Jump to content

Distance changes while not moving - GPSMAP 76csx


chrisandvern

Recommended Posts

I am hoping all of the wise people here can help explain this to me.

 

I am using a Garmin GPSMap 76csx.

 

I am a student and was interested in seeing how far I actually walked each day so started my GPS in the morning when I left for school and left it on until I returned home, about 6 hours later. I know that I walk 2.8 miles (round trip) to and from the university. On campus I estimate about .5 miles, for an estimated total of 3.3 miles. Most of the day is spent sitting studying or in class.

 

When I returned home from school the GPS said I had gone 5.4 miles! There is no way that was correct. I left the GPS on while studying at home and noticed that about every 10 minutes the distance increased by a tenth of a mile while sitting on the table.

 

What is going on? Is the rotation of the earth changing my position relative to the satellites and recording that as movement? If so, the accuracy of the distance traveled comes in to question. I may go for a day long hike and the distance says 10 miles. Is it actually 9 or 11 depending on which way I move?

 

I appreciate insight into why this is and helping me understand!

 

Chris

http://www.qualityitemsofmontana.com

http://blog.qualityitemsofmontana.com

Link to comment

This is an annoying condition known as "wandering" where you bounce around standing still. Turn your track log on and sit it on a desk and zoom all the way in. You'll really see it then. It has to do with the varying accuracy and the way the GPS interprets it. I wish there was a setting that pinned your location when speed dropped below a user determined level, like 0.5 mph or something. As far as I know you can't do this. I use a 60CSx, which is electronically identical to your 76CSx.

Link to comment

These things are just not 100% accurate at all times. Particularly indoors where you were likely seeing at least some small amount of multipath errors.

 

The average error on these small handheld units is in the 15 to 25 foot range. As the satallites move overhead the constellation of sats is ever shifting. This changes the strength of signals and the calculations the unit makes. Your unit will vary readings within that 15 to 25 foot range over time - (larger with errors indoors). That appears to be movement to the GPS. Over time it will accumlate enough to give an error in the recorded distance. Download the recorded track from your unit and zoom way in - you will see slight 10 to 30 foot shifts recorded when you were holding still for a long time - well within the error margin.

 

I'll wager if you set it outside with a clear view of the sky - you will also record movement over several hours but it will be less than what you observed with it indoors.

 

While you are actually walking briskly, the error pretty much stays with your line of movement and is far less noticeable.

Link to comment

I am hoping all of the wise people here can help explain this to me.

 

I am using a Garmin GPSMap 76csx.

 

I am a student and was interested in seeing how far I actually walked each day so started my GPS in the morning when I left for school and left it on until I returned home, about 6 hours later. I know that I walk 2.8 miles (round trip) to and from the university. On campus I estimate about .5 miles, for an estimated total of 3.3 miles. Most of the day is spent sitting studying or in class.

 

When I returned home from school the GPS said I had gone 5.4 miles! There is no way that was correct. I left the GPS on while studying at home and noticed that about every 10 minutes the distance increased by a tenth of a mile while sitting on the table.

 

What is going on? Is the rotation of the earth changing my position relative to the satellites and recording that as movement? If so, the accuracy of the distance traveled comes in to question. I may go for a day long hike and the distance says 10 miles. Is it actually 9 or 11 depending on which way I move?

 

I appreciate insight into why this is and helping me understand!

 

Chris

http://www.qualityitemsofmontana.com

http://blog.qualityitemsofmontana.com

 

I've read that you can minimize this wandering behavior by setting your GPS to track by distance rather than time. To do this on my 60CSX (which I've heard is similar to the 76) the key sequence is Menu|Menu|Tracks|Setup|Record Method|Distance then type an appropriate interval. You can set the units for the interval by Menu|Menu|Setup|Units|Distance/Speed.

Link to comment

I've read that you can minimize this wandering behavior by setting your GPS to track by distance rather than time.

 

No, that really won't work. Distance is recorded in feet, then 0.01 of a mile on the trip odometer once it gets to 0.01 miles. The distance still measures and adds up regardless of whether the track log "drops" track points every so much unit of time or distance. Heck, turn track off completely, reset the data on the trip page and let it sit for a while. It will still show you as having traveled some distance. Doesn't matter if the track log is on or off or set to track based on distance or time intervals.

 

What WOULD work is if Garmin included a feature that "pinned" your location once you fell below a pre-set speed. Say, 0.10 mph or even less, which is pretty danged slow. That way the GPS would "know" you aren't moving and could disregard recording "jumping around" in the track/distance log. Then, as soon as you start walking again, the distance will be measured. There's already a speed feature you can set that will switch to compass mode under a certain speed, so it certainly isn't like "pinning" you location under a certain speed is impossible.

Edited by scotthsi
Link to comment

I've read that you can minimize this wandering behavior by setting your GPS to track by distance rather than time.

 

No, that really won't work. Distance is recorded in feet, then 0.01 of a mile on the trip odometer once it gets to 0.01 miles. The distance still measures and adds up regardless of whether the track log "drops" track points every so much unit of time or distance. Heck, turn track off completely, reset the data on the trip page and let it sit for a while. It will still show you as having traveled some distance. Doesn't matter if the track log is on or off or set to track based on distance or time intervals.

 

What WOULD work is if Garmin included a feature that "pinned" your location once you fell below a pre-set speed. Say, 0.10 mph or even less, which is pretty danged slow. That way the GPS would "know" you aren't moving and could disregard recording "jumping around" in the track/distance log. Then, as soon as you start walking again, the distance will be measured. There's already a speed feature you can set that will switch to compass mode under a certain speed, so it certainly isn't like "pinning" you location under a certain speed is impossible.

 

Actually, setting the track log to "record by distance" does work, not perfectly, but pretty well. My 60CSX just spent another hour on a rock in my back yard. Actual movement was 0. My settings from the track setup menu were set for a distance of 9 yards. Over that hour the trip computer logged a trip odometer distance of 34 feet and a moving time of 1 minute 13 seconds. During that hour 4 "tracks" were recorded to the memory chip. One of those had two points, the other 3 had just one. The two point track distance was 10 feet.

Link to comment

Actually, setting the track log to "record by distance" does work, not perfectly, but pretty well. My 60CSX just spent another hour on a rock in my back yard. Actual movement was 0. My settings from the track setup menu were set for a distance of 9 yards. Over that hour the trip computer logged a trip odometer distance of 34 feet and a moving time of 1 minute 13 seconds. During that hour 4 "tracks" were recorded to the memory chip. One of those had two points, the other 3 had just one. The two point track distance was 10 feet.

 

Track on, off, distance, time, should not matter. I tried this earlier today. Track on or off, time or distance, the trip odometer still clicked up the feet as time went on.

 

The ONLY thing creating a track does is for reviewing your path-over-ground later. It has NOTHING to do with how far you (or the GPS unit) traveled. The track is just a "breadcrumb" recording feature which simply shows how you got there.

Link to comment

Actually, setting the track log to "record by distance" does work, not perfectly, but pretty well. My 60CSX just spent another hour on a rock in my back yard. Actual movement was 0. My settings from the track setup menu were set for a distance of 9 yards. Over that hour the trip computer logged a trip odometer distance of 34 feet and a moving time of 1 minute 13 seconds. During that hour 4 "tracks" were recorded to the memory chip. One of those had two points, the other 3 had just one. The two point track distance was 10 feet.

 

Track on, off, distance, time, should not matter. I tried this earlier today. Track on or off, time or distance, the trip odometer still clicked up the feet as time went on.

 

The ONLY thing creating a track does is for reviewing your path-over-ground later. It has NOTHING to do with how far you (or the GPS unit) traveled. The track is just a "breadcrumb" recording feature which simply shows how you got there.

 

It may not work for you, but it does work for me, and in previous posts on this thread, I've described how and why. I guess any other readers will just have to try it out, and make up their own minds.

Link to comment

I've described how and why. I guess any other readers will just have to try it out, and make up their own minds.

 

Um, NO, you haven't described "how and why". Only that it "appears" to make some difference for you. I've described "how and why" it would NOT work. The GPS unit records the distance traveled and displays it on the trip page via the odometer. That's it. Having a track should not make one bit of difference. Same as if I walked down a beach to a pre-determined point with a stick poking holes in the sand every five steps (distance) so I could later go back and see every twist and turn I took in the sand. Or making a poke mark every two minutes (or whatever) to create track via time intervals. Either way doesn't change the distance I've traveled. It would be EXACTLY the same if I followed my path EXACTLY each separate time.

 

Now, where again is your explanation of "how and why" dropping track point at distance intervals versus time intervals makes a difference other than "it just does"? Or even having the track function turned on at all? Hmmm? You already made the claim...back it up. :laughing:

Edited by scotthsi
Link to comment

What WOULD work is if Garmin included a feature that "pinned" your location once you fell below a pre-set speed. Say, 0.10 mph or even less, which is pretty danged slow. That way the GPS would "know" you aren't moving and could disregard recording "jumping around" in the track/distance log. Then, as soon as you start walking again, the distance will be measured. There's already a speed feature you can set that will switch to compass mode under a certain speed, so it certainly isn't like "pinning" you location under a certain speed is impossible.

I think they try and do that already. The problem is a GPS gives your approximate location only. Speed is calculated by figuring out the distance between two locations and dividing by the time between locations. My wild guess is that if you leave the GPS sitting all day the distance between two consecutive locations will occasionally be large enough to give a speed spike.
Link to comment

It has been reported numerous times before that you will often get different distances for the same trip on the Trip Computer and the Track Log. (Search these forums.) Try it yourself on pretty well any recent Garmin, and you will see the effect.

 

I believe the reason for the different distances is that the Trip Computer updates every second or so (on my Summit HC, there is no user-controllable way to adjust the frequency of updating in the Trip Computer), and if it records a new location, it will add the distance to the Trip Computer odometer. (This is quite independent of the Track Log function.) Even when you are standing still, your computed position wanders around a bit - a couple of metres here, a couple of metres there - even under ideal satellite viewing conditions. When satellite viewing conditions are sub-optimal (indoors, in built-up high-rise areas, in deep forests, in deep canyons, etc), the amount of "wander" or "drift" will increase. You will also note the GPSr computes a speed even when you are standing still - this will often read as 0.0 km/hr, but if it "thinks" you just moved 2 metres in 1 second, your instantaneous "speed" is computed as 2 m/s = 7.2 km/hr. Your GPSr has no way of knowing that you are actually stationary, so it just assumes that you really did move, and adds the distance to the Trip Computer odometer. (Incorporation of an accelerometer would be one way to improve the GPSr's ability to distinguish between "drift" and real movement.) When you are moving at a reasonable speed, these random "jumps" around your true position are of much less significance, being much smaller in effect than your actual movement. However, in the current software implementation, your Trip Computer will therefore tend to over-estimate your travel distance if you spend much time stationary.

 

The Track Log works a bit differently - you set it up to record Track Points at a time / distance frequency that works for you. The important thing is that the computed Track Log distance is the sum of the straight line distances between only the recorded track points - any computed movement which does not generate a Track Point does not contribute to the computed Track Log distance. Typically, when you stand still, for some time, your track log will still show a number of track points and track segments in a sort of "star burst" around your true stationary position. However, depending upon the recording frequency you have chosen, there are likely to be fewer such track segments than the full history of random second-by-second "jumps" that will be recorded by the Trip Computer odometer, which slavishly computes a new location and distance increment every second or so.

 

Note that BOTH the Track Log and the Trip Meter odometer will therefore over-estimate your travel distance if you spend long periods stationary (or indoors etc), but the Track Log distance will probably be a bit shorter than the Trip Computer odometer (with typical Track Point interval settings) - best to turn your GPSr off while stationary or indoors, if this bothers you.

 

In some earlier unit software versions, Garmin apparently tried to damp out the "wandering" effect in the Trip Computer exactly as you suggest - by ignoring motion when the computed speed was below some threshold. The trouble is - this tended to also damp out actual movement when you were moving quite slowly, and the Trip Computer odometer sometimes significantly under-estimated your travel distance as a result.

 

It seems we just have to accept the limitations of GPS technology for now - at least until consumer GPSrs see some additional technology (such as accelerometers) to assist in recognising the difference between real movement and GPS signal "drift".

 

Hope this helps!

Edited by julianh
Link to comment

Thank you very much everyone for your insightful and informative responses. It makes sense to me, even though it is frustrating.

 

Have a good week everyone!

 

It has been reported numerous times before that you will often get different distances for the same trip on the Trip Computer and the Track Log. (Search these forums.) Try it yourself on pretty well any recent Garmin, and you will see the effect.

 

I believe the reason for the different distances is that the Trip Computer updates every second or so (on my Summit HC, there is no user-controllable way to adjust the frequency of updating in the Trip Computer), and if it records a new location, it will add the distance to the Trip Computer odometer. (This is quite independent of the Track Log function.) Even when you are standing still, your computed position wanders around a bit - a couple of metres here, a couple of metres there - even under ideal satellite viewing conditions. When satellite viewing conditions are sub-optimal (indoors, in built-up high-rise areas, in deep forests, in deep canyons, etc), the amount of "wander" or "drift" will increase. You will also note the GPSr computes a speed even when you are standing still - this will often read as 0.0 km/hr, but if it "thinks" you just moved 2 metres in 1 second, your instantaneous "speed" is computed as 2 m/s = 7.2 km/hr. Your GPSr has no way of knowing that you are actually stationary, so it just assumes that you really did move, and adds the distance to the Trip Computer odometer. (Incorporation of an accelerometer would be one way to improve the GPSr's ability to distinguish between "drift" and real movement.) When you are moving at a reasonable speed, these random "jumps" around your true position are of much less significance, being much smaller in effect than your actual movement. However, in the current software implementation, your Trip Computer will therefore tend to over-estimate your travel distance if you spend much time stationary.

 

The Track Log works a bit differently - you set it up to record Track Points at a time / distance frequency that works for you. The important thing is that the computed Track Log distance is the sum of the straight line distances between only the recorded track points - any computed movement which does not generate a Track Point does not contribute to the computed Track Log distance. Typically, when you stand still, for some time, your track log will still show a number of track points and track segments in a sort of "star burst" around your true stationary position. However, depending upon the recording frequency you have chosen, there are likely to be fewer such track segments than the full history of random second-by-second "jumps" that will be recorded by the Trip Computer odometer, which slavishly computes a new location and distance increment every second or so.

 

Note that BOTH the Track Log and the Trip Meter odometer will therefore over-estimate your travel distance if you spend long periods stationary (or indoors etc), but the Track Log distance will probably be a bit shorter than the Trip Computer odometer (with typical Track Point interval settings) - best to turn your GPSr off while stationary or indoors, if this bothers you.

 

In some earlier unit software versions, Garmin apparently tried to damp out the "wandering" effect in the Trip Computer exactly as you suggest - by ignoring motion when the computed speed was below some threshold. The trouble is - this tended to also damp out actual movement when you were moving quite slowly, and the Trip Computer odometer sometimes significantly under-estimated your travel distance as a result.

 

It seems we just have to accept the limitations of GPS technology for now - at least until consumer GPSrs see some additional technology (such as accelerometers) to assist in recognising the difference between real movement and GPS signal "drift".

 

Hope this helps!

Link to comment
What WOULD work is if Garmin included a feature that "pinned" your location once you fell below a pre-set speed.
...The problem is a GPS gives your approximate location only ... if you leave the GPS sitting all day the distance between two consecutive locations will occasionally be large enough to give a speed spike.
If the GPSR had an accelerometer, it could filter those out also. If there was a sudden change in GPS calculated position from one second to the next, but the accelerometer didn't notice a corresponding bump in velocity, it could be filtered out.

 

At least two hand held GPSRs I know of contain accelerometers: DeLorme PN-40 and Apple iPhone. On the iPhone, the Motion-X GPS application includes an option to use the accelerometer to augment the GPSRs accuracy at low speeds. DeLorme indicated some months ago that they might implement acceleromete assisted track smoothing, but I don't think it's in their current firmware.

 

See also: http://appshopper.com/navigation/motionx-gps

See also: http://forums.delorme.com/viewtopic.php?t=19652

 

---

edit to add: I see JulianH mentioned accelerometers also, but in a "someday" context. Someday is close than you think.

Edited by lee_rimar
Link to comment

Having a track should not make one bit of difference. Same as if I walked down a beach to a pre-determined point with a stick poking holes in the sand every five steps (distance) so I could later go back and see every twist and turn I took in the sand. Or making a poke mark every two minutes (or whatever) to create track via time intervals. Either way doesn't change the distance I've traveled. It would be EXACTLY the same if I followed my path EXACTLY each separate time.

 

 

I am neither a hardware or software guy, but I'm pretty sure that with my 3 year old 60CSX the tracklog setting establishes the sampling method for the unit.

 

The purpose of setting the tracklog to "record by distance" is to set the sample rate to be larger than the error rate (point cloud) for a stationary GPSr. The track you record while walking down the beach may be the same when you are moving, but if you are only allowed to poke a hole every 30 feet, and you are sitting on the beach, you won't poke more than one hole. If you are poking a hole every 15 seconds, you will poke a lot of them whether you are moving or not.

 

In my last post to the following thread

 

http://forums.Groundspeak.com/GC/index.php...rimble+Planning

 

I reported results for a test with the GPSr set to record by "Auto". Those results, of a 55 minute test, described a point cloud in which there was about .5 miles of travel that never happened.

 

I also reported results above in this thread with the tracklog set to record by distance (the distance interval was set to 9 yards). During this hour long test, my GPSr generated no in memory tracklog, only recorded 4 points to the memory card, and only generated a spurious travel distance of about 30 feet.

 

Scotthsi, in response to your earlier comments "It just works" is a perfectly reasonable empirical response to a condition that is observed, but not well understood. To quote a time worn adage, "In theory, there is no difference between theory and practice. In practice there is." Then again, there may be some difference between the old SirfStar 60CSXs and the ones with the new chipset.

 

For the rest of the readers, I came across this method of dealing with damping the point cloud in another forum. Where it seemed to be generally accepted. Try it out, it "just might work" for you.

Edited by seldom_sn
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...