66s elevation data question, too accurate?

[wildgoose]

I recently acquired a used 66s. It's running on latest firmware 10.70 with a master reset.

Coming from the classic 60csx, I noticed an anomaly with the GPS's elevation field. It seems too accurate to be true.

I calibrated it the other day to the correct known elevation at my home, and I noticed now that it always shows the exact calibrated elevation at my home, no matter the circumstances (coming back from somewhere, or leaving home, or simply using the GPS indoor with satellite off).

Is the GPS smart enough to remember the calibration value at that location? Otherwise it's hard to believe it'll derive the exact same elevation from altimeter and/or GPS signal every time at this location. If so, how does it work in general? If the GPS is turned on when I am at home, does this mean the unit would quickly calibrate the barometer altitude to this known elevation automatically? That would be pretty neat and a huge improvement compare to 60csx.

Is there a way to see if any other location have a calibrated elevation value? What happens when I calibrate it at another location? Does it still remember the calibration value for my home location?

Also, when I move the cursor on the map screen and press enter, it'll show the elevation of the selected point, be it an POI, or simply blank area. Where is that elevation data coming from? (I have installed OpenTopoMap for Garmin. The elevation data is available whether I selected or de-selected this map in map setup. But not if I remove the SD card from the GPS).

Does the GPS uses the map elevation data at other times for auto-calibration? ie, if I turn it on at a completely new location after a flight, does it use Map elevation data for calibration, or solely based on barometer pressure difference from last known value?

[GPSrChive]

Is the GPS smart enough to remember the calibration value at that location? Otherwise it's hard to believe it'll derive the exact same elevation from altimeter and/or GPS signal every time at this location.

While this information is not found in any known documentation, your observations appear to be correct.

If so, how does it work in general? If the GPS is turned on when I am at home, does this mean the unit would quickly calibrate the barometer altitude to this known elevation automatically? That would be pretty neat and a huge improvement compare to 60csx.

Based on historical observations, this is likely a very safe assumption to make.

Is there a way to see if any other location have a calibrated elevation value? What happens when I calibrate it at another location? Does it still remember the calibration value for my home location?

One could easily test this if they commonly visit additional locations with different elevations that are calibrated in the same manner.

[Przekątny]

Also, when I move the cursor on the map screen and press enter, it'll show the elevation of the selected point, be it an POI, or simply blank area. Where is that elevation data coming from? (I have installed OpenTopoMap for Garmin. The elevation data is available whether I selected or de-selected this map in map setup. But not if I remove the SD card from the GPS).

If after removing the SD card on which OpenTopoMap is placed there is no altitude data, it means that there is no gmapdem.img and base map in the device.

The altitude data is read from maps containing DEM data,
the map with DEM data does not have to be turned on, it is enough that it is in the device.

The remaining considerations have no justification.

The altitude data from the altimeter is used only during activity recording.

Did you have the recording turned on in the described situations.

The altitude measurer takes into account changes in atmospheric pressure related to changes in altitude above sea level, so the altitude value is indirectly related to the place you mentioned, but this place is not related to performing calibration in this location.

[GPSrChive]

Any time you are at ground level, your altitude is always zero.

The vertical distance to MSL is your Elevation.

The altitude data is read from maps containing DEM data,

What does the 'E' in 'DEM' reference?

[Nail]

What does the 'E' in 'DEM' reference?

You probably know the answer.
Digital Elevation Model

[wildgoose]

Looks like the unit auto calibrates to the map DEM elevation in quite a few cases (with auto calibration set to Once), even when user has not explicitly calibrated it at that location.

For example:

• When powered on, and GPS acquires a location lock.
• When switching to GPS simulation mode.
• When ending an activity.

Today at a different location, on the top of a parking structure, I turned the unit on, and it got a location lock, and indicated elevation of 20ft. Which is obviously too low, since the building itself is several stories. I moved the cursor on the map slightly, and sure enough the unit said the elevation at that point is 20ft, which is what the GPS calibrated itself to at power on.

For the entire drive from the parking structure to home, the shown elevation is about 20-30 ft lower than actual, and when I arrived home, it was showing a lower elevation as well. But as soon as I stopped the activity, it reset to the correct elevation. In the middle of the drive, I turned GPS reception off/on again, that did not reset the location. Only stopping the activity did.

Overall it seems pretty smart and I like it. The moral of the story, one should be at ground level when turning on the GPS.

[wildgoose]

While the elevation data field is fairly accurate, during an actual hike, I noticed Total Ascent and Total Descent to be TOTALLY off.

Did a hike today in an area with almost complete open sky, with very mild gentle elevation gains. Elevation range is 157ft to 506ft. However total ascent on GPS shows at 1250ft, which is significantly higher than what it should be.

Spoiler

134.png

143.png

When I examine the track on Garmin Explore on phone, I can see why the total number maybe so high. There are a lot of noisy micro elevation up/downs, while the overall elevation plot when you zoom out is fairly smooth.

Spoiler

20250405_175707_1477_iphone13pm_ray.PNG

Why are there so many noisy elevation up/downs on the graph? If the values are smoothed out then I think the total number will be more correct.

Would it be a device issue? It's running on latest software GPS-2.9/10.7 version on a 66s. I suppose this is bug-109?

Anyone know if the same issue exist on the 67?

[JungleJim]

The GPSMAP 67 also suffers from this issue. It has been discussed extensively, see
Total Ascent Values Exaggerated by 50% [Bug 29].

On the GPSMAP 67 elevation is also quite accurate, but ascent and descent seems to incorporate every small elevation change.

Also the ascent/descent data fields show data changes even if no activity is being recorded. The manual indicates that this should not be the case. In the FIT file only ascent and descent as registered during activity recording is recorded though.

Garmin did release a firmware update for the 67 which included an improvement for ascent and descent. It has gotten a bit better but the value can still differ from what would be expected.

[wildgoose]

TLDR: multiple keypress in quick succession (such as when cycling through pages) will create these micro elevation changes during an activity.

Details

I found something interesting today. I saw an earlier activity where the 1st half had lots of small elevation change, but the 2nd half was pretty smooth. Now this was me just walking around the neighborhood in loops, with no tall buildings around and excellent sky view.

As I couldn't recall exactly what I did to the GPS in the middle back then, I went out for a similar walk today to see if I can reproduce the micro elevation changes. I did 4 loops total.

Here's the elevation data:

Spoiler

IMG_1491.PNG

I didn't actually do the experiment I wanted to do, which is to compare GPS-only vs with GLONASS (my theory). But instead, I did GPS only, but with and without WAAS.

But if you think WAAS was the culprit, you are sadly mistaken.

- loop 1 and 2: GPS-only, with WAAS enabled. (lots of micro elevation change up/down)
- loop3, GPS-only, WAAS disabled (relatively smooth)
- loop4, GPS-only, with again WAAS (still smooth)

The ONLY difference between 1st half of the hike, and 2nd half, is I realized a bit into loop 3 that, any quick multiple key press on the GPS (for which I did a lot of during the first half of the hike to cycle through pages), produced one of these micro elevation changes that lasts quite a few seconds. ie, elevation will jump down then back up, and settle down after 5-10 seconds.

Once I realized this, I simply stopped pressing any key on the GPS for the remaining of the hike (holding it in my hand like a true scientist), and that resulted in the much smoother elevation data. The elevation reading on the GPS did not change much during my walk.

The jump in elevation is about 30ft (min to max), before settling down (when viewed on the phone in Explorer).

This is not reproducible when the device is not recording an activity. Key press simply don't change elevation at all.

To reproduce this, go to map page, with small data fields so you see elevation.

• find a location with good GPS coverage. Wait a minute for reading to stabilize.

• press [enter] and start an activity.

• wait for elevation to stabilize (10-20s?)

• now press [QUIT] multiple times to cycle through the pages (I had ribbon turned off).

Observe:
* After you cycle back to the map page (say in about 1-2s), you'll notice the elevation reading to be a bit lower than what it was before.
* It'll then start to climb back up after a few seconds...

The multiple key press in fast succession is needed to more easily reproduce the issue. If you simply press QUIT then PAGE to go back to map page, it doesn't show up as much.

I believe the multiple key press creates a slight positive pressure inside the unit, resulting in a reading that indicates a slight elevation drop (due to higher pressure). The pressure difference stabilizes within a few seconds, afterwards which the reading goes back to normal.

Is this a known issue, or have I discovered something interesting? Or is my unit defective?

[GPSrChive]

Very interesting....

Some initial testing with a 66i does result in dropping my stationary elevation reading 10-20 feet after some aggressive button mashing for ~ 30 seconds, and once I stop, the elevation does slowly readjust back toward the original value.

I am curios what your Setup > Altimeter and Setup > Recording settings are during these tests?