60CSx Atlimeter

[+DarthGator]

New to the sport. Just purchased a 60CSx on Amazon at a pretty decent price.

 

Does anyone know the best way to calibrate the altimeter? I'm not convinced I am getting the most accurate reading. One thought I had was using a contour line from the topo map to set my known altitude.

 

Any other thoughts?

[+Dryphter]

Go somewhere with a known elevation such as an airport. Then type that into the altimeter calibration screen. I looked up the elevation for my town and figured that was close enough.

[+Red90]

IMO.... leave it to auto calibrate using the GPS signal.

[toddm]

It's a crapshoot in my experience. In certain places the GPS signal elevations are very good, in others they are way off. On average over time though the GPS signal elevations are probably more accurate.

 

The problem with using the barometric altimeter is that if a weather front comes through it will throw things off. I've seen those vary by a couple hundred feet just sitting on my desk for a day if weather comes in or leaves. Over short periods of time with stable weather using a survey benchmark or sea level itself and a barometric altimeter is more accurate in many cases than the GPS signal, but only if the weather is consistent and you calibrate at least daily. For most people that's just not feasible, so the gps signal elevations are the best bet for less hassle.

 

Altimeters are really much better suited to recording ascent and descent data than keeping a true record of your actual exact elevation.

 

Also there is the chicken and the egg issue. I've seen topo maps off a couple hundred feet from survey benchmarks as well, so if you are looking at gps signal elevation and comparing it to a topo map, there is always the question of if the topo map is wrong, or the gps signal is wrong.

 

Basically it's not an exact science, at least not in our consumer gps units. I've come to the conclusion that if I let the unit autocalibrate as Red90 suggested it keeps me generally within +/- 200' of actual elevation.

[toddm]

Double post, my bad

Edited April 4, 2008 by toddm

[hikerhiker]

My 1st GPS had a barometric altimeter. My new one, 60cx doesn't. I wanted a GPS that would tell me my elevation but instead I had to keep telling the one with the altimeter what its elevation was. I gave up, now I just use the GPS elevation. Close enough !

[OGBO]

The barometric altimeter in the 60CSx (and other Garmin altimeters) has got to be my biggest complaint with them. After a bad experience with the 60CSx in Antarctica, I had a series of exchanges with the Garmin tech folks and never got a satisfactory response (I'll describe the problem below). As long as you have a 3D fix with a reasonable constellation (SVs spread over the satellite display), the accuracy of the GPS-derived altitude is as good or better than the mapping accuracy of USGS maps (refer to the USGS mapping website for the accuracy standards - basically 1/2 contour interval at surveyed points, which means +/- 20 ft for most quads and could be much more in many areas, vs the GPS error budget of 1.5 x horizontal error, or +/- 30 ft 1 sigma). If you derive altitudes from the DEM files or do stereographic tracing (as done for the contour lines), you can be off by much more. Obviously if your estimated error is very large, due to a poor constellation view (DOP, to remind the old-timers, which used to be displayable), your altitude error can be large as well.

 

Having said that, the basic complaint is that the 60CSx displays the barometric altitude all the time. There is no option for displaying the GPS-derived altitude in the window on any screen. You have to recalibrate frequently, especially in mountainous terrain. I have seen many times an altitude set at, say, a lake with a well-derived lake level altitude, then climbed up a few thousand feet to a nearby peak, only to find the altitude displayed off by 200 or 300 feet, then descend back to the lake and find it displaying the lake level altitude. This is a peculiarity of all barometric altimeters, including my favorite pocket Thommens - the real lapse rate (altitude vs pressure) is frequently different from the Standard Atmosphere tables. There are all sorts of other error sources for barometric altimeters, such as change of pressure during the day, or due to frontal passage, or even standing on the windward vs on vs leeward side of a pass when the winds are strong (venturi effect).

 

Now you can look at the GPS altitude without re-calibrating. Go to the Satellite page. Push the menu button. Scroll down the menu to "GPS Elevation" and push "Enter". The displayed altitude is not continuously updated, but is instead the computed altitude at the time you pushed "Enter".

 

The Antarctic incident - It is widely stated that at high latitudes (more than 70 deg north or south), the absolute pressure at a given altitude is less than at the same physical altitude above Sea Level. I have confirmed this by taking the absolute pressure, for example, in ascending from Mt Vinson Base Camp up to High Camp, then comparing the surveyed altitude (derived during the Project Omega mapping study of the Sentinel Range) with the altitude the absolute pressure would indicate using the Standard Atmosphere tables. At High Camp, the absolute pressure corresponds to an altitude about 1500 feet higher than the surveyed altitude. With the 60CSx, if you attempt to set ("calibrate") the indicated altitude to the surveyed altitude, within a few seconds, the displayed altitude switches to the barometric altitude as you would get from the ICAO Standard Atmosphere Table, whether or not "automatic recalibration" is shut off. This happens whether you choose "I know the altitude", "I know the barometer setting", or "use the GPS altitude", as long as the barometric altitude is more than 300 meters or so higher than you set it, and apparently above some minimum indicated altitude (Garmin tech was very reticent on this point). Garmin's response to my suggestion that an option to display only the barometric altitude was to dismiss it as being of interest to a tiny number of users, probably no more than a dozen.

[+hogrod]

Just leave it on auto calibration, in this mode it uses the elevation it gets from the GPS signal to average its way to accuracy. I have found after about 30mintues its pretty accurate.

 

If you really want to mess with the pointless calibration I used the posted elevation from a benchmark, then on the way home I drove by the post office and noticed it also had the elevation listed on its sign and my GPS was right on. when I got home I made a mental note of the elevation of my house. after checking my topo 2008 maps were pretty spot on for elevation.

[+julianh]

New to the sport. Just purchased a 60CSx on Amazon at a pretty decent price.

 

Does anyone know the best way to calibrate the altimeter? I'm not convinced I am getting the most accurate reading. One thought I had was using a contour line from the topo map to set my known altitude.

 

Any other thoughts?

I have a Summit HC. My procedure is simply that I calibrate the altimeter at the start of the day, and have auto-calibrate turned on. I don't do any other manual calibrations at all, until the next time I turn the unit on.

 

Methods of calibrating at the start of the day are:

 

1) If you know your current elevation, use that. (I know my home elevation is 50 metres).

2) If you have access to a topo map (either paper map, or topo map loaded into your GPS), interpolate between the contours to get your current elevation.

3) If you have neither of the above, wait until you have a good 3D fix, and use the GPS elevation for initial calibration.

 

The Garmin auto-calibrate algorithm on current high-sensitivity units appears to adjust the sea-level barometer pressure on a more or less continuous basis, using the trends of the GPS elevation as a guide, but it seems hat the algorithm is "damped" so as to not over-react to rapid or spurious changes in GPS elevation, which you can get frequently depending on number of satellites in view, spread of satellites, multi-path reflections, etc. I don't now exactly how the algorithm works, but if you set your unit up to log elevation at a single location, and compare a trace of the auto-calibrated barometric elevation to the raw GPS elevation, you will see that the barometric elevation is MUCH more stable than the GPS elevation, and DOES correct for changes in air pressure over the course of the day.

 

I typically find the auto-calibrated barometric altimeter will remain accurate to about plus or minus 5 metres accuracy all day under most conditions, maybe plus or minus 10 metres worst case. This is MUCH better than you will ever get with GPS elevation alone. (With good satellite visibility, a good spread of satellites, and no multi-path errors, etc, GPS elevation of plus or minus 10 to 20 metres is about as good as I have ever seen. With poor satellite visibility, such as under heavy tree cover or in canyons etc, plus or minus 50 to 100 metres is not uncommon. I have NEVER seen my auto-calibrated altimeter that far out.)

 

I can't comment on how it works in Antarctica, or other areas with "unusual lapse rates" etc, but it works extremely well for me all over Australia, and in Singapore and Indonesia (which are the only places I have tested it so far).

 

Hope this helps!

Edited April 5, 2008 by julianh

[+julianh]

The problem with using the barometric altimeter is that if a weather front comes through it will throw things off. I've seen those vary by a couple hundred feet just sitting on my desk for a day if weather comes in or leaves. Over short periods of time with stable weather using a survey benchmark or sea level itself and a barometric altimeter is more accurate in many cases than the GPS signal, but only if the weather is consistent and you calibrate at least daily. For most people that's just not feasible, so the gps signal elevations are the best bet for less hassle.

With Garmin units, set the altimeter to "auto-calibrate", do an initial calibration at the start of the day, and you should find that the barometric altimeter auto-corrects for changes in barometric air pressure all day, without requiring additional manual calibrations (as long as you leave it turned on, with a more or less continuous 3D fix).

 

Hope this helps!

[RGK]

The barometric vs. GPS altitude has been a thorn since the 60CS first came out. I have sent multiple emails to Garmin on this topic, the only result has been to add a clumsy static display on the Satellite page. The forced barometric altitude use is especially troublesome when hiking near mountain tops, saddles, high winds in mountains, during rapid weather changes, and when in a pressurized aircraft. In each of these cases where the barometer-derived altitude is used for position solution, the lat/long solution can be severely in error; I've seen it miss by miles when in a pressurized a/c. How hard can it be to give us the option to select which altitude to use?

 

I am (was) a Garmin Colorado victim, I've had it for 2 months, had about a week's worth of marginal use in that period, it is now in the Garmin Hospital with an ETA of 6 weeks after I had sent it to them. But that is another story. The CO has the same idiotic altitude policy as its (alleged) predecessors; but I can trick it into recording GPS altitude on track logs. My CO failed before I could get enough data to determine whether it used GPS altitude for lat/long computation, my suspicion is that it doesn't. The trick is to go to: Shortcuts>Setup>Altimeter>Barometer Mode>Fixed Elevation. This forces a recorded GPS altitude on the bloated .gpx log file. At least the recorded value seems to be close to GPS altitude. There may be some other side effects that I would have discovered if the thing hadn't failed.

 

Anybody want to trade a new 60CSx + $150 for a Colorado 300 + some 2GB SD cards?

 

A recovering Garmin Early-Adopter

[+julianh]

I can trick it into recording GPS altitude on track logs. My CO failed before I could get enough data to determine whether it used GPS altitude for lat/long computation, my suspicion is that it doesn't. The trick is to go to: Shortcuts>Setup>Altimeter>Barometer Mode>Fixed Elevation. This forces a recorded GPS altitude on the bloated .gpx log file. At least the recorded value seems to be close to GPS altitude. There may be some other side effects that I would have discovered if the thing hadn't failed.

You can do the same thing on a Summit HC (and presumably a Vista HCx and other Garmins with barometric altimeters).

 

By setting it to "fixed elevation" mode, and using a fixed time interval for track-logging, and then leaving the unit at a fixed location for 12 hours or so, you can get a trace of GPS elevation at a fixed location. Do the same again but set the altimeter to "variable elevation" mode. Compare the two traces - in my experience, as long as the altimeter is reasonably well calibrated to start with, I get a MUCH more stable elevation plot using the auto-calibrated altimeter than I ever see with the GPS elevation trace. This is basically why I am convinced that a properly set up barometric altimeter with auto-calibration on really does work.

[+julianh]

The forced barometric altitude use is especially troublesome when hiking near mountain tops, saddles, high winds in mountains, during rapid weather changes, and when in a pressurized aircraft. In each of these cases where the barometer-derived altitude is used for position solution, the lat/long solution can be severely in error; I've seen it miss by miles when in a pressurized a/c. How hard can it be to give us the option to select which altitude to use?

Yes, the barometric altimeter is of no use when you are in a pressurised aircraft, apart from telling you the equivalent cabin altitude air pressure. (I assume the major airlines are not using Garmin consumer hand-held hiking / geocaching GPSrs for their aircraft navigations systems! )

 

However, I have NEVER seen an error with the barometric altimeter cause any accuracy whatsoever with the latitude / longitude fix. As far as I know, the barometric altimeter is NOT an input into the location fix. Quite the reverse - the 3D GPS location fix is used as an input to the altimeter auto-calibrate routine.

[TomOfTarsus]

For my part, I'm wondering how in the world the barometric altimeter could correct itself for barometric pressure changes during the day. Something would have to tell it, "hey, a front just went through, the pressure at elevation x is now 29.91" Hg..."

 

Way back in the olden days, (actually, I think they still do this), as a pilot approcahed an airport, he was given the local barometric pressure, which he then dialed in to his altimeter. I don't see how Garmin's barometric altimeter could compensate without an external source of information.

 

Just curious, I don't even have a unit yet, but I keep reading discussions such as thsi so I'll have better idea when the time has come...

[+julianh]

For my part, I'm wondering how in the world the barometric altimeter could correct itself for barometric pressure changes during the day. Something would have to tell it, "hey, a front just went through, the pressure at elevation x is now 29.91" Hg..."

 

Way back in the olden days, (actually, I think they still do this), as a pilot approcahed an airport, he was given the local barometric pressure, which he then dialed in to his altimeter. I don't see how Garmin's barometric altimeter could compensate without an external source of information.

 

Just curious, I don't even have a unit yet, but I keep reading discussions such as thsi so I'll have better idea when the time has come...

See my Post #9 in thus thread for my best attempt to explain how I think it probably works.

http://forums.Groundspeak.com/GC/index.php...p;#entry3407362

 

There is also a "robust" discussion of the behaviour of Garmin's barometric altimeters here:

http://forums.Groundspeak.com/GC/index.php...p;#entry3389827

 

The unit uses the GPS elevation as the "external source of information" needed to correct for changing air pressure. The algorithm appears to know that GPS elevation is not always accurate or stable, but will tend to oscillate around your true elevation in the long term. I think the algorithm compares the current GPS elevation with the current barometric elevation (probably also using its knowledge of the current spread of satellite positions being received at your location to make an estimate of accuracy and reliability of the current GPS elevation), and adjusts the sea level air pressure to keep the trends compatible, but uses a heavily "damped" correction so as to not over-compensate for excessive fluctuations in the accuracy of the GPS elevation.

 

In my experience, it really DOES succeed in compensating for changes in atmospheric pressure over the course of a day. I have used my altimeter on a number of trips involving some fairly wild weather, including cold fronts moving through, and it has ALWAYS given me good results. (I haven't taken it through a cyclone or up Mt Everest yet!)

 

It doesn't compensate for pressurised aircraft cabins.

 

Error in a moving car at highway speeds can be of the order of 5 to 10 metres (depending on whether you have windows open or closed, etc).

 

Strong wind gusts can give short-term apparent height fluctuations. Others have suggested this may lead to unreliable elevation traces in bad weather at the tops of mountains etc. I guess this is plausible, but I have never seen this effect myself (although I am not a mountaineer).

 

In my experience - in "normal" use, with a basic calibration at the start of the day, with auto-calibration turned on, and with a reasonable GPS fix maintained more or less continuously throughout the day, I almost always see 5 metre vertical accuracy maintained all day, occasionally blowing out to 10 metres under very poor reception conditions with my Summit HC, but this self-corrects when reception conditions improve. (Leave the GPSr turned on all day - don't turn it on and off to save batteries, because the auto-calibration algorithm needs to see continuous records of both the barometer and the 3D GPS fix to allow the trends in changing GPS elevation to correct the barometer for changes in barometer pressure.)

 

Hope this helps.

Edited April 7, 2008 by julianh

[OGBO]

In my discussions with Garmin, they said basically the same thing julianh said, except they gave me the update intervals (basically, at frequent fixed intervals if the unit has a 3D fix, the barometric altimeter is re-set to the GPS-derived altitude). There are several problems with that approach. First, referring back to my discussion of the problem in polar regions, if it really worked as Garmin described, the altitude display should be relatively in agreement with the surveyed altitudes. But it is NOT, at least in latitudes above 70 deg. Second, if it worked as described, the units with this "feature" would not show the behavior I observed of displaying momentarily the GPS-derived or manually set altitude, then within 15 seconds (yes, I timed it) switching to the purely barometrically-derived altitude (yes, it matched the measured absolute pressure, which I was simultaneously measuring). What really happens is that the Garmin algorithm uses the criterion that if the barometric altitude is greater than the GPS-derived altitude by more than a certain amount (roughly 1000 feet if the absolute pressure is less than the equivalent of 7000 ft, meaning lower pressure than that in the ICAO Standard Atmosphere Tables), the barometric value over-rides the GPS-derived value, thus losing the calibration, whether manual or automatic. Of course, this means that this will have no importance for most use in most of the land masses of the Earth except for Antarctica and northern Alaska, Canada, Scandinavia, and Siberia.

 

As also pointed out by others, the displayed altitude shows anomalies in mountainous areas in windy conditions, particularly in the area of passes (due to Venturi effect).

 

But there are anomalies in the GPS-derived altitude as well. The displayed position (3D, meaning lat/lon/alt) is part of the full 4D solution that the unit derives iteratively. If the horizontal position is off (potentially, and by my observation in certain locations) because of multipath, the altitude will also be off - all 4 positions (lat/lon/altitude/time) are interlocked in the solution. The time solution will not be noticed by the user of consumer units, since the time display only shows the nearest second at best, and the deviations are of the order of nanoseconds. The position equivalent is close to one foot per nanosecond. In older units where it was possible to lock into a 2D display (in the case where you "know" the altitude, such as on the open ocean), I ran tests on the horizontal position error induced by entering erroneous altitudes (and later by running through simulations). As a rough rule of thumb, it turns out to be roughly 1 km horizontal error for 1 km offset in altitude (not surprising if you know how the solutions work). In other words, it works both ways - error in altitude (perhaps from using the barometric altitude in polar regions) induces horizontal error, and horizontal error (perhaps from multipath) induces altitude error. Of some comfort, though, the Garmin techs I spoke to stated that the solution is obtained using the GPS values, not substituting the barometric altitude in the computation.

 

I will also note that the early units that displayed barometric altitude allowed displaying either GPS-derived or barometric altitude. So it was a case of eliminating the choice of displaying the GPS-derived altitude

Edited April 7, 2008 by OGBO

[+julianh]

OGBO,

 

Thanks for an informative update. I have a couple of questions and comments:

 

Firstly, I can't comment on the behaviour of current units at high latitudes, not yet having visited these areas with my Summit HC.

 

How long ago was your conversation with the Garmin techies? The reason I ask is that my current model Summit HC with current software gives MUCH better barometric altimeter performance than my older B&W Vista (which I have since disposed of, so I can't run side-by-side tests). Of course, the Summit HC has a high sensitivity receiver (which the B&W Vista did not), and this improves the quality of the 3D fix, and this undoubtedly helps overall elevation stability as well.

 

I suspect the latest Garmin software revisions improve the auto-calibration algorithm. In particular, I think my old Vista used to appear to over-respond to fluctuations in the GPS elevation, whereas my Summit HC with current software appears to be quite heavily "damped" in this respect, so it seems to respond to significant trends in GPS elevation, but not to short term changes (e.g. when you walk into a location with poor satellite visibility, so you get sudden apparent GPS elevation shifts due to blockage and / or multi-path errors). That is, by a combination of using barometric pressure for short-term elevation changes, underpinned by longer-term auto-calibration using "damped" GPS elevation, the unit gives me excellent behaviour all day.

 

I have NOT seen the barometric altimeter shift to match the GPS elevation - even instantaneously. Rather, my own analysis of some test barometric elevation traces and corresponding GPS elevation traces leads me to conclude that the unit periodically auto-calibrates the barometric altimeter TOWARDS the GPS elevation, but not to actually match it. This may also be a change in the newer algorithms?

 

(I will gladly attach some sample traces, if someone can tell me how! The following discussion relates to some testing I conducted on two consecutive nights. T tried to upload some plots, but couldn't work it out - sorry!)

 

I set my GPSr up at the same fixed location on two consecutive nights (so the visible satellite constellation is essentially the same on both traces). On the first night, I recorded GPS elevation, and on the second night, I recorded the auto-calibrated barometric elevation. The location has "fair to good" satellite visibility - partially blocked to the East by my house, some partial tree cover to the North and West. I was attempting to simulate "real world" observations, not "best possible" conditions. The true elevation is 50 metres above sea level.

 

The GPS elevation trace shows sudden spikes from time to time, presumably as satellites drop out of visibility. The barometric elevation trace is much more stable, and responds only slowly to the effect of poor satellite visibility giving a poor GPS elevation. There is a period of very poor GPS elevation at about 4:00 am to 4:30 am - presumably, I had a very poor visible satellite constellation for this period. This also affects the auto-calibrated barometer elevation, but there is a significant time lag, and the maximum error in the barometric elevation is a lot less.

 

In "typical real world" conditions, I generally get significantly better performance than this, because even though the instantaneous satellite visibility can vary a lot as you move in and out of tree cover etc, the GPSr can typically track all satellites currently visible, even though individual satellites may drift in and out of view periodically, whereas at my fixed location, some satellites were totally blocked for significant periods (or subject to significant multi-path errors) due to my fixed location with partial blockage.

 

My 2D position in both cases was absolutely "rock solid" - maximum position drift was only a couple of metres. I don't believe the barometric elevation (calibrated or uncalibrated) is used in any way as part of the 2D position solution. Deliberately entering a wildly erroneous elevation estimate (or when travelling in a pressurised aircraft, for example) seems to have no effect on my 2D fix accuracy at all.

 

Yes, when you are in areas of high wind speed, or travelling in a car, this will affect the barometric elevation, because there is a direct correlation between wind speed and air pressure. In my experience, short gusts are not an issue, because the "damping" seems to handle this, but I guess sustained high wind speeds would correlate to an apparent rise in elevation. I haven't had the opportunity to test this. Travelling in a car at 100 km/hr can lead to an apparent elevation shift of about 5 to 10 metres - still not bad, compared to the accuracy of GPS elevation!

 

Yes, when your 2D position accuracy is poor (limited satellite visibility, multi-path errors, etc), the GPS elevation accuracy suffers badly. However, in my experience, because the unit seems to only use longer term trends in GPS elevation to auto-calibrate the barometer, this short-term effect does not seem to carry over to the auto-calibrated barometric elevation accuracy.

 

Yes, it might be nice to be able to see GPS elevation AND barometric elevation at will. By all means, keep pushing Garmin to get this feature activated. However, for my purposes, the auto-calibrated barometric elevation works great!

 

Hope this helps! (And sorry I couldn't work out how to attach the plots!)

[+TrailGators]

My 1st GPS had a barometric altimeter. My new one, 60cx doesn't. I wanted a GPS that would tell me my elevation but instead I had to keep telling the one with the altimeter what its elevation was. I gave up, now I just use the GPS elevation. Close enough !

This is how I feel about it too. I never use the barometric altimeter anymore.

[TomOfTarsus]

(with intervening post by Trailgators)

 

Well, that's my point. Without some external source of information, there's no way a barometric altimeter can compensate for changes in barometric pressure that isn't due to a change in altitude. It's merely a pressure sensor. So I guess the whole idea is to start well, then have the GPS-derived altitude update the barometer as needed.

 

Now on the Rino, they could use data from the weather radio, could they not? Maybe not the alert channel, but aren't there local weather station dat brodacasts (encoded, not voice) that give the current barometric pressure (corrected to sea level, etc). I know next to nothing about that, just taking a shot.

 

And yes, I'd like to be able to see both GPS and barometric altitde figures, as well as the choice, if I ever get a unit.

[OGBO]

....

How long ago was your conversation with the Garmin techies?.

 

.......

I have NOT seen the barometric altimeter shift to match the GPS elevation - even instantaneously.

 

.......

The GPS elevation trace shows sudden spikes from time to time, presumably as satellites drop out of visibility.

 

....

My 2D position in both cases was absolutely "rock solid" - maximum position drift was only a couple of metres. .....

 

ONly time for some brief responses -

 

Discussion with the Garmin tech folks started shortly after I got back from Antarctica (Jan 2007) and continued until sometime in Aug 2007, when I gave up on getting an informative or rational response.

 

You generally will not see the barometric altimeter update unless you watch the altimeter continuously through a couple of full cycles. I saw it in Antarctica on my 60CSx and on a couple of other people's Vistas and 60CSx's, because it is so obvious under the high latitude/high altitude conditions. In general, if you set the barometric altitude with the "Do you know the correct altitude?" option at lower 48 altitudes/latitudes, the offset remains fixed (even if you set the altitude too high by 10,000 ft - recall the over-ride I described happens when the barometer reads a pressure that is too low by a large percentage compared to what it expects from the GPS-derived altitude, and apparently only at high latitudes and high altitudes).

 

Yes, there is a jump in 3D position (both altitude and lat-lon) when one of the "best 4" satellites is dropped and replaced by another visible SV (even when the new one is part of the "over-determined" group). Lots of reasons for this, but not enough time or space to go into it now.

 

You cannot place the current (last 4 or 5 years' models) of Garmin (or some other manufacturers) GPSRs in a true 2D mode. Even if you set the barometric altimeter to some wildly different altitude (via the "do you know the correct altitude" selection), the unit will show "no fix" if you have only 3 SVs (which you have to be stuck at for true 2D, which means you are locking the altitude). If you try to set a mark when you have 3 or fewer SVs (say, right after you turn the unit on), you get a warning message ("Do you want to proceed anyway?" which you might want to do if entering a known lat-lon from a book or map). One way to see this is to offset the altitude by a large amount (while you do have a fix), press "Mark", then select "Average" and look at the altitude - it will be averaging on the GPS-derived altitude, not the off-set altitude. If for some reason you want the saved waypoint to be at a different altitude, you have to edit the altitude in the waypoint. So your observation of accurate 2D horizontal position is really a 3D position.

 

I haven't been back to Antarctica and won't be until at least 18 months from now (or more likely a year beyond that), so I don't know if Garmin has made an appropriate change in the firmware. But there is no mention of it in the list of changes for the last couple of releases. Besides, the 60 CSx and Vista, etc models are "obsolete" now. Who knows if the Colorado or whatever the successors will have the change. Garmin won't tell us about bugs that were fixed between models (except for the usual "NEW! IMPROVED!" comments - yeah, "improved" how?)

[OGBO]

....

I have NOT seen the barometric altimeter shift to match the GPS elevation - even instantaneously. ....

 

Try this - Go to the Calibrate Altimeter window. Select "Know Correct Altitude?". Enter an altitude at least 2000 ft lower than your actual altitude and select "OK", and when it says "Successful" enter "OK". Then watch the altitude window (it helps to start with a page that has an altitude window, like the trip altitude profile where you have set a window to show altitude. Even if you have "autocalibrate" turned off, you will see the altitude reset (probably to a slightly incorrect value).

[+julianh]

....

I have NOT seen the barometric altimeter shift to match the GPS elevation - even instantaneously. ....

 

Try this - Go to the Calibrate Altimeter window. Select "Know Correct Altitude?". Enter an altitude at least 2000 ft lower than your actual altitude and select "OK", and when it says "Successful" enter "OK". Then watch the altitude window (it helps to start with a page that has an altitude window, like the trip altitude profile where you have set a window to show altitude. Even if you have "autocalibrate" turned off, you will see the altitude reset (probably to a slightly incorrect value).

Interesting ...

 

The behaviour of my Summit HC with current unit software is quite different.

 

If I deliberately enter a wildly inaccurate calibration elevation, the unit just shows the manually entered elevation continually (plus or minus about 5 metres), and will do so indefinitely (several hours at least). It does NOT show the GPS elevation, even for an instant. This is also basically what I see on pressurised aircraft - the barometric altimeter just "maxes out" at about 2,000 metres elevation (presumably, that is the cabin pressure in jet airliners at cruise altitude?), and continues to show that elevation until the plane drops back down to earth again.

 

However, if I just enter a "reasonable" but very bad guess for my calibration altitude (e.g. an error of 50 or 100 metres, say), the unit starts correcting immediately, but not very rapidly. Even so, it does NOT show the GPS elevation, even for an instant - and yes, I have watched it continuously for 10 minutes or so through this process, and have taken continuous elevation traces at 30-second intervals as it corrects back to true elevation.

 

An initial calibration error of about 100 metres will take about 1 hour to self-correct, and the elevation display shows a "decay curve" trace back to the true elevation, suggesting some sort of "damped" correction, such that the unit corrects the altimeter toward the current GPS elevation, but limiting the short-term correction rate so as to not over-compensate for poor satellite visibility etc. At no time does the altimeter trace match the GPS elevation - at least not until the auto-calibration process has brought the two elevation figures into the same range, and then the barometric altimeter trace and GPS elevation will more or less agree with each other, apart from when the GPS elevation gets significant errors due to poor satellite visibility, etc.

 

My guess is that if the algorithm cannot correlate between the manually-entered calibration elevation and the automatically determined GPS elevation, the algorithm does not attempt to adjust the barometric altimeter. However, when the two elevations are in "reasonable" agreement, the GPS elevation is used as a "damped base line" to "anchor" the barometric altimeter, thereby compensating for changes in barometric pressure.

 

Regardless, of whether the barometric altimeter can auto-calibrate or not (depending on whether the initial calibration error is "huge"), at no time does the elevation error have any effect on the calculated latitude and longitude at all. That is, I don't believe the barometric altimeter is used as part of the location fix in any way.

 

Like I said before, this is all at "moderate" latitudes (currently 27 degrees south), and using auto-calibration turned on. I am not particularly interested in the behaviour without auto-calibration, as this seems to defeat the intended mode of operation of the unit.

 

I can't speak for the behaviour at extreme latitudes, and I haven't seen a 60 CSx with current software. I believe the 60 CSx has a different GPS receiver to the Summit HC, and presumably the software is different as well. However, I have certainly seen that my new Summit HC is MUCH better behaved in this respect than my old B&W Vista (which, while an older model, was bought brand-new just a couple of months ago), so I suspect that Garmin have improved the auto-calibration algorithm in recent months / years.

[OGBO]

Somewhere (I thought it was in this thread, but don't see it right now), possibly somewhere else on the internet dealing with this or a similar question, I noted that when averaging a waypoint, the altitude saved for the 60CSx is the GPS-derived altitude, not the barometric altitude. This implies that the the displayed 2D location is derived using the GPS-derived altitude (the only sensible thing to do), not whatever the barometeric altimeter is displaying (no matter how calibrated and no matter how erroneous an offset). This explains why the experiments you are doing do not seem to be affected by whatever you do to the barometric altimeter setting - the unit is NOT using the barometric altimeter for position calculations at all, only the GPS-derived value, no matter what the display shows (In other words, Garmin is not letting you see what is really going on behind the displays). The way to see this is to offset the altitude (which is the barometric altimeter) by some large amount, mark a waypoint then go to "average" and watch.

 

This is also the explanation for the weird EPE circles compared to the displayed position error. Garmin has always been very secretive about the meaning of their "position error" - not saying they are the 1 sigma, 2 sigma, or whatever error circles, not saying what the DOP of the constellation configuration is, or much of anything else. "Just believe and accept what we let you know".

 

By the way, you say the Vista does not do what the 60CSx does when offset by a large amount. Were you careful to set the altitude calibration ("Known altitude" setting) to a large amount below your actual altitude? In other words, if you are close to sea level as my house is, did you set the altimeter to minus 2000 ft or some other value a couple thousand feet lower? Or did you forget the minus sign? The 60CSx behaves differently for altitudes much lower than for those much higher, as far as the autocalibrate and averaging are concerned.

Edited April 14, 2008 by OGBO

[+julianh]

By the way, you say the Vista does not do what the 60CSx does when offset by a large amount. Were you careful to set the altitude calibration ("Known altitude" setting) to a large amount below your actual altitude? In other words, if you are close to sea level as my house is, did you set the altimeter to minus 2000 ft or some other value a couple thousand feet lower? Or did you forget the minus sign? The 60CSx behaves differently for altitudes much lower than for those much higher, as far as the autocalibrate and averaging are concerned.

You are right - I had never tried deliberately setting the initial calibration altitude to a value that is way too low - I had only experimented with elevations that are too high. (In part, because my typical elevation is between about sea level and 100 metres above sea level - setting a level of more than 500 metres below my true level typically puts me into significantly negative elevation, and this is outside the expected working range of a GPS. (Lowest point in Australia is Lake Eye at -15 m).

 

Having now tried, it seems that the unit will accept an elevation up to about 400 metres lower than my true elevation, and will then auto-calibrate it's way back to a realistic elevation using the process as I described before. However, if I try to perform an initial calibration with an elevation that is 500 metres or more too low, then it jumps straight to an elevation value that is "roughly" correct, and continues to auto-calibrate from that new elevation. Interestingly, as far as I can tell, the elevation it jumps to is NOT the current GPS elevation - I have no idea where it gets the number from.

 

I can't help wondering whether the algorithm is based on assuming you are somewhere on the Earth's surface (not an reasonable assumption for a hand-held hiking unit!), and it knows that the lowest place on the planet is the Dead Sea at about -418 metres? Coincidentally, -420 meters seems to be the lowest elevation I can set as an initial calibration, before it will jump straight back to a figure near sea level. (Just speculation.)

 

Anyway, it's all a bit academic for me. As far as I am concerned, with a basic calibration at the start of the day, and with auto-calibration turned on, I get excellent elevation accuracy with my Summit HC in normal use, from sea level to 3,500 metres (apart from in pressurised aircraft), at "moderate" latitudes. Should I ever get to visit the high Arctic / Antarctic, I will be sure to report back on my observations!

[+julianh]

I noted that when averaging a waypoint, the altitude saved for the 60CSx is the GPS-derived altitude, not the barometric altitude. This implies that the the displayed 2D location is derived using the GPS-derived altitude (the only sensible thing to do), not whatever the barometeric altimeter is displaying (no matter how calibrated and no matter how erroneous an offset). This explains why the experiments you are doing do not seem to be affected by whatever you do to the barometric altimeter setting - the unit is NOT using the barometric altimeter for position calculations at all, only the GPS-derived value, no matter what the display shows (In other words, Garmin is not letting you see what is really going on behind the displays). The way to see this is to offset the altitude (which is the barometric altimeter) by some large amount, mark a waypoint then go to "average" and watch.

 

This is also the explanation for the weird EPE circles compared to the displayed position error. Garmin has always been very secretive about the meaning of their "position error" - not saying they are the 1 sigma, 2 sigma, or whatever error circles, not saying what the DOP of the constellation configuration is, or much of anything else. "Just believe and accept what we let you know".

Yes, as far as I can tell, when you first create a waypoint, it is saved using the barometer elevation, but if you then average its position, it starts to use the GPS elevation instead.

 

GPS elevation is always available (through the satellite page), whenever you have a location fix (because as you rightly point it, the unit won't give you a location until it has a full 3D fix). In general, your auto-calibrated elevation should be more reliable and stable than your GPS elevation (which is subject to instantaneous vagaries of poor satellite visibility, multi-path errors, etc, whereas your auto-calibrated barometric elevation will stay current and reasonably accurate even if you move indoors out of sight of all satellites).

 

I don't know exactly what the radius of the EPE circles on the map page represent (in terms of 50% / 90% confidence, or whatever), but I don't think the radius of the EPE circle has any relationship to whether you have a barometric elevation or not - I think these are based solely on GPS location and the unit's calculation of EPE / DOP, which is not affected by barometric elevation.

[OGBO]

julian, I think we are on the same page now. As for what the display jumps to when you set an elevation far too low, I suspect that it is the altitude of the absolute pressure. And I am now suspecting that it is not affected by the latitude as I had originally guessed. What I think is going on is this - If you set the altitude too low, you are telling the unit that the absolute pressure is higher than the barometric sensor is measuring. It does not matter whether the calibration is from "known altitude", "known barometer", or "use GPS-derived altitude". As long as the pressure actually measured is no more than a certain amount less than the value set, that is the altitude set is more than a certain amount lower (your guess of the Dead Sea is probably a good one), the displayed altitude will be happy. If you set a higher altitude, meaning that in effect you are telling it the "real" pressure is lower than what it is measuring, the algorithm is happy and won't switch to the barometric altitude. But if the difference is more than the amount designated, the unit will switch to the ICAO table value for the measured absolute pressure and auto-calibrate from then on (if autocalibrate is turned on).

 

This would explain (1) why there is no switch when the altitude is too high (you might be in a pressurized cabin of a plane) and (2) why the switch shows up at ground level when you are in polar regions where the pressure-altitude relation is substantially different from the lower latitude relationship (that is, the pressure at a given physical altitude is substantially lower than the Standard Atmosphere Table gives). So what I suspected was a latitude effect programmed into the GPSR is not that at all, but is in fact the physical effect of latitude on the atmospheric lapse rate. This would explain why I did not see the switching on Denali (the thinning of the atmosphere is not great enough at the 60 deg latitude of Denali to trigger the switch, even at the summit), but do see it on Mt Vinson (the thinning of the atmosphere is great enough at the 80 deg South latitude of Vinson to show up as a large enough lower pressure difference to trigger the "correction".

 

There was a similar problem some years ago with Garmins (but not Magellan, Trimble, or Lowrance) where Garmin GPSRs would not show altitudes below sea level. Their lowest altitude displayed (and it was a GPS-derived altitude) was 0. Garmin was a boating and aircraft instrument company at that time, so "no one" of their customers would ever be below sea level. Those of us in California who would go to Death Valley or the Salton Sea complained to Garmin for several years before they started allowing altitudes below sea level.

 

By the same token, I guess they are right - there are probably only a few hundred people in the world who would notice the problem with the displayed altitude because they are in one of the few mountain ranges in the Arctic that are high enough or in the mountain ranges in Antarctica or on the South Polar Plateau, where the effect also shows up. Still, how hard can it be to allow showing the GPS-derived altitude in the display window (rather than having to search it out as at present), or to allow calibration of, say 2000 meters lower than what the absolute pressure says (that would cover all surface locations, though not aircraft flying in polar regions).

Edited April 14, 2008 by OGBO

[RGK]

The forced barometric altitude use is especially troublesome when hiking near mountain tops, saddles, high winds in mountains, during rapid weather changes, and when in a pressurized aircraft. In each of these cases where the barometer-derived altitude is used for position solution, the lat/long solution can be severely in error; I've seen it miss by miles when in a pressurized a/c. How hard can it be to give us the option to select which altitude to use?

Yes, the barometric altimeter is of no use when you are in a pressurised aircraft, apart from telling you the equivalent cabin altitude air pressure. (I assume the major airlines are not using Garmin consumer hand-held hiking / geocaching GPSrs for their aircraft navigations systems! )

 

However, I have NEVER seen an error with the barometric altimeter cause any accuracy whatsoever with the latitude / longitude fix. As far as I know, the barometric altimeter is NOT an input into the location fix. Quite the reverse - the 3D GPS location fix is used as an input to the altimeter auto-calibrate routine.

 

Translated to Aussie-talk to accommodate the replier :

 

Have you ever actually had your unit on a pressurised aircraft? If so, a review of track logs would show you that as the a/c moves from sea level to cruising altitude the recorded position takes big jumps as true altitude moves away from (or toward) cabin pressure altitude. This is prima facie evidence that Garmin are forcing baro altitude into the position solution. It is not possible for a civilian a/c to change position at the recorded rates without shedding its wings! Besides that, basic aerodynamics would require high-g banking to change position at the recorded rates, I would have noticed such a manoeuver.

The fact that the unit can compute and display GPS-derived altitude makes it all the more maddening, the capability is there and Garmin stubbornly deny us the option to use it.

 

BTW, I remain absent my 300, sent it to Garmin over a month ago, I was promised a 10 day turn around, it is now 4 weeks and no 300, they haven't even had to courtesy to email an excuse or ETA, 2 requests to their product support group have gone unanswered. I'm stacking on a blue of monumental proportions.

[+julianh]

Have you ever actually had your unit on a pressurised aircraft? If so, a review of track logs would show you that as the a/c moves from sea level to cruising altitude the recorded position takes big jumps as true altitude moves away from (or toward) cabin pressure altitude. This is prima facie evidence that Garmin are forcing baro altitude into the position solution. It is not possible for a civilian a/c to change position at the recorded rates without shedding its wings! Besides that, basic aerodynamics would require high-g banking to change position at the recorded rates, I would have noticed such a manoeuver.

The fact that the unit can compute and display GPS-derived altitude makes it all the more maddening, the capability is there and Garmin stubbornly deny us the option to use it.

Yes,I have taken my Summit HC on pressurised jet and turbo-prop aircraft, several times, and no, this is not what I generally see on my track logs. (I have just gone back and reviewed several track logs that I have saved from recent flights.)

 

As long as I am getting a reasonable signal (and that basically means if I have a window seat, and have the unit in my shirt pocket or on the tray table, for example), my track-logs are generally smooth and continuous, with occasional drop-outs, because the signal inside an aluminium tube is far from perfect! The paths are generally straight and true, and while I can't independently confirm the accuracy to within a couple of metres, what I see looking out the window exactly correlates with what the map display shows me. Significant bends in the flight-path typically happen directly above civilian airports, and this makes sense to me as the basis for planning trans-continental flight-paths in Australia.

 

However, now that I look closely at some sections where we are coming in to land, or shortly after take-off, I can see the apparent jumpiness you describe at a couple of points. The biggest sideways "jumps" in my track-logs are a couple of hundred metres (not miles) - basically the same size leg length as the previous and following legs, but the bearing has suddenly shifted - but I guess this could be affected by how you have your track-log intervals are set up. (Of course, I have no way of knowing which is more accurate - the track-point at the start of the jump, or at the end of the jump - or neither.) I had previously dismissed this effect as being due to poor reception as I was packing the unit away (e.g. signal dropping to just 3 or 4 closely-grouped satellites, all in one sector of the sky), and looking closely at the track-logs again now, I am not convinced this is not the case. Looking closely at the elevation profile for one jumpy part of the track, the elevation is showing as dropping from 0 metres to about -80 metres elevation over a "wonky" section. I can assure you this was not the case - while we were certainly dropping, we were still several hundred metres above the ground (and the sea!)

 

However, I guess it is possible that rapid increase in cabin air pressure was confusing the barometric altimeter, and maybe this was having an effect on the 2D fix, once the GPS unit decided that the ambient air pressure and the GPS elevation were close enough to each other to have a go at producing a correlated barometric altimeter elevation / GPS elevation.

 

Anyway, like I said, when used within the circumstances for which the unit is designed (i.e. outdoors, no artificial pressure), I am totally happy with the accuracy of my barometric altimeter AND my location.

Edited April 15, 2008 by julianh

[RGK]

 

BTW, I remain absent my 300, sent it to Garmin over a month ago, I was promised a 10 day turn around, it is now 4 weeks and no 300, they haven't even had to courtesy to email an excuse or ETA, 2 requests to their product support group have gone unanswered. I'm stacking on a blue of monumental proportions.

 

 

Shortly after sending this complaint, a replacement CO arrived on my porch. It was 4 weeks to the day after I had sent it in for repair. A quick check of battery life shows that it still shuts down at a relatively high battery voltage, there is a lot of energy left in those 'dead' batteries. SO I am skeptical that battery life has been fixed. Will do a more comprehensive test before sending it back for a refund.

[Pacific Stereo]

I am simply dumbfounded that the Garmin people remove the option of using GPS altitude instead of barometric altitude. There is absolutely no reason for this. When I got my 60Csx, the Cx model was a few bucks cheaper, and now I am wishing I had gotten it. While I love the altitude history feature of the Csx (which the Cx lacks), I am furious that I can't turn the bloody barometric altimeter off when I am in an airplane!

 

There are a number of reasons why someone would want GPS altitude, and it is completely retarded that you can't have the unit do this. If Garmin thinks it's too complicated for ordinary people to sort out, then make it simple. Add "airplane mode" to the altimeter setup page so that we can not be stuck at 8000 feet for the duration!

 

My guess is that it might take one of their programmers about 15 minutes to sort this out. Oh, wait... add a menu item. 20 minutes, then.

 

I am not a happy camper.

 

-Pacific

Edited June 17, 2008 by Pacific Stereo

[Pacific Stereo]

I decided to add my name to the virtual list at Garmin of people asking for GPS altitude support in the sensor series of GPS products by submitting a support request.

 

The tech at Garmin says, "Oh, yeah, we've already done that. If you want GPS altitude, just go to the satellite page and press Menu, and then GPS Altitude."

 

Hello? (Knocks on monitor) Anyone in there?

 

Good Lord. That tech knew exactly what I wanted, why I wanted it and why I was asking about it. And that's the response?

 

Probably my first and last Garmin product.

 

-Pacific

[Hertzog]

I decided to add my name to the virtual list at Garmin of people asking for GPS altitude support in the sensor series of GPS products by submitting a support request.

 

The tech at Garmin says, "Oh, yeah, we've already done that. If you want GPS altitude, just go to the satellite page and press Menu, and then GPS Altitude."

 

Hello? (Knocks on monitor) Anyone in there?

 

Good Lord. That tech knew exactly what I wanted, why I wanted it and why I was asking about it. And that's the response?

 

Probably my first and last Garmin product.

 

-Pacific

As you say, this has been a much requested and discussed feature that some of us would like to see added; should be simple as you indicate, but Garmin seems to be reluctant to add it for some reason.

 

Meanwhile, there is a work-around that is available with the 60CSx (don't know about the other "Sx" models: If you are flying, put it in "Fixed Elevation" mode; in this mode the GPS altitude will be logged in the active track and the gpx file in place of hte barometric altitude. Not perfect, as the data fields will still display only the barometric altitude, but at least you get a record of the filght (you can also see it in the elevation profile, but only if you play around with the scroll button).

[gallet]

IMO.... leave it to auto calibrate using the GPS signal.

 

 

Seconded

[Pacific Stereo]

As you say, this has been a much requested and discussed feature that some of us would like to see added; should be simple as you indicate, but Garmin seems to be reluctant to add it for some reason.

It really does not make any sense. I've got two out of the three original founders of Magellan sitting across the room from me, we all write code, and we are all scratching our heads.

 

I did get another response from support (very nice guy, I must say) which basically said, "Oh. Now I understand. Yes, for some reason, Engineering thinks differently about making that available."

 

It looks like we will never get our wish.

Meanwhile, there is a work-around that is available with the 60CSx (don't know about the other "Sx" models: If you are flying, put it in "Fixed Elevation" mode; in this mode the GPS altitude will be logged in the active track and the gpx file in place of hte barometric altitude. Not perfect, as the data fields will still display only the barometric altitude, but at least you get a record of the filght (you can also see it in the elevation profile, but only if you play around with the scroll button).

Oh, this is FABULOUS! This is exactly what I want. I can live with the fact that I don't have correct altitude unless I fiddle, but I want the recorded track to be correct!

 

Now why didn't support tell me about this when I asked for a work-around?

 

Thank you, thank you. I will report results the next time I fly. Yay!

 

-Pacific

[gallet]

Brilliant, I was searching the forum for this post that I knew was here and couldn't find it. The gf's coming back on a plane today!

 

 

My guess is that it might take one of their programmers about 15 minutes to sort this out. Oh, wait... add a menu item. 20 minutes, then.

 

Nah, that's way off buddy, it couldn't be done in less than half an hour to 45mins ;¬)

[Pacific Stereo]

I can report that the "fixed altitude" workaround WORKS! I'm sorry that it took so long to do the test. Two flights ago, I forgot my amplified antenna (the built-in on the 60 doesn't work well inside the airplane) and couldn't get a decent fix. But this latest trip, I brought the antenna, and did the log. The track does indeed show the CORRECT altitude.

 

Fabulous! Thanks again.

 

-Pacific

[Pacific Stereo]

Ugh. Dupe. The database timed out for some reason, and the reply got double-posted.

Edited November 18, 2008 by Pacific Stereo

[+JSWilson64]

But this latest trip, I brought the antenna, and did the log. The track does indeed show the CORRECT altitude.

 

It shows the correct altitude, or the GPS altitude?

 

<rimshot!>