Jump to content

Colorado location errors


Recommended Posts

Here's another image from the gallery.

 

It superimposes tracks from three different sources:

 

60cs: Green

Colorado w/o backlight: Blue

Colorado w/backlight: Red

 

The image shows about 7-9 different tracks from each source over as many days.

 

GO$Rs

 

colorado%20tracklog.jpg

 

Wow. It certainly looks like the blue line (CO w/o backlight) is tighter than the green (60cs). I thought the only difference b/w the 60cs and 60csx was the removable sd card?

 

That red line is crazy! I wonder if you just have a defective unit? I'm no expert, but it would seem to me that your problem might be caused by the wires for your backlight and antenna being either too close or not properly shielded.

Link to comment

Wow. It certainly looks like the blue line (CO w/o backlight) is tighter than the green (60cs). I thought the only difference b/w the 60cs and 60csx was the removable sd card?

 

No! The 60csx has a high sensitivity Sirf GPS. The 60cs is a prior generation unit, very different.

 

That red line is crazy! I wonder if you just have a defective unit? I'm no expert, but it would seem to me that your problem might be caused by the wires for your backlight and antenna being either too close or not properly shielded.

 

It is a possibility that I have a defective unit which is why I was posting the issue here and trying to see if others were seeing the same or similar results with the backlight cranked up.

 

GO$Rs

Edited by g-o-cashers
Link to comment

Not sure of multiple tracks over many days is a valid test. Different satellite positions, slightly different paths, etc could cause variance. The ones where you carry two different GPSRs simultaneously are far more telling in my opinion.

 

Agreed but I do think with enough samples you can start to see a trend. If I had a 60csx or some other decent reference the results would be better but I do think that pictures show how severe the tracking error can be on the Colorado. The fact that I never had any significant error in the past week with backlight off says something. There is actually a minor error in the blue track to the east of the waypoint "kiosk" but its the kind of stuff I can live with, certainly not the level of the errors in the red track over multiple days.

 

GO$Rs

Link to comment

GO$Rs, I've been paying attention to your thread and have looked over my tracks that I've recorded side-by-side with my 300 and 60Cx. I have to say my 300 records a better more consistent track, regardless of using backlighting or not. My first 300, and now my replacement 300 wer/are better at recording tracks thatn my 60Cx.

 

I have to believe your unit is faulty. I'll post some tracks later to show you, gotta run to my kids concert.

Link to comment

I also am unable to duplicate this problem. I just finished a systematic test.

 

Compass off

WAAS off

Tones off

Altimeter on in variable altitude mode (no setting in GO$Rs post)

Track up

Backlight limiting off

Displaying map continuously, with EPE and signal strength data fields

GPSr hanging from a belt loop for consistency. Eyeballed data fields about 6 times per loop.

 

Reasonably fresh 2700 mAh MAHA NiMH (3 bars to start, 2 at the end).

 

Running 2.40/2.60 rather than 2.51 beta. I do software for a living, so I avoid other people's betas when possible.

 

Tested in suburban area with moderate tree cover. Good constallation. The course has both sharp corners and long, sweeping curves. Walked on sidewalks, so the tracks on the ground are repeatable to within 18 inches.

 

EPE settled out at 11-12 feet standing still, both at the start and at the end of the test. EPE generally ranged from 14-18 feet while moving. Saw one spike to 24 feet with backlight off. Never saw anything higher than 18 feet with backlight on. Signal strength varied between 4 and 5 bars while moving.

 

Walked a .4 mile loop, 6 times in the space of 45 minutes. First three loops with backlight timeout at 15 seconds, dialed all the way down. Second three loops with backlight on continuously, dialed all the way up.

 

Although I have not uploaded the tracks for analysis, I can already tell you that there were no significant differences between the tracks with backlight off and on. All six tracks are tightly grouped. I see a couple of individual deviations, but nothing which exceeds what's reasonable for the observed EPE. Eyeballing the map on the unit at 50 foot scale, I don't see any deviation (from the tight grouping) of more than about 20 feet.

 

If anybody has tested this with 2.51b/2.60, we'd all be interested in seeing the results...

Link to comment

Thanks Timpat and twolpert. I'm actually hoping it is my unit especially after the test I ran tonight -- a direct comparison with my 60cs. I changed one thing from my previous experiments, I ran the tracklog on both units to gather a sample every 3 seconds to get a more uniform track. My altimeter is in VA mode as well.

 

As I was running the test with backlight off I ran into the location error that I haven't seen for over a week. I don't believe that running the tracklog at a higher frequency would have enough impact on power that it would trigger the issue, so its probably that I've just been getting lucky or that it is just more sensitive with the backlight up.

 

The error I saw tonight was on the smaller side of what I've seen in the past -- I was about 150' off, but the image below shows it pretty clearly, CO is red and the 60cs is black. The track runs counter-clockwise starting and ending in the northeast corner. On the return trip you can see where the CO started to get in trouble and got stuck 150' south of the 60cs track. In this last .3m of the trip (which is wide open) I had 15' EPE on my 60cs and my CO was showing 30' of EPE.

 

Let's hope my CO is bad.

 

080512p.jpg

 

GO$Rs

Link to comment

This may not be very scientific but here is a map showing part of a hike my son and I took this weekend. It was in the Adirondacks in forest with hemlock, pine, and various hardwood trees with leaves coming on, under constant canopy and no open sky at all. Both units were hooked to my backpack straps.

 

We had our Map60Cx (Blue tracks) and Colorado 300 (Red tracks). Both had WAAS enabled and getting corrections continuously. Compass off on the 300, backlighting (which we used) set to 30 seconds. (No backlight on the 60Cx-none ever needed!). Track recording settings were the same at Distance Auto and Most Often. Running Topo 2008 on both viewing the map screens constantly and switching all the time to the Trip Odometer data screens.

 

Look here for my Track Map. The length of this sample is about 1.3 mi. This shows each unit track going up and back, and there was a time lapse of about 4 hrs in between. The color pairs of tracks trend together, with a general separation of about 40-50'. I wonder why this is? (While standing still both show identical coordinates).

 

I think for heavy tree cover my two units behaved fairly alike, and the tracks were generally consistent with no great overliers. I feel the 300 records a cleaner track in most cases.

Link to comment

Celtic Cacher,

 

Do you use the backlight on both of your CO's? If so, would be possible to run one with backlight and one without for a few days to see how they compare?

 

GO$Rs

 

I used your settings except for the backlight and ran one with and one without. Same batteries in both, freshly charged. We went out for two caches for just over a mile and a half walk. There were no problems and the tracks were pretty much dead on with each other. This is different for us in that both units were pretty close with each other at GZ.

Link to comment

I'm using 2700mAh Maya NiMH's, if these don't work I'm not sure what would!

 

Thanks Timpat and Celtic Cacher for posting your results. Definitely feels like an issue with my unit although it bugs me that people with the eTrex HCx's and some CO's are seeing similar problems. Maybe it is something that Garmin has fixed or improved in the newer units.

 

GO$Rs

Link to comment

Last weekend I was caching with a large group and there were two of us with Colorado 400T's. A bunch of other folks had some 60csx's (and other units). After a half mile bushwhack to a cache the guys with the 60's stopped and started looking for the cache and I asked what they were doing since my Colorado still showed another 180 feet to ground zero. As I said this they pulled the cache from about 5 feet where I was standing.

 

At this point I looked at the other Colorado in the group and seen her's was just as far off as mine and we were both getting an EPE of about 80 feet. Power cycling the units fixed the problem. Since paying closer attention to this issue I see it happening more often. I also get more erratic behavior when moving slow. For example, the compass needle will be pointing away from the cache while the distance will be counting down as I walk closer. I always keep the electronic compass off. Sometimes the distance does not move either requiring a power cycle.

 

In the examples above both Colorado's are running the beta software (2.51) and GPS 2.6. My back-light was set on with 30 second shutoff, not sure about hers. Electronic compass off, waas on.

 

I am a huge Colorado supporter and have had it since day 1 when it hit the REI in Michigan. However this location problem has to be fixed soon as it is a problem I can't live with. I'll be calling Garmin today.

 

Edited to add: I use regular alkaline batteries and the other user had the low discharge rechargeables in her unit.

Edited by TeamLegend4
Link to comment

I also get more erratic behavior when moving slow. For example, the compass needle will be pointing away from the cache while the distance will be counting down as I walk closer. I always keep the electronic compass off. Sometimes the distance does not move either requiring a power cycle.

 

Turn off the compass and your accuracy will improve when moving slowly.

Link to comment

Last weekend I was caching with a large group and there were two of us with Colorado 400T's. A bunch of other folks had some 60csx's (and other units). After a half mile bushwhack to a cache the guys with the 60's stopped and started looking for the cache and I asked what they were doing since my Colorado still showed another 180 feet to ground zero. As I said this they pulled the cache from about 5 feet where I was standing.

 

Sounds exactly like what I have seen. We probably have similar vintage 400t's since I bought mine at REI the first day it was available, mine S/N is 18Z001xxx. I'm beginning to think that these early units may have this problem, issues with battery life and the loss of time when powered off issue to varying degrees. I wouldn't be surprised if all of these problems are somehow related to a power regulation problem in the CO hardware. Mine's going back.

 

GO$Rs

Link to comment

I am also seeing larger error margins wih the backlight on. In addiiton I have to clibrate the compass each and every time the unit is turned on to ensure accuracy. I have always run with the compass on so will now try with it off permanently. other settings:

- Backlight: on full (no auto shutoff)

- WAAS: onn

- Compass: on

- Orientation: Track Up

- Tones: off

- Visible page: Map

Link to comment

I also get more erratic behavior when moving slow. For example, the compass needle will be pointing away from the cache while the distance will be counting down as I walk closer. I always keep the electronic compass off. Sometimes the distance does not move either requiring a power cycle.

 

Turn off the compass and your accuracy will improve when moving slowly.

As I stated the compass is and always has been off. I've never had success using the compass on a Garmin unit.

Link to comment

Last weekend I was caching with a large group and there were two of us with Colorado 400T's. A bunch of other folks had some 60csx's (and other units). After a half mile bushwhack to a cache the guys with the 60's stopped and started looking for the cache and I asked what they were doing since my Colorado still showed another 180 feet to ground zero. As I said this they pulled the cache from about 5 feet where I was standing.

 

Sounds exactly like what I have seen. We probably have similar vintage 400t's since I bought mine at REI the first day it was available, mine S/N is 18Z001xxx. I'm beginning to think that these early units may have this problem, issues with battery life and the loss of time when powered off issue to varying degrees. I wouldn't be surprised if all of these problems are somehow related to a power regulation problem in the CO hardware. Mine's going back.

 

GO$Rs

 

Yup, mine is an early unit as well S/N 18Z001xxx. It was the first one purchased at the REI in Troy, MI. I didn't get the time to contact Garmin today but I'm sure this one is going back.

Link to comment

I did a simple test yesterday with my Colorado 300 and 60CSx. I installed them side by side on my bicycle and ran a path 5 times at the same very slow speed. The test took about 25 minutes. Before start I let them find satellites for 15 minutes. During the test 60CSx had an EPE between 2-3 meter (without showing Ds on satellite bars), Colorado had Ds (see picture) but EPE 5-10 meter. Both units was WAAS/Egnos enabled, no compass, max backlight, fresh batteries, showing the satellite screen and so on. The terrain is a perfectly flat open field with very good satellite reception. My Colorado (serial: 169006xxx) is bought in Sweden and use the beta software 2.51 and shows more or less the same problems that other has described in the forum.

 

The result:

tracktest.jpg

669.jpg

Link to comment

and shows more or less the same problems that other has described in the forum.

--------------

Sorry, but where is this problem ? From your tracks I see that your GPSr keeps within 5mtr EPE.

 

Here you see calculation of EPE 4mtr:

 

50cepee7.gif

 

For EPE 4mtr it means, that your position is with probability 99% within 10mtr on BOTH sides when you move, or 10mtr around you if you stay. And 50% probability of position within 4mtr.

You can calculate for 5mtr (it is 12mtr around).

 

Christopher.

Edited by sokolo0
Link to comment

Mine has been fairly accurate however, I did notice it jump around a few times. Last night did 3 cahes. I found all three. However, on the last one of the night my 400T lead me a spot that said I was with in 2 feet of the cache. I looked for a few minutes and nothing. There was no tree cover in this area. It was next to a river with rocks on the edge. I was standing on the rocks looking underneath some of the large openings.

 

I then decided to walk about 4feet back to the sidewalk, and go further down the side walk and come back. The GPS lead me back to the rocky area next to the water. Looked another minute. Then I decided to go back to the sidwalk and turn the back light off. My GPS then lead me 180 degrees the other direction from the water. So the waters edge is south and now my GPS led me North. I went about 30 feet and my 400T zeroed me in real close to the cache and I spotted it. Now these are not huge distances to be off but I may have not found the cache had I not read this forum before hand and new to try a couple things with the GPS. I also can't remember or say for sure if this has not happened before with my other GPS (etrex Legend). It seems to me that in the past I always remember that some times the things just loose position and have to settle a bit. What was weird is as soon as I turned the backlight off I got a new position and found my prize.

 

400T

compass off

WAAS off

Geocaching profile

3 maps on

NIMH 2400-2700 (don't remember which set is in)

Link to comment

Sorry, but where is this problem ? From your tracks I see that your GPSr keeps within 5mtr EPE.

My point was to compare the Colorado and 60CSx side by side under the same conditions with the background light switched on. The question is why the track from Colorado is not as good as the track from 60CSx. I know from other tests I made (undocumented) that if I switched off the light in Colorado the tracks are approximately as from 60CSx. So it seems to be something strange with the light.

 

Another strange thing is that occassionaly my Colorado is 80-100 meter off track (with EPE around 6 meter). The only way to have it work normally again is to restart the unit. Very annoying when Geocaching and even worse navigating in the archipelago by boat.

 

Bengt

Link to comment

Kaffekopp:

The question is why the track from Colorado is not as good as the track from 60CSx. I know from other tests I made (undocumented) that if I switched off the light in Colorado the tracks are approximately as from 60CSx. So it seems to be something strange with the light.

 

Possibly voltage problem. I use sometimes Colorado connected to car 12 - 14V. Screen is fantastic ! Very bright. Seems it is prepared for higher voltage, than from 2 AA batteries. And Colorado is very similar in many functions to Nuvi, as I understand from discussions on forums.

Christopher

Link to comment

<snip>

I'm beginning to think that these early units may have this problem, issues with battery life and the loss of time when powered off issue to varying degrees. I wouldn't be surprised if all of these problems are somehow related to a power regulation problem in the CO hardware. Mine's going back.

 

GO$Rs

Just as a point of information, can you confirm that the unit had this issue before the upgrade to 2.51b? Since there was no chip code included, I doubt that 2.51b affected the actual position accuracy. But it seems like they're always tweaking the power management stuff.

Link to comment

Was out today on Mt Monadnock NH today and the location error appeared again. With a 360 degree view, no trees and a 1500 foot elevation rise over nearby terrain my 400t had a GZ of .14 mi. and the 60CS was 120 feet from the cache. Wass and backlight were on. This has happened to this unit at least two times before. After shutting it down it had 119 feet. I have an RMA and will also be returning this early unit(3XXX).

 

Anniebananie

Link to comment

Some days ago, we were on a hike with 2 CO300 (V 2.51) and 1 60CSx.

It was a walk in the woods, but the reception was generally OK.

 

For the first view kilometers all units had their normal deviation from the "real" world.

At a certain moment when walking at a straight road, both CO300 started to deviate from the road in the same way, more and more, up to 85 meter. The 60 CSx still followed the road (as expected)

 

We compared the 3 units intensively, for at least 10 minutes.

The deviation did increase until this 85 meter and stayed that large.

At the end of the straight track, we turned left (getting some other sats, I think).

First my CO jumped back on the road and half a minute later the other one did.

 

The MT3318 is said to be the best available GPS chipset at the moment. If that is true, the management software around it has some severe issues.

... or Garmin can't handle it. Wouldn't surprise me when I see the issues of the CO.

 

Cheers.

Link to comment

Sounds similar to the issues I've seen. Did you have the tracklogs running on all 3 units? If so would it be possible to post a link to an image with the tracklogs overlaid on one another?

 

How long did it take for the CO's to snap back into position? In some cases I can walk for 10-20 minutes with an open view of the sky and the error still remains.

 

I had my fourth conversation with Garmin on this issue yesterday and they've thrown up their hands and told me to return the unit (which I was planning to do under another RMA anyway). They claim they haven't heard of the problem, so if you are seeing it please give Garmin a call and let them know. My fear is the next unit is going to behave the same way, especially since Vista HCx users are seeing the same issue. Seems to be related to the chipset and it's software.

 

GO$Rs

Edited by g-o-cashers
Link to comment

i had off-position from more than 150-200 meters (version 2.51better)- waiting up to 5 minutes did not solve the poblem.

after power off/on it was under 10 meters

seems the gps-fw is still under construction

i did not inform garmin because their answer (text module) is annoying.

Edited by freeday
Link to comment

I have not seen any strange accuracy problems. Backlight on or off. I do mapping and closely examine my tracklogs.

 

I agree. I have been using mine in the car, on foot backlight on AND off, batteries and external power and I am seeing none of these errors. I DID cache all weekend with my friend who has a 60CSX and his ALWAYS read further away from the cache and as we got closer the gap shortened, but I was standing on top of the cache while his still said to go 12-20 feet further.... so take from that what you will.

Link to comment

I've noticed that the HCX Etrex models,according to another thread,have been suffering accuracy problems as well since the latest firmware upadate. It gets a bit confusing trying to determine if this is a chipset problem in both the Etrex and Colorado's,as it seems the Etrex's were not having accuracy problems prior to the update and have the same chipset. Anyone have a way of explaining this further,considering both models apparently have the same chipset.

Link to comment

I've noticed that the HCX Etrex models,according to another thread,have been suffering accuracy problems as well since the latest firmware upadate. It gets a bit confusing trying to determine if this is a chipset problem in both the Etrex and Colorado's,as it seems the Etrex's were not having accuracy problems prior to the update and have the same chipset. Anyone have a way of explaining this further,considering both models apparently have the same chipset.

 

I'm not as familiar with the history of the HCx GPS software updates but on the Colorado we had GPS Software 2.5 on 1/20 when the unit first came out and then it was upgraded to 2.6 when Software 2.3 came out on 1/23. Given that there weren't many Colorado owners during that first week or two and there were so many other bigger issues I'm guessing no one noticed the transition to GPS Software 2.6 on the Colorado and the changes associated with it.

 

GO$Rs

Edited by g-o-cashers
Link to comment

Sounds similar to the issues I've seen. Did you have the tracklogs running on all 3 units? If so would it be possible to post a link to an image with the tracklogs overlaid on one another?

 

I allways have a tracklog running (unless my CO shuts it down without asking me, of course. happens all the time). Here it is:

 

1fcc0688-e74f-4bd7-b5a5-9ae48a4ea5b7.jpg

 

The long straight line with the word baan along is the road we were walking on.

The 2nd CO was off in the same direction and distance.

The 60CSx was exactly on the road

 

After the CO decided to behave normally again, position was restored within half a minute. Before that is had been wrong for 10 minutes or so.

 

I will write Garmin on this and show the tracklog. Perhaps this will reactivate some memory cells with them.

 

Cheers,

Stormy

Link to comment

I had my fourth conversation with Garmin on this issue yesterday and they've thrown up their hands and told me to return the unit (which I was planning to do under another RMA anyway). They claim they haven't heard of the problem, so if you are seeing it please give Garmin a call and let them know.

GO$Rs

 

I know I complained three times and only got the standard replies such as, calibrate, reset, auto-locate, etc. The rep I spoke to each time didn't really seem very knowledgeable. finally I just told them to send me a new unit which they did. I received it yesterday with software version 2.5 (not the 2.5.1 beta) and GPS 2.6. I will watch it closely to see if the problem still exists. My guess is it will. :D

 

This unit is going to end up going back as well though as the fit for the battery cover is really awful. I even tried the back from the first Colorado and still got the same fit issue so it is not the cover piece that is the problem it is the unit itself. The back slides off extremely easy, not that nice tight fit I was used to and the gap near the top is a joke, water will pour into this one. :D

 

2510115936_598d2c5dd4_m.jpg

Edited by TeamLegend4
Link to comment

I had an interesting mountain bike ride this afternoon. I had both my 60CSx and Colorado 300 in handlebar mounts. As you can see below, both units were tracking fine from the trailhead. I rode a counter-clockwise loop. Before long, my Colorado's track (in blue) was straying. When I noticed it, I powered the unit off and back on, which corrected the problem for the remainder of the ride. Sorry I cut the scale out. At its worse, right before power cycling, the tracks are off by about 430'.

 

Colorado-chipset-issue.jpg

 

I'm running beta firmware 2.51 and chipset firmware 2.6.

 

There are many Colorado threads here, and it's hard to keep up with them all. What's the consensus on this issue? Is it chipset related and can it be fixed with new firmware? Should I send the unit back?

Link to comment

There are many Colorado threads here, and it's hard to keep up with them all. What's the consensus on this issue? Is it chipset related and can it be fixed with new firmware? Should I send the unit back?

 

Welcome to the club Rich! This has the exact same signature as the problems that I've seen. I don't think anyone has a fix or knows what the issue is, including Garmin. I'm sending my unit back but I'm skeptical it is going to behave any differently because of this thread.

 

HCx owners (same MediaTek chipset as the Colorado) are seeing this same problem. It seems to have showed up when the 2.6 GPS firmware update was introduced (again same firmware as the Colorado). The problem they describe on the HCx sounds identical to the issue being discussed on this thread. That would suggest it is a chipset or chipset firmware issue.

 

Please call Garmin with your data! I haven't figured out a way to convince them it is an issue and I've had 4-5 calls, provided photos, tracks and exact configurations but so far the best they can come up with is "replace the unit".

 

GO$Rs

Link to comment

I've had my 400t for about two weeks and though I was able to find a few caches with it. I am not pleased with its accuracy for geocaching. In one case I was initially close to the cache but suddenly the compass pointed in another direction 30 feet away. I followed the compass and when I got to the other point after a while it pointed back to close to where I was originally. I eventually found the cache but it was not due to the Colorado's performance. This happened several times. Today I gathered up all my gps devices and went to the middle of a football field where I took 4 readings with the Colorado, a 350 Nuvi, A Dell Axim running CachNav with a Navibe Bluetooth GPSr, and a Garmin Forerunner 305. The non Colorado units all use the Sirf III chipset. All were reading about 12 to 15 feet accuracy except the CachNav which read about 26 feet. I marked a waypoint with all four units and averaged the coordinates. I then went to the four corners of the field with the Colorado and navigated to the averaged waypoint in Recreation mode using the compass which I calibrated. The compass was pointing in the right direction and the distance seemed to be about right at 150 ft or so. I then followed the compass pointer and walked toward the waypoint. In all four cases the compass went crazy when I got to about 30 feet or closer to ground zero. As I walked past the way point and away from it following the compass needle, at about 45 feet or further the bearing and compass begin pointing in the direction of the waypoint and the distance was also about right. The GPS accuracy was about 12 to 15 feet throught the test. I ran the same test with the CachNav and Bluetooth reciever. The bearing and distance got a little closer before the varibility started at about 15 to 20 feet. The instability of the Colorado compass when close in will drive you nuts. When I'm about 50 to 60 feet away, I now will try to pace off the distance and bearing when I can to see if that helps. I still am hoping that Garmin will fix the compass and heading instability issue in upcoming software revisions but this may be a hardware issue. I like the Colorado but the accuracy for geocaching needs Garmin's attention.

 

While at the football field, I zeroed the Trip Computer and cleared the current track of the Colorado. I also cleared the Forerunner 305 trip computer. I then walked around the footbball field at a slow pace. When I stopped the 305 and the Colorado trip computers read the same which seemed about right and the Colorado track was a little higher due to rounding. Thats pretty good news. However, I had noticed earlier that if you stopped and the Colorado was still on, the trip computer and the track continued to increase. I had stopped for 30 mins and the trip computer had gone up about 1 mile and the current track had gone up about 2 miles. I assume if you turn it off when you stop for a break or lunch, that it would start back up where it left off. I thought I'd report what I found recently.

 

After using my CO 400t for the last two weeks I've been very pleased with the GPSr itself. I've seen no issues with grabbing and holding lock, EPE has always been sub 20', normally in the teens with a minimum of 5-6 satellites at good strength. I've been walking the same trail over and over since getting the CO and the track logs for each of these are right on top of each other (within 10-20'). So far so good, until this morning.

 

On the walk out everything looked fine. I went off onto some new trails and starting marking waypoints until I came back to the original trail. I noticed at that point my position didn't seem to match earlier walks but I went a little further to an intersection and I could see that my position on the map was way off based on the old tracks.

 

When I looked at the satellite page I was surprised to see the EPE in the 70-80' range. Based on my previous tracks I would estimate my position error was off by more than 300' ! At that point I stopped for a few minutes to see if things would settle out but there was no change. I had good strong lock on 7 satellites with one directly above, 3 at ~45 degrees and 2-3 closer to the horizon. I put the GPS in demo mode several times to see if that would bring it out of its funk but I got exactly the same results. Finally I power cycled the GPS and when it came back on I had EPE of 18 and the constellation/strength readings looked identical to me. My position on the map relative to previous tracks was exactly where I thought I should have been.

 

These issues will drive you nuts searching for a cache. Sli23sli

 

GO$Rs

Link to comment

Did you try the same experiment with the electronic compass turned off? The electronic compass introduces a entirely new set of potential issues. I would suggest turning it off to see how the unit behaves.

 

Most of the issues described in this thread are with the compass turned off, and while the stability of the compass in the Colorado may be an issue it shouldn't be confused with the accuracy of the GPS.

 

GO$Rs

Link to comment

Maybe I'm missing something but it seems to me the GPS receiver in the Colorado is controlling the compass pointer to the waypoint so the intergrated accuracy of the compass/gps combination is the issue

 

In any event, I went back and redid the experiment with the compass turned off. I installed Distance and Heading in Degrees as my datafields on the map page. I used a compass to navigate to the waypoint and I stepped off the distance. I had even more confusing results than with the compass on. I occationally got to the waypoint by this method but most of the time it would take me around in random paths 30 to 50 feet away from the waypoint. Sometimes I was following the heading and the distance was increasing so I turned around and the heading would change and be in the correct direction. It seems to me the heading to the waypoint should be independent of what direction you are holding the Colorado. Another way to navigate to the waypoint is to try to follow the course line to the waypoint and turn so the position pointer is going directly toward the waypoint. That seemed to work better than using the heading and a compass but its slower since you have to turn the Colorado and wait for it to turn the screen to assure you are going in the right direction. The GPS accuracy was in the 15 foot range during this test.

 

I turned the compass back on and used it and was able to consistantly get close to the waypoint by stepping off distance in the direction of the compass when I was in the 40 to 60 feet range. I was hoping that the Colorado would give more consistant results than this. Sli23sli

 

 

Did you try the same experiment with the electronic compass turned off? The electronic compass introduces a entirely new set of potential issues. I would suggest turning it off to see how the unit behaves.

 

Most of the issues described in this thread are with the compass turned off, and while the stability of the compass in the Colorado may be an issue it shouldn't be confused with the accuracy of the GPS.

 

GO$Rs

Link to comment

Maybe I'm missing something but it seems to me the GPS receiver in the Colorado is controlling the compass pointer to the waypoint so the intergrated accuracy of the compass/gps combination is the issue

 

In any event, I went back and redid the experiment with the compass turned off. I installed Distance and Heading in Degrees as my datafields on the map page. I used a compass to navigate to the waypoint and I stepped off the distance. I had even more confusing results than with the compass on. I occationally got to the waypoint by this method but most of the time it would take me around in random paths 30 to 50 feet away from the waypoint. Sometimes I was following the heading and the distance was increasing so I turned around and the heading would change and be in the correct direction. It seems to me the heading to the waypoint should be independent of what direction you are holding the Colorado. Another way to navigate to the waypoint is to try to follow the course line to the waypoint and turn so the position pointer is going directly toward the waypoint. That seemed to work better than using the heading and a compass but its slower since you have to turn the Colorado and wait for it to turn the screen to assure you are going in the right direction. The GPS accuracy was in the 15 foot range during this test.

 

I turned the compass back on and used it and was able to consistantly get close to the waypoint by stepping off distance in the direction of the compass when I was in the 40 to 60 feet range. I was hoping that the Colorado would give more consistant results than this. Sli23sli

 

 

Did you try the same experiment with the electronic compass turned off? The electronic compass introduces a entirely new set of potential issues. I would suggest turning it off to see how the unit behaves.

 

Most of the issues described in this thread are with the compass turned off, and while the stability of the compass in the Colorado may be an issue it shouldn't be confused with the accuracy of the GPS.

 

GO$Rs

Link to comment

I have recently run into similar problems on GPS accuracy. (I had posted in this same thread an earlier test that I had done under ideal GPS conditions on a football field and was frustrated at the compasses varibility when less than 35 feet from the cach under ideal conditions. GO$Rs had pointed out that that was probably not a GPS accuracy issue. I understand his point and now am reporting what recently happened to me regarding GPS accuracy.) On two occations while geocaching in the woods I ran into a situation where I was closing in on ground zero with a GPS accuracy of 25 to 30 feet and the cache distance about 40 to 50 feet away when my GPS accuracy suddenly went to 90 feet and the cache distance went to 150 to 200 feet away. I moved to a clearer spot and the GPS accuracy came back to the 25 to 30 feet range within a 10 minute period. I went back toward the cache and was able to ultimately find it. A similar occurance happened yesterday with similar results.

 

A different but GPS accuracy related issue happen this morning on a hike on a paved bicycle path where I had zeroed the track and trip computer before starting the hike in recreation mode. The trip computer and track measured similar mileage at the end of the walk which I think is close to being correct. I archived the track at the trailhead before leaving. I had walked on a paved path in the woods and was keeping an eye on the accuracy which was between 25 and 30 feet during the walk with the Colorado on a belt loop. The paved path was 8 feet wide and I walked on one side going and the other side comming back. I would estimate the difference between my tracks going and comming to be no more than 6 feet apart. When I got home I estimated the difference in the track between the going and comming track on the Colorado to be about 60 feet or so using the scale mark on the map screen. I downloaded the archived track onto Mapsource and measured it with the measutrement tool to be 50 feet between the going and comming tracks. This is an innacuracy of actual 6 feet vs 50 feet between the tracks or a factor of 8 or so. Given the GPS accuracy during the walk of (25 to 30 feet) , the resulting track separation of 50 to 60 feet seems high. You'd expect some of these inaccuracies to cancel each other out and not be at a max separation. I regularly walk that track and will record additional tracks for comparison. I guess I was a little surprised that the tracks were that far apart. Sli23sli

 

 

quote name='g-o-cashers' date='Feb 1 2008, 05:47 AM' post='3292867']

After using my CO 400t for the last two weeks I've been very pleased with the GPSr itself. I've seen no issues with grabbing and holding lock, EPE has always been sub 20', normally in the teens with a minimum of 5-6 satellites at good strength. I've been walking the same trail over and over since getting the CO and the track logs for each of these are right on top of each other (within 10-20'). So far so good, until this morning.

 

On the walk out everything looked fine. I went off onto some new trails and starting marking waypoints until I came back to the original trail. I noticed at that point my position didn't seem to match earlier walks but I went a little further to an intersection and I could see that my position on the map was way off based on the old tracks.

 

When I looked at the satellite page I was surprised to see the EPE in the 70-80' range. Based on my previous tracks I would estimate my position error was off by more than 300' ! At that point I stopped for a few minutes to see if things would settle out but there was no change. I had good strong lock on 7 satellites with one directly above, 3 at ~45 degrees and 2-3 closer to the horizon. I put the GPS in demo mode several times to see if that would bring it out of its funk but I got exactly the same results. Finally I power cycled the GPS and when it came back on I had EPE of 18 and the constellation/strength readings looked identical to me. My position on the map relative to previous tracks was exactly where I thought I should have been.

 

These issues will drive you nuts searching for a cache.

 

GO$Rs

Link to comment

Both of these sound like they could be the same issue I've described, although the one definitive characteristic of the issue I've seen which in my mind differentiates it from normal "wander" is that power cycling the GPS clears the issue immediately.

 

This really matters because it says to me that under nearly the exact same external conditions (the only minor difference is the time required to power cycle) the unit can give you drastically different readings. To me this suggests that there has to be an internal explanation (ie. a bug, design issue, hardware problem) that would account for the discrepancy.

 

GO$Rs

Link to comment

I can't say that I've really paid attention to the speed when I see the issue. I have compared the average and max speed to my 60cs during a walk and I did notice a couple of things:

 

1) Over a 1-2 mile walk the average speed and total distance are usually within 1-2% with the CO being on the longer/faster side.

2) Over the same walk the instantaneous speed of the 60cs and CO differ significantly. It is hard to characterize but my sense was that the CO fluctuated more and that it consistently read lower than the average. At least mentally the 60cs speed field seem to fluctuate around the average while the CO speed spent a lot of time below the average and would jump above for short periods of time. The fluctuation during the walk was as much as 3mph.

 

GO$Rs

Link to comment

I had my fourth conversation with Garmin on this issue yesterday and they've thrown up their hands and told me to return the unit (which I was planning to do under another RMA anyway). They claim they haven't heard of the problem, so if you are seeing it please give Garmin a call and let them know.

GO$Rs

 

I know I complained three times and only got the standard replies such as, calibrate, reset, auto-locate, etc. The rep I spoke to each time didn't really seem very knowledgeable. finally I just told them to send me a new unit which they did. I received it yesterday with software version 2.5 (not the 2.5.1 beta) and GPS 2.6. I will watch it closely to see if the problem still exists. My guess is it will. :D

 

This unit is going to end up going back as well though as the fit for the battery cover is really awful. I even tried the back from the first Colorado and still got the same fit issue so it is not the cover piece that is the problem it is the unit itself. The back slides off extremely easy, not that nice tight fit I was used to and the gap near the top is a joke, water will pour into this one. :)

 

2510115936_598d2c5dd4_m.jpg

 

Actually I believe you'll find that waterproof seal is inside the unit and not where the gap is shown in your picture. I also was unhappy about this gap, but realised it is not the seal. Why they didn't sort the tolerance out here and add another seal, who knows. On the accuracy side, my unit seems okay, and I've only had problems when under heavy tree cover. Why the track jumps by 100 odd meters in a second I have no idea (software logic could prevent that, based on previous averages, etc.). Hopefully improvement will continue to come from Garmin. More concerned if these are fixed by hardware revisions - where would that leave those of use with current units?

Link to comment

I'm beginning to wonder if the amount of stationary "soak" time the unit receives under good conditions before subjecting it to difficult reception conditions has an effect. I seem to see the problem less if I let the unit "soak" outside with good satellite reception before taking it into the woods. Normally this is something I try to do but I've noticed several of the times that I've had problems recently involved a quick power on and dash into the woods. Given the nature of the problem, which sort of feels like an issue of averaging in error, it would sort of make sense that averaging in a good signal might be beneficial.

 

Food for thought the next time any of you see this issue.

 

GO$Rs

Link to comment

When I looked at the satellite page I was surprised to see the EPE in the 70-80' range. Based on my previous tracks I would estimate my position error was off by more than 300' ! At that point I stopped for a few minutes to see if things would settle out but there was no change. I had good strong lock on 7 satellites with one directly above, 3 at ~45 degrees and 2-3 closer to the horizon. I put the GPS in demo mode several times to see if that would bring it out of its funk but I got exactly the same results. Finally I power cycled the GPS and when it came back on I had EPE of 18 and the constellation/strength readings looked identical to me. My position on the map relative to previous tracks was exactly where I thought I should have been.

I know this is an old post, but I was linked here from another thread. Has anyone thought of the why the EPE is bad? Don't they base the EPE of the four sats they use for the calculation? The sats move and the sat directly above is totally worthless. Maybe that sat was picked before it moved and not thrown out when moved straight up. I wish they showed the constellation they use and not just all of the sats they see.
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...