution. I generally wait about 5 minutes at each new location giving the unit time to stablize before marking and averaging the position.

[GeoBobC]

So here's a question from a 60c owner considering the purchase of a 60cx: with a 60cx (no magnetic -in compass) does the map orientation change when stopped? In other words, if the map is oriented as "heading up" and I stop walking while heading south, does the orientation remain as "south up" or keep flipping around as the position wanders slightly? That would be REALLY irritating.

[Snipe315]

So here's a question from a 60c owner considering the purchase of a 60cx: with a 60cx (no magnetic -in compass) does the map orientation change when stopped? In other words, if the map is oriented as "heading up" and I stop walking while heading south, does the orientation remain as "south up" or keep flipping around as the position wanders slightly? That would be REALLY irritating.

 

I just brought my brand new GPSMAP 60CSx home. I've had it sitting on a table and not moving for several minutes.

 

The odometer says I've moved over 50 feet.

 

The Map (set to Track Up) hasn't rotated or moved.

 

While I find this "wandering" annoying, I can live with it for now.

 

[+dhbaird]

My 60Cx wanders a lot too... when it has an "adequate lock" indoors in my living room sitting on my coffee table. I expect it to wander a little in that environment as the signals are not being received directly from the satellites. This is called multipath error and it's caused by the additional time it takes the signal to bounce off of objects and be deflected to the GPSr.

 

By adequate lock, I mean that my GPSr has a lock on the satellites and is providing me with navigation data. Please define your term "adequate lock" and describe the environment in which your GPSr wanders.

 

This was my test...

 

I plugged my 76CSX into a Glisson Antenna. Provided external power. Got full 8 bar lock on 10 statellites and then left it sitting in position for 8 1/2 hours. Estimated Positional Error is saying +/- 9 feet.

 

It too wandered, significantly more than my 76CS. Longest leg was 135 meters! Many legs of 20 to 30 meters. Lots more of 10 to 15. And yes, my reset odometer also logged in distance travelled, I wish I had saved a screen shot of that, but I did not. All while sitting dead still and full lock on all satellites each time I checked.

 

The 76 CS in the same test, before I sold it was locked on 6-8 statellites. It did have a WAAS lock this was before Sat 35 was moved) and showed D's and Estimated Positional Error of +/- 12 to 15 feet. Longest leg was 20 meters and the cluster of legs recorded in the tack was so dense that it was very difficult to pick out individual jumps.

 

The comparison with known bench marks has also been very disappointing.

 

I have the benefit to having 3 adjusted benchmarks either on my property or on my walking route. With my old 76CS unit, I could go to the Benchmarks then do find for the Benchmark waypoints in my unit and get a real positional error of less than 15 feet each time. Estimated Positional error was registered at 10 feet or less.

 

With the new unit, I can have it resting on the top of the Benchmark (3 different ones) and it consistently tells me that the BM is 50 to 75 feet away. (Yes I have changed the Datum) As I move in the direction of the pointer, the compass seems to get totally disoriented. Recalibration sometimes helps, but I can move in a circle around the bench mark and the pointer tells me that the waypoint is 75 feet way in several directions. All this is while the unit tells me that Estimated Positional error is less than 10 feet. If this were an actual bench mark search in the field such an error would have quadrupled my area of search over what I used to get with my 76CS.

 

But as someone else said, in the field I got 3 caches fairly easily. Perhaps that is because with more experience once I am within 20 or so feet, caches are beginning to be mre quickly identified.

 

I am beginning to wonder if the SIRF chip itself is prone to these wandering instability problems.

[Snipe315]

UPDATE.

 

After the first "wandering" session my GPSMAP 60CSx seems to have decided to stay put!

 

It hasn't moved at all over several minutes!

 

It was almost like the device was checking the surrounding area and trying to verify it's location within that zone.

 

BTW...

 

I LOVE THIS DEVICE!

[planewood]

Here's my test.

 

The white line represents 19'. The actual location of the Gilsson antenna on top of a 10' pole is just about where the red dot is on the bottom of the white line. The google images are registered quite well in this area. The unit was showing 7'-9' accuracy all of the time with 8 or 9 bars maxed out. This was between 2:00 and 6:00 today.

 

I let the 60cx track in 5 minute intervals for about 4 hrs and it collected 52 points. The image shows all 52 points. (the white spot is my Purple Martin house)

 

[+YuccaPatrol]

Great test! Thanks for that google image.

 

Although I don't have all of the data points to look at, I count less than 10 points further than approximately 10 feet from the antenna.

 

This level of accuracy over time seems very reasonable and acceptable to me

[planewood]

Great test! Thanks for that google image.

 

Although I don't have all of the data points to look at, I count less than 10 points further than approximately 10 feet from the antenna.

 

This level of accuracy over time seems very reasonable and acceptable to me

 

If you notice, about 80% of the points are within 10'.

[+zahadoom]

 

The author of that work finds it weird that we're discussing him in the third person. :-)

 

Continuous position information was used becuase that's the most practical way to get data out of the unit for study. Anything involving manually pushing buttons is impractical when you're working with several receivers at once and trying to not block the reception of any of them.

 

Though I've not done any math to prove it, I suspect 'waypoint averaging' is just taking the mean of successive position fixes and then examining mean deviation or a standard deviation (this is tradionally where Fizzymagic steps in and delivers a wedgie) to assign a "goodness" factor to the numbers based on HDOP/VDOP and time stability of the values. When you have reasonable reception, I find the difference between just letting the receiver sit for a few moments until the numbers stabilize and messing with the averaging function to be a distraction.

 

In short, if you place a cache with one of these while you have a good view of the sky and a low EPE, you'll be fine. Just learn to ignore the logs that go:

 

"GPS took me to a spot, but I could not find the ammo box. I twisted my ankle while examining that 2 foot circle becuase I kept tripping over a hollow stump 3 feet away that had 4,000 parallel sticks on it..."

 

Sorry Didnt realize I was talking to THE Robertlipe! Nice article BTW, thank you. So you do think I should get a Mac or Garmin? I really used to be under tree cover alot, I think I need the sirf III but with everyone saying it off from actual benchmarks makes me wonder.

 

Heres the test Id like to see run.

 

If someone takes a 60Cx or 60Csx to a KNOWN bench mark, Places it on the bench mark hopefully as vertical as possible, Set it to Average mode.

 

Record the starting Co-ord's and approx accuaracy.

 

And then let it run lets say 5 min's (at least, just record how long) (and not sure if your current tracklog setting will effect how many points are recorded in this time, turn it to max for the test) ...

 

After the 5 min's(or alotted time) re-record the data, and compare it to the actual becnhmark.

 

I'd like to see those results, if anyone is up for it? This should be a true test of what the BEST accuaracy in the field you can get from the 60Cx line no?

Edited April 20, 2006 by zahadoom

[planewood]

 

Sorry Didnt realize I was talking to THE Robertlipe! Nice article BTW, thank you. So you do think I should get a Mac or Garmin? I really used to be under tree cover alot, I think I need the sirf III but with everyone saying it off from actual benchmarks makes me wonder.

 

Heres the test Id like to see run.

 

If someone takes a 60Cx or 60Csx to a KNOWN bench mark, Places it on the bench mark hopefully as vertical as possible, Set it to Average mode.

 

Record the starting Co-ord's and approx accuaracy.

 

And then let it run lets say 5 min's (at least, just record how long) (and not sure if your current tracklog setting will effect how many points are recorded in this time, turn it to max for the test) ...

 

After the 5 min's(or alotted time) re-record the data, and compare it to the actual becnhmark.

 

I'd like to see those results, if anyone is up for it? This should be a true test of what the BEST accuaracy in the field you can get from the 60Cx line no?

 

Zaha -

 

Go to this image file on my website: http://pages.sbcglobal.net/mike_in_katy/Pl...od/Clip0011.jpg

 

When the builders put in this addition they surveyed all the streets and cul-d-sac centers using differential GPS from a nearby triangulation station with something like 6" accuracy. I got a copy of their plat when I bought the house. The large dot represents the center of the cul-d-sac. I think the dot is within 2' - 3' from truth.

 

The points were recorded over a couple of weeks in various weather. All the outlyers were on heavily overcast days where the clouds were dark and heavy with moisture. The ones closest to the center were on bright clear days. Lot of the points closest to the center are overlaying each other. 3' is about all the accuracy you get with 5 decimal places of accuracy. The 60cx was showing 20' accuracy when the outliers were recorded and 8'-10' when the others were recorded.

[robertlipe]

Didnt realize I was talking to THE Robertlipe!

Let's assume for a minute it's a common name. How common is it in GPS/Geocaching kinds of circles? THAT is why I use my real name everywhere. :-)

 

So you do think I should get a Mac or Garmin?

"Should I buy a Honda or Pampers?"

"Do you prefer Pink Floyd or fried food?"

"Did you walk to school or bring your lunch?"

If someone takes a 60Cx or 60Csx to a KNOWN bench mark, Places it on the bench mark hopefully as vertical as possible, Set it to Average mode.

 

I've run a large (larger than any single reporter to date in the "What do you own" thread) number of GPSes at a known location. Granted, it's not a surveyed location so I don't know beyond all certainty the coordinates of that location. But for any GPS where I'm able to actually get a lock, they'll typically be within a digit or so of each other. Today this one will be +.001 minute and an hour from now it'll be -.001 minute. It's just not a big deal.

 

Maybe it's time for another page in the article. The outline will be:

 

If you don't have a lock, you don't have jack...

If you do have a lock (and your receiver isn't lying to you you about having one) you're probably OK.

 

I need to find some way to put an absolute perspective on the size of that red scatter plot....

 

Oh, and you're welcome.

[+zahadoom]

So anybody up for that test I was talking about a few posts up there.... Post #25.

 

Take the 60Cx or 60Csx to a known benchmark, position it vertically, set it on AVERAGE mode/screen for at minimum of 5 mins(or longer if you want), record starting co-ords', and then final as well as the posted accuracy and length of test.

 

And of course the benchmark co-ords, so we can compare if the AVERAGE mode can actually improve accuracy.

 

Should be some interesting results.

Edited April 21, 2006 by zahadoom

[planewood]

So anybody up for that test I was talking about a few posts up there.... Post #25.

 

Take the 60Cx or 60Csx to a known benchmark, position it vertically, set it on AVERAGE mode/screen for at minimum of 5 mins(or longer if you want), record starting co-ords', and then final as well as the posted accuracy and length of test.

 

And of course the benchmark co-ords, so we can compare if the AVERAGE mode can actually improve accuracy.

 

Should be some interesting results.

 

Zaha -

 

One thing you could do is turn on 1 second tracking at the time you turn on averaging then turn them both off at the same time. Averaging the track coordinates yourself should show you how the averaging is performed. The advantage with averaging the track yourself is that you could throw out the outliers. I'd bet a plug nickel that the Garmins are just providing a simple average.

 

If you average the track yourself, I would pick a point in the interior of the biggest cluster and then calculate the distance, in feet, from that coordinate to each of the track coordinates. Then, throw out the furthest 10%. Or, you could calculate the simple average and then calculate the Standard Deviation at the 2/3rds confidence level and throw out all points above that.

 

But, to get a real improved average, you'd have to let it track through an entire circuit of satellites and I don't think it would be enough better to warrant going to all that trouble. In the amount of time it would take to do that you could find a dozen caches.

[+zahadoom]

 

Zaha -

 

One thing you could do is turn on 1 second tracking at the time you turn on averaging then turn them both off at the same time. Averaging the track coordinates yourself should show you how the averaging is performed. The advantage with averaging the track yourself is that you could throw out the outliers. I'd bet a plug nickel that the Garmins are just providing a simple average.

 

If you average the track yourself, I would pick a point in the interior of the biggest cluster and then calculate the distance, in feet, from that coordinate to each of the track coordinates. Then, throw out the furthest 10%. Or, you could calculate the simple average and then calculate the Standard Deviation at the 2/3rds confidence level and throw out all points above that.

 

But, to get a real improved average, you'd have to let it track through an entire circuit of satellites and I don't think it would be enough better to warrant going to all that trouble. In the amount of time it would take to do that you could find a dozen caches.

 

Well I pick mine up this week. Ill let you know what I learn from the 60cx average screen.

 

I understand what your saying, but to me, and I just want to point out here.... that according to the 60cx manual on page #13 and 14 (shows a screen even)....

http://www.garmin.com/manuals/GPSMAP60Cx_OwnersManual.pdf ... that it will give you accuracy down to increments of .1 ft, and also tells you how many measurements it has taken.

 

I was only wondering, and asking people who own them, how low on the AVERAGE screen have they seen the unit report its accuracy.

 

Well to me, if the GPS math is good, which we know it is. As long as YOU the user can ENSURE that you have not MOVED the GPS on earth. That if you keep taking measurements that have a slight error, the more of them you take evenutally you could be pretty SURE of your location.

 

For me this isn't for looking for the cache, but to place them. Or if your really stumped it could help you if the hide'e took an accurate reading.

 

Thats all, just wondering on how the screen has worked for people and how low and close this has reached to a benchmark.

Edited April 22, 2006 by zahadoom

[planewood]

 

Well I pick mine up this week. Ill let you know what I learn from the 60cx average screen.

 

I understand what your saying, but to me, and I just want to point out here.... that according to the 60cx manual on page #13 and 14 (shows a screen even)....

http://www.garmin.com/manuals/GPSMAP60Cx_OwnersManual.pdf ... that it will give you accuracy down to increments of .1 ft, and also tells you how many measurements it has taken.

 

I was only wondering, and asking people who own them, how low on the AVERAGE screen have they seen the unit report its accuracy.

 

Well to me, if the GPS math is good, which we know it is. As long as YOU the user can ENSURE that you have not MOVED the GPS on earth. That if you keep taking measurements that have a slight error, the more of them you take evenutally you could be pretty SURE of your location.

 

For me this isn't for looking for the cache, but to place them. Or if your really stumped it could help you if the hide'e took an accurate reading.

 

Thats all, just wondering on how the screen has worked for people and how low and close this has reached to a benchmark.

 

Zaha - I'd take the 1/10 unit on the estimated accuracy with a grain of salt. I've seen 3' actual accuracy when it was reporting 15' estimated accuracy and 20' actual accuracy when it was reporting 7'.

 

Overall, the estimated accuracy it reports is directly proportional to actual accuracy. The great majority of the time you will be getting better actual accuracy if the estimated accuracy shown is low than if it is high. Basically, if you plot actual accuracy against estimated accuracy and fit a best-fit straight line through the cloud of points, the line will have a positive slope.

 

It's all a game of statistics and percentages!

 

7' is the lowest I've seen my 60cx report for estimated accuracy. But, I've gotten actual accuracies lower than that.

[+Team BlackZ]

I know this is a little older thread, but I just noticed last night that my 60CS wanders to when indoors using a Gillson Antenna hooked on a window. I was watching the GPS using Mapsource and it was all over the place. Every once in a while it would say I was moving 1.2 mph. This may have been discussed before using an EXT. antenna, but thought I would share my experience.

Edited April 28, 2006 by Team BlackZ

[+dumketu]

I sat my GPS 60csx and my older Garmin Etrex Vista together oustside for a while and watched the coordinates. Both "wandered around by about the same amount - 0.4 seconds of long/lat. The 60csx was reading 3-4 meter resolution the Vista was reading 11-17 meter resolution. The other difference was the track. With both set to auto normal, the Vista did not drop as many points as the 60csx and the 60csx had a much tighter grouping of the points dropped. The vista had a couple of really wayward points from when the unit first started acquiring. I did not see this with the 60csx.

 

Brian

[+dhbaird]

 

Above is my test yesterday imposed upon a Google Earth image. Test ran for 14 hours, 76CSX... Glisson External antenna. Full lock on 8-9 sateliites. WAAS enabled. No WAAS lock during the time I monitored. The GPSr said I had an error of +/- 9 feet.

 

It was sitting on top of a surveyed spot that is: N 41.54.697 W71.53.946 Altitude of 307.54 feet. Told me I was 45 feet from the BM. The average waypoint with some 5000 measurements was N 41.54.696 W71.53.947 Altitude of 318 feet.

 

The first half of the log file is below...

 

Hope this helps

D

 

 

1 4/28/2006 4:02:10 PM 82 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

2 4/28/2006 4:02:12 PM 99 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

3 4/28/2006 4:02:14 PM 113 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

4 4/28/2006 4:02:16 PM 127 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

5 4/28/2006 4:02:18 PM 138 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

6 4/28/2006 4:02:20 PM 152 ft 0 ft 0:00:02 0 mph 0° true N41 54.696 W71 53.947

7 4/28/2006 4:02:22 PM 164 ft 0 ft 0:00:03 0 mph 0° true N41 54.696 W71 53.947

8 4/28/2006 4:02:25 PM 175 ft 0 ft 0:00:03 0 mph 0° true N41 54.696 W71 53.947

9 4/28/2006 4:02:28 PM 186 ft 0 ft 0:00:03 0 mph 0° true N41 54.696 W71 53.947

10 4/28/2006 4:02:31 PM 201 ft 3 ft 0:00:03 0.6 mph 90° true N41 54.696 W71 53.947

11 4/28/2006 4:02:34 PM 211 ft 0 ft 0:00:04 0 mph 0° true N41 54.696 W71 53.947

12 4/28/2006 4:02:38 PM 225 ft 3 ft 0:00:04 0.5 mph 90° true N41 54.696 W71 53.947

13 4/28/2006 4:02:42 PM 237 ft 3 ft 0:00:04 0.5 mph 90° true N41 54.696 W71 53.946

14 4/28/2006 4:02:46 PM 248 ft 0 ft 0:00:03 0 mph 0° true N41 54.696 W71 53.945

15 4/28/2006 4:02:49 PM 258 ft 0 ft 0:00:06 0 mph 0° true N41 54.696 W71 53.945

16 4/28/2006 4:02:55 PM 268 ft 0 ft 0:00:05 0 mph 0° true N41 54.696 W71 53.945

17 4/28/2006 4:03:00 PM 279 ft 0 ft 0:00:05 0 mph 0° true N41 54.696 W71 53.945

18 4/28/2006 4:03:05 PM 289 ft 4 ft 0:00:10 0.3 mph 180° true N41 54.696 W71 53.945

19 4/28/2006 4:03:15 PM 298 ft 0 ft 0:00:04 0 mph 0° true N41 54.695 W71 53.945

20 4/28/2006 4:03:19 PM 303 ft 0 ft 0:00:07 0 mph 0° true N41 54.695 W71 53.945

21 4/28/2006 4:03:26 PM 307 ft 5 ft 0:00:34 0.1 mph 217° true N41 54.695 W71 53.945

22 4/28/2006 4:04:00 PM 311 ft 0 ft 0:00:04 0 mph 0° true N41 54.695 W71 53.946

23 4/28/2006 4:04:04 PM 324 ft 0 ft 0:00:01 0 mph 0° true N41 54.695 W71 53.946

24 4/28/2006 4:04:05 PM 319 ft 0 ft 0:01:30 0 mph 0° true N41 54.695 W71 53.946

25 4/28/2006 4:05:35 PM 313 ft 0 ft 0:00:33 0 mph 0° true N41 54.695 W71 53.946

26 4/28/2006 4:06:08 PM 317 ft 5 ft 0:17:14 0.0 mph 90° true N41 54.695 W71 53.946

27 4/28/2006 4:23:22 PM 306 ft 20 ft 0:09:51 0.0 mph 249° true N41 54.695 W71 53.945

28 4/28/2006 4:33:13 PM 317 ft 15 ft 0:25:24 0.0 mph 45° true N41 54.694 W71 53.949

29 4/28/2006 4:58:37 PM 327 ft 0 ft 0:06:15 0 mph 0° true N41 54.695 W71 53.947

30 4/28/2006 5:04:52 PM 316 ft 5 ft 0:06:38 0.0 mph 143° true N41 54.695 W71 53.947

31 4/28/2006 5:11:30 PM 317 ft 11 ft 1:15:36 0.0 mph 14° true N41 54.695 W71 53.946

32 4/28/2006 6:27:06 PM 307 ft 0 ft 0:23:13 0 mph 0° true N41 54.697 W71 53.945

33 4/28/2006 6:50:19 PM 297 ft 7 ft 0:26:06 0.0 mph 236° true N41 54.697 W71 53.945

34 4/28/2006 7:16:25 PM 287 ft 7 ft 0:16:02 0.0 mph 124° true N41 54.696 W71 53.947

35 4/28/2006 7:32:27 PM 277 ft 7 ft 0:29:18 0.0 mph 0° true N41 54.695 W71 53.945

36 4/28/2006 8:01:45 PM 287 ft 3 ft 0:01:20 0.0 mph 90° true N41 54.697 W71 53.945

37 4/28/2006 8:03:05 PM 284 ft 3 ft 0:03:17 0.0 mph 90° true N41 54.697 W71 53.945

38 4/28/2006 8:06:22 PM 284 ft 3 ft 0:01:01 0.0 mph 270° true N41 54.697 W71 53.944

39 4/28/2006 8:07:23 PM 286 ft 4 ft 0:06:04 0.0 mph 0° true N41 54.697 W71 53.945

40 4/28/2006 8:13:27 PM 285 ft 8 ft 0:09:38 0.0 mph 270° true N41 54.697 W71 53.945

41 4/28/2006 8:23:05 PM 295 ft 8 ft 0:36:44 0.0 mph 20° true N41 54.697 W71 53.947

42 4/28/2006 8:59:49 PM 300 ft 4 ft 0:02:27 0.0 mph 180° true N41 54.698 W71 53.946

43 4/28/2006 9:02:16 PM 299 ft 5 ft 0:05:35 0.0 mph 144° true N41 54.698 W71 53.946

44 4/28/2006 9:07:51 PM 297 ft 3 ft 0:05:21 0.0 mph 270° true N41 54.697 W71 53.945

45 4/28/2006 9:13:12 PM 295 ft 3 ft 0:05:05 0.0 mph 90° true N41 54.697 W71 53.946

46 4/28/2006 9:18:17 PM 297 ft 11 ft 0:04:08 0.0 mph 180° true N41 54.697 W71 53.945

47 4/28/2006 9:22:25 PM 298 ft 4 ft 0:00:57 0.0 mph 0° true N41 54.695 W71 53.945

48 4/28/2006 9:23:22 PM 300 ft 5 ft 0:00:49 0.1 mph 37° true N41 54.696 W71 53.945

49 4/28/2006 9:24:11 PM 301 ft 0 ft 0:00:26 0 mph 0° true N41 54.697 W71 53.945

50 4/28/2006 9:24:37 PM 302 ft 0 ft 0:00:50 0 mph 0° true N41 54.697 W71 53.945

51 4/28/2006 9:25:27 PM 303 ft 0 ft 0:00:31 0 mph 0° true N41 54.697 W71 53.945

52 4/28/2006 9:25:58 PM 304 ft 0 ft 0:00:26 0 mph 0° true N41 54.697 W71 53.945

53 4/28/2006 9:26:24 PM 304 ft 5 ft 0:00:34 0.1 mph 217° true N41 54.697 W71 53.945

54 4/28/2006 9:26:58 PM 300 ft 0 ft 0:00:39 0 mph 0° true N41 54.696 W71 53.945

55 4/28/2006 9:27:37 PM 302 ft 0 ft 0:00:24 0 mph 0° true N41 54.696 W71 53.945

56 4/28/2006 9:28:01 PM 301 ft 0 ft 0:00:27 0 mph 0° true N41 54.696 W71 53.945

57 4/28/2006 9:28:28 PM 301 ft 0 ft 0:00:34 0 mph 0° true N41 54.696 W71 53.945

58 4/28/2006 9:29:02 PM 301 ft 0 ft 0:00:50 0 mph 0° true N41 54.696 W71 53.945

59 4/28/2006 9:29:52 PM 302 ft 0 ft 0:00:37 0 mph 0° true N41 54.696 W71 53.945

60 4/28/2006 9:30:29 PM 301 ft 4 ft 0:00:48 0.1 mph 0° true N41 54.696 W71 53.945

61 4/28/2006 9:31:17 PM 299 ft 0 ft 0:00:37 0 mph 0° true N41 54.697 W71 53.945

62 4/28/2006 9:31:54 PM 301 ft 3 ft 0:00:31 0.1 mph 270° true N41 54.697 W71 53.945

63 4/28/2006 9:32:25 PM 302 ft 3 ft 0:01:16 0.0 mph 90° true N41 54.697 W71 53.946

64 4/28/2006 9:33:41 PM 303 ft 5 ft 0:01:07 0.0 mph 217° true N41 54.697 W71 53.945

65 4/28/2006 9:34:48 PM 303 ft 0 ft 0:14:50 0 mph 0° true N41 54.696 W71 53.946

66 4/28/2006 9:49:38 PM 292 ft 7 ft 0:16:16 0.0 mph 180° true N41 54.696 W71 53.946

67 4/28/2006 10:05:54 PM 302 ft 5 ft 0:02:03 0.0 mph 37° true N41 54.695 W71 53.946

68 4/28/2006 10:07:57 PM 301 ft 7 ft 0:09:07 0.0 mph 304° true N41 54.695 W71 53.945

69 4/28/2006 10:17:04 PM 302 ft 9 ft 0:05:21 0.0 mph 143° true N41 54.696 W71 53.947

70 4/28/2006 10:22:25 PM 305 ft 7 ft 0:29:20 0.0 mph 304° true N41 54.695 W71 53.945

71 4/28/2006 10:51:45 PM 310 ft 7 ft 0:05:03 0.0 mph 56° true N41 54.695 W71 53.947

72 4/28/2006 10:56:48 PM 313 ft 8 ft 0:10:54 0.0 mph 20° true N41 54.696 W71 53.945

73 4/28/2006 11:07:42 PM 303 ft 3 ft 0:01:22 0.0 mph 270° true N41 54.697 W71 53.945

74 4/28/2006 11:09:04 PM 305 ft 11 ft 0:04:22 0.0 mph 0° true N41 54.697 W71 53.945

75 4/28/2006 11:13:26 PM 312 ft 3 ft 0:01:38 0.0 mph 270° true N41 54.699 W71 53.945

76 4/28/2006 11:15:04 PM 314 ft 11 ft 0:02:17 0.1 mph 166° true N41 54.699 W71 53.946

77 4/28/2006 11:17:21 PM 310 ft 11 ft 0:09:06 0.0 mph 48° true N41 54.697 W71 53.945

78 4/28/2006 11:26:27 PM 313 ft 7 ft 0:00:34 0.1 mph 236° true N41 54.698 W71 53.944

79 4/28/2006 11:27:01 PM 313 ft 5 ft 0:08:09 0.0 mph 323° true N41 54.698 W71 53.945

80 4/28/2006 11:35:10 PM 316 ft 11 ft 0:13:34 0.0 mph 180° true N41 54.698 W71 53.945

81 4/28/2006 11:48:44 PM 313 ft 5 ft 0:08:27 0.0 mph 90° true N41 54.697 W71 53.945

 

 

So anybody up for that test I was talking about a few posts up there.... Post #25.

 

Take the 60Cx or 60Csx to a known benchmark, position it vertically, set it on AVERAGE mode/screen for at minimum of 5 mins(or longer if you want), record starting co-ords', and then final as well as the posted accuracy and length of test.

 

And of course the benchmark co-ords, so we can compare if the AVERAGE mode can actually improve accuracy.

 

Should be some interesting results.

[gps_lover]

My GPSMAP60CSx (purchased in late March) frequently suffered from

a wandering problem, too. This occurred on several occasions, even

handheld outdoors in the clear, with many strong satellite signals indicated.

 

Typical initial symptom was a gradual deviation of the track from true,

by amounts up to several hundred feet, followed by increasingly bizarre

behavior in which the track log showed movements totally unrelated to

my actual movements. On two occasions I even saw the directional arrow

moving flat-out sideways! I also have seen speed indications as high as

4.7 mph when standing still.

 

I have eliminated several of the usual suspects by being sure that WAAS

and "lock on road" were turned off.

 

I got little help from Garmin; their only suggestion was to get the latest

software. But their release notes said nothing about addressing this sort

of problem, and I did not want to take the risks associated with a failed

download, so I did not go this route.

 

In time I developed a strong suspicion that there is something wrong with

the way the engine (SiRF Star III) was working, so I returned my unit for

a refund while I still had a chance.

 

My unit came with Garmin software version 2.50. I stupidly did not note the

SiRF receiver software version- I sure wish I had thought to do so. Sorry about

that!

 

Does anyone know if the SiRF receiver software is user-upgradeable too, or

just the Garmin part?

[+zahadoom]

I find that it always wanders (if north is up) to the NNW...

 

I also tryed out averaging for the first time yesterday and wasnt that impressed..

 

After about 50 measurements the unit was down to 3.9meters of accuracy. Then it started going up, went as high as 4.4m and then started to drop again, I stoped at about 4.0 and 80 measurements..

 

And after testing out that waymark, its really not 100% on... its within the reported accuracy though...

 

I just think the wandering issue still continues while averaging and it ruins those calulations because its moving when it shouldnt be.