Jump to content

Garmin Colorado drift?


Crid

Recommended Posts

Last June I bought a Garmin Colorado. Like several other owners, I sometimes experienced drift when in woodland (on one occasion over 600ft). A GPS software upgrade (not a firmware upgrade) released around October or November appeared to sort the problem out and after that my Colorado seemed to be on a par with my 76CSx.

 

A couple of weeks ago I managed to damage my Colorado and sent it in for a replacement. Yesterday I got the replacement and it seems to be a brand new unit (serial number starts with 16911), loaded with 2.90 firmware and 2.90 GPS software.

 

Today I took it out for a test walk along with my 76CSx, in woods that I have previously been in with my old Colorado without incident.

 

Everything looked good for about 50 minutes to an hour, under various levels of canopy (and some open sky). Then I happened to look at the Colorado and noticed that it was saying accuracy was about 75ft, but the track was clearly a lot further off than that. I have previously mapped the path in question, and the 76CSx was on track.

 

The Colorado drifted further and further off course. Eventually I turned it off and back on and let it reaquire the satellites. At that point it was back on track and was OK on my walk back to the car.

 

I thought this was all sorted out now? I'm disappointed that this problem seems to have returned for me (even though it's on a different physical unit to before). I can't upgrade the firmware or GPS software because they are already up to date.

Edited by Crid
Link to comment

My route was anti-clockwise. Green track is the 76CSx, red track is the Colorado...

 

Colorado-Drift.jpg

 

As you can see, the drift started when I turned the corner at the top, and steadily got worse. It was 544ft off by the next corner, where I turned. By the next corner it was 819ft off and showing no sign of coming back. At that point I powered the unit down and back up, it reaquired the satellites and was back on track.

Link to comment

It's quite annoying, sometimes even troubling when your gps receiver does that. Unfortunately, it seems to be the price you have to pay for the “Higher” sensitivity units. There is only so much you can do with actual reception, then they start doing some interesting things with data manipulation to provide you a position solution even when there might be some issues with the position. For the most part, people seem to prefer this over signal loss.

 

The first receiver I had that displayed a tendency like this was my sportrak maps. They would have glitches like this from time to time, but overall people seemed to prefer the sportraks over say a B/W etrex, which would simply loose reception in areas where the sportrak would make stuff up. Myself I preferred no data to bad data, but I was in a minority with that preference. Overall, you have to admit these little handheld units are amazing. Not that many years ago, navigation where you couldn’t navigate off terrain features was difficult at best, and getting within miles of your destination was quite the feat. Now we get frustrated with errors measure in feet.

Link to comment

Wow, that is some serious drift!

 

Searching_ut is right though, I' too have noticed more anomalies with my PN-40 (high sensitivity unit) than with my etrex legend as far as tracking goes.

 

The software has come a long way in reducing a lot of the timing/multipathing issues that exist with the higher sensitivity units, but it's still not quite perfect.

 

I would estimate 9/10 of my hikes using the PN-40 are perfectly accurate. One of them usually turns out a little buggy though. A lot of it depends on the terrain and surroundings.

Edited by Tahoe Skier5000
Link to comment

I guess I'll have to keep an eye on it. Maybe this was a one-off, but it's disconcerting considering this was my first real test of the replacement unit.

 

I took the unit out today in different terrain and it was fine. The trip was an out-and-back, which allowed me to compare the outbound track to the return one. It actually seemed more consistent than the 76CSx, but there wasn't tree cover (just a lot of scrub - some of it pretty high).

 

I'm going to try keeping WAAS switched off, just in case that's the cause of the problem. (I wouldn't expect WAAS to be active in woodland anyway).

 

I must say that I hadn't seen my old Colorado do this since the software upgrade last year, and I've never seen my 76CSx drift anywhere near as much as this in the 2 years I've had it.

Link to comment

It looks like you took these readings on 2 separate occasions and this wasn't a side by side comparison. There are many reasons why a GPSr would be reading perfectly fine one day and drift the next.

The only way to do a proper comparison would be to turn both units on and carry them next to eachother and do a track. Then compare the tracks.

 

I would be interested to know a side by side comparison.

Link to comment

It looks like you took these readings on 2 separate occasions and this wasn't a side by side comparison. There are many reasons why a GPSr would be reading perfectly fine one day and drift the next.

The only way to do a proper comparison would be to turn both units on and carry them next to eachother and do a track. Then compare the tracks.

 

I would be interested to know a side by side comparison.

It WAS a side-by-side comparison. I was carrying both GPSr units at the same time. 76CSx was fine. Colorado drifted off.

Edited by Crid
Link to comment

It looks like you took these readings on 2 separate occasions and this wasn't a side by side comparison. There are many reasons why a GPSr would be reading perfectly fine one day and drift the next.

The only way to do a proper comparison would be to turn both units on and carry them next to eachother and do a track. Then compare the tracks.

 

I would be interested to know a side by side comparison.

It WAS a side-by-side comparison. I was carrying both GPSr units at the same time.

 

Oh, I read things wrong. My bad.

 

Much more interesting then.

 

I have also had my Colorado drift. What I found odd is that when it drifts it refuses to auto correct. Mine also needed to be turned off and back on.

 

There are BETA updates available but they don't appear to address drift issues or the need to power off and back on but the BETA updates still could include a fix for hat issue nonetheless.

Link to comment

I've experienced the same with an Oregon 300. The old unit never experienced drift, but the battery connectors broke - and I got a warranty replacement. The new one is now giving me many more occurrences of drift (2 out of the last 3 walks). I *think* it has got worse on the latest beta firmware (3.13), but I'm not sure.

Link to comment

It's quite annoying, sometimes even troubling when your gps receiver does that. Unfortunately, it seems to be the price you have to pay for the “Higher” sensitivity units. There is only so much you can do with actual reception, then they start doing some interesting things with data manipulation to provide you a position solution even when there might be some issues with the position. For the most part, people seem to prefer this over signal loss.

I disagree that "this is just how it is". The problem here is not the occasional abrupt shift in reported position which is to be expected with high-sensitivity receivers in multi-path conditions. The problem here is that the firmware is apparently trying to damp these shifts in a manner which leads to gradually but continuously increasing error in some situations. The fact that power-cycling the unit results in an instant -- and correct -- fix tells you that there is no significant issue with reception at that time.

 

Although Crid does not mention it, the unit takes a very long time to self-correct the drift, once started, even when you return to near-ideal reception conditions. In my experience, the time it takes far exceeds my patience in the field. This also supports the theory that it's an "averaging" kind of problem, not an issue with reception per se. When in doubt, I power-cycle early and often :D

 

I'm not saying that the firmware problem is obvious or easy to fix. Obviously it's not. But it clearly is a firmware problem, not an inevitable consequence of the receiver design.

Link to comment

Whoa!......to blame the drift on (ALL) the "Higher sensitivity chips" is a very uninformed statement!

 

The 76CSx has a high sensitivity chip, and you'll notice that it's not the one being described as having the problem. (the 60 series doesn't have the problem either)

 

The problem lies with the Mediatek & Cartesio "brand" chipsets. It started with the ETrex H series and continues thru the PN-40 and Colorado & Oregon series. (I'll also bet that the newer 60's with the M-Tek chips start having "problems")

 

That's also the express reason I refuse to buy a new model GPS until those problems are corrected. They are simply not as accurate and dependable as my 76CSx. (with the SIRFlll chipset)

 

Until the manufacturers can correct the problems or change to more accurate chipsets, you'll continue to hear complaints about "drift" and WAAS reception , etc.

Link to comment

Although I have seen significant drift on my Colorado (M chipset) a few times, I have never seen it on my PN-40. Maybe just lucky, though. Or maybe Grasscatcher is tarring two different chipsets with one brush. :D

 

The generalization doesn't entirely hold water on the WAAS front, either. The PN-40 had (and to a much lesser degree after a chipset f/w update, continues to have) problems with the ability to consistently receive and use WAAS corrections. On the other hand, my Colorado (side by side with my PN-40) has no such problems. This might be due to differences in antenna type, differences in the chipset, or whatever. Regardless, it's not accurate to state that all units which use these chipsets have WAAS problems.

 

I'm pretty sure the concensus is that the Cartesio chipset firmware did (and to some degree still does) have WAAS issues in multiple brands and models. I'm not sure I've seen such a broad concensus on the drift issue, even with regard to Mediatek chipsets, let alone both Mediatek and Cartesio.

 

Just my two cents.

Link to comment

I've seen drift issues on MTK, Cartesio and Sirf. If you use any of them long enough you'll see tracks with errors > 100'. The difference I've seen with the Colorado (although I'll have to say I haven't seen any recently) is the length and duration of the drift. The Oregon and 60csx tend to be less than 100' and are usually a few 10's of seconds in duration. The Colorado drifts could be minutes and several hundred feet. I would agree with twolpert that it seems to be some sort of averaging algorithm problem (not a receiver accuracy problem) given that a restart would correct it immediately.

Link to comment

Worst I have seen lately on the Colorado is about 100 feet over the space of a quarter mile under light hardwood cover. Don't know how much worse it would have gotten if I had not noticed. True to my impatient nature, I waited a minute or two standing still in relative open, the power-cycled :D

Link to comment

Twolpert- Check out post # 9 above....so it's not me painting with a broad brush.

I had a PN-40(very early) for a short while, and noticed the problem immediately.

An easy way to see the problem is to log/record an intricate track in dense cover, or at least difficult reception conditions, then download the track and view it displayed over aerial photos. Zoom in to where the trail is clear and visible on the photo, and see how well the track/trail match.

 

You will start to see things that you really won't be proud of. Note: GE photos often are not georeferenced accurately enough to use for this purpose.

 

NEVER.....have I seen a "drift" problem with my 76CSx. When mapping trails / logging accurate tracks, I regularly also use an external antenna.

 

I have been in conditions where there was absolutely no reception(Slot canyons,etc), where even the 76CSx won't work.

Link to comment

Twolpert- Check out post # 9 above....so it's not me painting with a broad brush.

I had a PN-40(very early) for a short while, and noticed the problem immediately.

An easy way to see the problem is to log/record an intricate track in dense cover, or at least difficult reception conditions, then download the track and view it displayed over aerial photos. Zoom in to where the trail is clear and visible on the photo, and see how well the track/trail match.

I think we have to make a distinction between the persistent, large-error drift on the Colorado and expected inaccuracies in challenging conditions.

 

I have done the kind of comparison you suggest often with geo-referenced imagery from the DeLorme library, both on the unit and on the PC. While I have seen tracks which diverge from reality, I have never (knock on wood) seen one which diverges much farther than I would expect with the EPE at the time. In general, the worst-case errors I have seen are in the 40-50 foot range in very poor conditions (heavy cover and bad geometry). What I have never seen is the kind of ever-increasing, non-self-correcting drift which occurs occasionally on my Colorado.

 

Which is not to say it can't happen. But I've never seen it.

Link to comment

Happens frequently on my vista HCX (at least 1 in 4 hikes). I've seen it as bad as 800 feet, and regularly 300-400 feet. One way to judge just how bad is to compare GPS elevation to actual elevation, because this issue also impacts the altimeter. My altimeter can be 500-600 feet off when left in autocalibrate mode because it is using bad GPS elevation data to maintain it's calibration.

 

As always, a power cycle instantly corrects the problem.

Link to comment

Hi everybody,

 

I thought I would jump in here with my experience with my Colorado 400t and the 2.9 GPS firmware. I've had my Colorado for well over a year and I have to say that I've never seen the bad drift problems that other have seen, or at least I've never noticed it. Last week I upgraded from 2.8 to 2.9 and now I see drift errors over 500 feet. Here is an example that compares a Forerunner 205 with the Colorado.

 

http://picasaweb.google.com/wazkaren/Color...805270652402146

 

The blue is the Forerunner and the red is the Colorado. The Forerunner is right on. In the northern most part of the track I was out of the woods in an open strip of land as is clearly visible in the picture. The Colorado shows my over 500 feet south in the woods. This is with an unobstructed view of the sky.

 

Here's another example that shows the Colorado track is a mess:

 

http://picasaweb.google.com/wazkaren/Color...798497198786178

 

I've been in contact with Garmin support but so far I'm getting the usual delaying tactics (re-acquire satellites, hard reset). I've trying to get them to help me downgrade to 2.8 but so far they won't do it.

 

I've been through the various FAQs and this forum but I have not yet been successful to downgrade the GPS firmware. I tried the updater.exe but got a error saying it was trying to write to a protected area of memory. Is the GPS firmware the same file that's used on the vista HCX? Has anyone been able to downgrade their Colorado?

 

Thanks,

Greg

Link to comment

What does Garmin say?

 

I would be happy to help you out with your colorado today. The chipset is not something that we suggest backdating as it would have the opportunity to negatively effect your sat. acquisition. I would suggest at this you perform an Auto Locate to obtain new satilite information.

How do I perform an Auto Locate?

1) Access the Shortcut Menu

2) Select Others

3) Select Satellites

4) Press Options (Left Soft Key)

5) Select Use With GPS Off

6) Immediately press Options (Left Soft Key) again

7) Choose Autolocate Unit

8) Set the unit outside in plain view of the sky for 20 minutes so that it can record fresh satellite data

 

then when that didn't work:

 

Thank you for contacting Garmin International. I am happy to further assist you with this. Since the auto-locate did not correct the issue, I suggest you master reset the unit. Please keep in mind that resetting the unit will erase all user data such as your waypoint, routes, and tracks. Please be sure to back up any information you may want to save before continuing.

 

To master reset the Colorado:

....

Link to comment

I have the old software versions but they do not update the GPS firmware. I have only updated the GPS firmware through the webupdater and can't find a standalone version. I've seen them for the Vista HCX but not the Colorado. That's why I was wondering if the Vista HCX and Colorado GPS firmware was the same.

 

Greg

Link to comment

I did a hike in the same park today and my results were very good, I don't see anything that I would call a drift problem:

 

http://picasaweb.google.com/wazkaren/Color...172551708975794

 

Blue track is Forerunner, Red is Colorado.

 

The differences between this track and the previous are:

1. I downgraded the Colorado software from 2.95 to 2.8. GPS Firmware stayed the same at 2.9.

2. The speed is different: I walked the trail instead of running.

3. The path I took was slightly different from the previous day.

 

Could it be the 2.95 software that's the issue? I'll run with 2.8 more this week and see if I have the drift issues again with the 2.8 software.

 

Greg

Link to comment

OK, I'm totally frustrated with Garmin support and I give up. Has anyone successfully downgraded the chipset firmware on a Colorado? I have the GPSChipsetTypeM2_RegionFile__280.rgn file and I tried using Updater.exe using the instructions on the Colorado Wiki but that fails with 'The GARMIN device is unable to program the requested memory region....'

 

I have gotten responses from 5 different Garmin support people. The summary of the "help" they've given:

 

1. We don't recommend downgrading chipset firmware

2. A pointer to the unit software .exe files (the unit software which I didn't ask for)

3. A pointer to an.rgn file (on http://www.gawisp.com/perry/chipset_firmware/Type_M) with no other instructions

4. Then when I asked with to do with the .rgn file they said to click on the link and select run (huh?)

 

Thanks,

Greg

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...