GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Discussion related to the Garmin GPSMAP 66sr GPSr
Post Reply
ewald.gruen
Posts: 61
Joined: Sat Jun 18, 2022 2:58 pm

GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by ewald.gruen »

The 66sr seems to have exactly the same problem as the 66s/st. To find out what was going on, I walked up and down the stairs (32 steps, each 16.5 cm high) 10 times with a 64s (firmware version 6.60) and a 66sr (firmware version 3.90) and then analysed the two tracks. The 64s showed 33 m, the 66sr showed 20 m. Manual calculation from the GPX output without any smoothing gave 53 m for the 64s and 27 m for the 66sr, and this is what the time-synchronized and slightly smoothed (simple moving average with a window size of 3 s) elevation plot looks like:

altimeter_analysis.png
66sr.png
User avatar
GPSrChive
Site Admin
Posts: 2798
Joined: Fri Jan 20, 2017 11:29 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by GPSrChive »

You did not include any information about how the altimeters for each device were configured, and if/when/how they were calibrated before the test.

Also, 3.90 is not the most current firmware for the 66sr.

If anyone were to report your test results to Garmin software engineers, they would want all of this information and request the 66sr be updated before testing again.

They would also request the unedited GPX files directly from the devices for analysis.
asprin624
Posts: 123
Joined: Sun Dec 29, 2019 11:34 am

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by asprin624 »

From the Garmin website
Barometric Altimeter Accuracy of Outdoor Products
With proper calibration, the barometric altimeter of an outdoor watch or handheld will report elevation readings ranging from -2,000 to 30,000 feet with an accuracy of +/-50 feet.1 Elevation values greater than 30,000 feet can be generated, but may not be accurate due to environmental factors.
NOTE: Garmin outdoor products are designed as recreational GPS devices to be used as aids in navigation. They should not be used for any activity requiring precise measurements, such as surveying or weather prediction.
Factors Affecting Barometric Altimeter Accuracy
.Changes in Weather or Air Pressure
.Changes in Ambient Air Temperature
.Calibration Method


Complete info can be found here:

https://support.garmin.com/en-US/?faq=W ... xGtwB7hLY5
ewald.gruen
Posts: 61
Joined: Sat Jun 18, 2022 2:58 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by ewald.gruen »

GPSrChive wrote: Mon Jun 27, 2022 1:42 am You did not include any information about how the altimeters for each device were configured, and if/when/how they were calibrated before the test.
Because it doesn't matter in this case (unless there's a major flaw in the 66sr's altimeter calibration logic, which I doubt). But I obviously set them up the same way (see attachments).
GPSrChive wrote: Mon Jun 27, 2022 1:42 am Also, 3.90 is not the most current firmware for the 66sr.
If you tell me how I can upgrade to 5.50 and later downgrade back to 3.90 without connecting the device directly or indirectly to the internet I might repeat the test with 5.50. But I would like to note that version 3.90 is two years younger than the first post in this thread and consequently there would have been enough time to fix the problem if they wanted to and if it was possible without a change in hardware design.
GPSrChive wrote: Mon Jun 27, 2022 1:42 amThey would also request the unedited GPX files directly from the devices for analysis.
I'm not going to share them and apart from the fact that I think it's highly unrealistic to believe that Garmin's engineers didn't already know this before the product launch, it literally takes 5 minutes to repeat this test yourself.
With proper calibration, the barometric altimeter of an outdoor watch or handheld will report elevation readings ranging from -2,000 to 30,000 feet with an accuracy of +/-50 feet.
I think I speak of all of us (the ones who complain about the massive step backwards of the response speed of the inbuilt barometer) when I say that we do not really care much about the accuracy of the absolute elevation. We don't expect miracles, but simply that the device works as well or better than its predecessor - but maybe I just have to realize that the 66sr is not a full-fledged replacement for the 64s after all and that - among other things - that's exactly why the 65s exists.
.Changes in Weather or Air Pressure
.Changes in Ambient Air Temperature
.Calibration Method
Neither of these explain the slower response time of the barometric sensors in the 66xx devices and besides that all of them were exactly the same during the (less than 4 minute long) test.

194.png
202.png
User avatar
GPSrChive
Site Admin
Posts: 2798
Joined: Fri Jan 20, 2017 11:29 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by GPSrChive »

ewald.gruen wrote: Mon Jun 27, 2022 7:53 am If you tell me how I can upgrade to 5.50 and later downgrade back to 3.90 without connecting the device directly or indirectly to the internet I might repeat the test with 5.50.
That information is available at GPSrChive > GPSMAP 66sr > Firmware

ewald.gruen wrote: Mon Jun 27, 2022 7:53 am I'm not going to share them and apart from the fact that I think it's highly unrealistic to believe that Garmin's engineers didn't already know this before the product launch, it literally takes 5 minutes to repeat this test yourself.
It actually took the better part of an hour to repeat the test with our 64s, 65s, and 66sr, including time to analyze and document the results.

64s vs 65s vs 66sr - 220627.png

ewald.gruen wrote: Mon Jun 27, 2022 7:53 am maybe I just have to realize that the 66sr is not a full-fledged replacement for the 64s after all and that - among other things - that's exactly why the 65s exists.
The 66 series was never intended to be a 'replacement' for the 64 series. If it were, they would not still be producing the 64 series. And no, the 65 series is also not a 64 replacement, it is a 64 with enhanced GNSS capabilities.

The 64 series, 65 series, and 66 series are all still in production, and none have been 'replaced'.
ewald.gruen
Posts: 61
Joined: Sat Jun 18, 2022 2:58 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by ewald.gruen »

GPSrChive wrote: Mon Jun 27, 2022 11:50 pm That information is available at GPSrChive > GPSMAP 66sr > Firmware
Thank you! I didn't realize until now that there are also direct download links there. I will update my 66sr and repeat the test.
GPSrChive wrote: Mon Jun 27, 2022 11:50 pm It actually took the better part of an hour to repeat the test with our 64s, 65s, and 66sr, including time to analyze and document the results.
And what was the true total ascent?
ewald.gruen
Posts: 61
Joined: Sat Jun 18, 2022 2:58 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by ewald.gruen »

I have updated my 66sr to firmware version 5.50 and the response time of the barometer is at least as bad as with firmware version 3.80:

altimeter_analysis.png
66sr_about.png
66sr_altimeter.png
64s_altimeter.png
ewald.gruen
Posts: 61
Joined: Sat Jun 18, 2022 2:58 pm

Re: GPSMAP 66sr: Total ascent way too low (by 20% - 30%) [Bug 64]

Post by ewald.gruen »

For reference, I repeated the test once again with a 64s (firmware version 6.60) and a 65s (firmware version 4.50):

altimeter_analysis.png
total_ascent.png
64s_altimeter.png
65s_altimeter.png
Post Reply

Return to “GPSMAP 66sr”