Garmin accuracy.

[peetee]

Forgive me if this is in the booklet somewhere but I am a ‘typical bloke’ who tends to do the hands-on thing rather than read instructions.
Anyhow, I have a Garmin Edge Tour which has the great feature of recording climbing & descending en route. However, I did a circuit the other day, all on road, front door to front door and it was way out; recording 640m climbing and 560 descending.
Has it just lost satellite reception somewhere? If so it must be a considerable chunk of road and somewhere I can probably locate as it was for the most part an out-and-back ride. It should also show on the route map shouldn’t it? How does it even record elevation?

[Psamathe]

My Garmin 1030 has a barometric altitude sensor. https://support.garmin.com/en-GB/?faq=sFMkjQFdnZ99DcunfAue66 lists Edge devices that use such a sensor.

GPS altitude is always pretty rubbish (nothing to do with Garmin).

I find the same on my Edge 1030 - circular route home back to home and your ascent and descent are moderately different where on my iPhone track recording app ascent and descent are the same (the iPhone app is set to get altitude from map data).

Not only is there the issue of ascent and descent being different for a circular route but I have a very strong/I'm convinced that the ascent displayed massively over calculates your ascent. e.g. level road, up a 3m hill then levels again and the "ascent" displayed increases by over double the actual height of the small hill the cumulative ascent might increase by 6+m. That said I tend to regard ascent/descent figures from any system only useful for comparison with the same system as so much depends of the algorithms/methods used to calculate ascent and altitudes (even from mapping data). So yesterday's ride Garmin was 279m ascent and todays was 321m so today's was more climbing than yesterday's. Only compare Garmin ride altitude figures with same Garmin figures for other rides, only compare cycle.travel figures for one route with other cycle.travel figures for other routes, only compare MapOut figures with other ride MapOut figures, etc.

Ian

[Nigel]

My experience of Garmin's barometric measure, on older hand-held Etrex device, is that wind conditions directly near the sensor hole matter a great deal. With the GPS in the top of a rucksac, the wind effects are negligible, and readings decent. With GPS outside in a shoulder pouch on windy days, the readings could be awful.

I suspect this is the wind leading the fluctuations in pressure, which the device interprets as height changes. Add that to variations in pressure over the day as the weather changes (and pressure changes in the UK happen frequently), and the readings can be variable.

[andy74]

If you use Garmin connect to view your data you can set it to correct the elevation. I believe that it does this from map data.

I'm a trendy consumer. Just look at my SM-G973F using hovercraft full of eels.

[RickH]

If you use Garmin connect to view your data you can set it to correct the elevation. I believe that it does this from map data.

I'm a trendy consumer. Just look at my SM-G973F using hovercraft full of eels.

If your device has a barometric altimeter Garmin Connect will automatically use that data. But you can opt to correct to map data (I can't remember if you can set that as the default). If you don't have a barometric altimeter then Connect will automatically use map data & I don't think you have the option to use the device's GPS altitude.

[andy74]

If you use Garmin connect to view your data you can set it to correct the elevation. I believe that it does this from map data.

I'm a trendy consumer. Just look at my SM-G973F using hovercraft full of eels.

If your device has a barometric altimeter Garmin Connect will automatically use that data. But you can opt to correct to map data (I can't remember if you can set that as the default). If you don't have a barometric altimeter then Connect will automatically use map data & I don't think you have the option to use the device's GPS altitude.

I think you may be right. An interesting article about the subject here. https://www.dcrainmaker.com/2010/05/und ... e-gps.html

I'm a trendy consumer. Just look at my SM-G973F using hovercraft full of eels.

[Mick F]

Garmin accuracy?
It depends on the GPS reception and the sensitivity of the device.

Here's a track from my Garmin Edge 20 (cheap), followed by the track from my Garmin Montana (expensive).
Riding left to right on the road past Hingston Down quarry. Edge 20 on Mercian, and Montana on Moulton.

[pwa]

I don't think anyone should view Garmin elevation data as anything more than a bit of entertainment. Even with the barometric models. I have known cyclists do a ride together and come back with different figures.

[peetee]

Thanks for the comments. I don’t know how the Edge Tour sources it’s data but the barometric system seems like it’s designed to be inaccurate so perhaps that’s the answer as on that particular day the weather did take a noticeable dip as I got on the coast road.

[philvantwo]

Mick F........was the top photo after a session in the Wetherspoons!
Are you not on your third Montana?

[simonineaston]

Aviation fora used to be full of posts warning of potential dire consequences, resulting from reliance on GPS altitude readings, presumably prompting the manufacture of consumer products to dabble with inexpensive pressure sensors. However, as far as I know, ( a chums's brother works for ESA, on their Galileo project, which of course we have recently voted ourselves out of...) the potential is there for high levels of accuracy from the system - thousands of commercial airliners flying safely being good evidence, not to mention pin-point accuracy from smart munitions... We are all dependant on what the American government sees fit to give us access to and of course they keep the best for themselves. Foreign end-users, like us, get the crumbs off the table! The linked article (albeit from 2014) includes a chart quoting respective vertical accuracy, if you wish to peruse the sort of thing we'll be missing out on!
https://www.gsa.europa.eu/newsroom/news ... d-services

[Psamathe]

Aviation fora used to be full of posts warning of potential dire consequences, resulting from reliance on GPS altitude readings, presumably prompting the manufacture of consumer products to dabble with inexpensive pressure sensors. However, as far as I know, ( a chums's brother works for ESA, on their Galileo project, which of course we have recently voted ourselves out of...) the potential is there for high levels of accuracy from the system - thousands of commercial airliners flying safely being good evidence, not to mention pin-point accuracy from smart munitions... We are all dependant on what the American government sees fit to give us access to and of course they keep the best for themselves. Foreign end-users, like us, get the crumbs off the table! The linked article (albeit from 2014) includes a chart quoting respective vertical accuracy, if you wish to peruse the sort of thing we'll be missing out on!
https://www.gsa.europa.eu/newsroom/news ... d-services

This site says general signal is accuracy <1m (encrypted that we (domestic free users) wont have access to <20cm https://www.deingenieur.nl/artikel/after-13-years-galileo-satellite-navigation-complete-at-last.

My Garmin 1030 now supports GPS & Galileo (&Glonass) but the accuracy (displayed by the device) is always 9ft which is GPS NOT Galileo but my 1030 is set to "GPS + Galileo" and I thought Galileo was meant to be mostly deployed by now.

Ian

[Mick F]

Mick F........was the top photo after a session in the Wetherspoons!
Are you not on your third Montana?

Wrong on the first point, and correct on the second.

First Montana was absolutely fine except the power button had a crack in the outer case.
I asked Garmin for a new case, but was offered a refurbished unit complete, at a good price.
It worked for a while and then was crap, so they sent me another ............. which is as good as the original one ................. correct me if I'm wrong and my memory has had it .............. my original Montana was 2012.


-----------------------------------

Back to the question in hand.
The elevation system needs to be calibrated.
My previous unit to the Montana was an Edge 705.

By knowing the exact elevation, you can either set a waypoint at the accurate elevation at your front gate like with the 705, or with my Montana, you can calibrate the altimeter when you boot it up.
My Edge 20 has no BP altimeter and relies solely on the GPS elevation ............. poor accuracy.

With a decent device that you can calibrate if you know the accurate elevation, the BP and the GPS continually monitor each other. Belt and braces.

When I ride with my Montana and calibrate the elevation before I leave the front gate, it reads the same when I get back.
If I don't calibrate it and leave it still and non-moved ............. and the weather has changed .............it'll take 15mins before it finds the correct elevation.

[philvantwo]

What's your altitude at the start then Mick F?

[Mick F]

If I told you that, all you would need to do is look at the contour lines around Gunnislake, and you could probably guess exactly where we live!

What I did, was leave my Montana running outside at handlebar height for a few hours.
The system notes the GPS elevation and the ambient air pressure.
As the device isn't moving, the GPS elevation gets more and more accurate over time.
The barometric pressure also gets more accurate.
Looking at the OS map to confirm the figure is a belt and braces approach and it's a nice confirmation to see it.

Eventually, the BP is correct and so is the GPS elevation, so the BP corrected elevation is correct too.
The data fields I have displayed show BP elevation.

Before setting off, I go to the setup page and go to Altitude/Calibrate. I input the correct BP elevation and this short-circuits the long drawn out method described above .............. but you must know the correct elevation to do this. Get it wrong, and it'll be trying to get it correct for you.

When I had my Edge 705, you have to do similarly in the long method, as you can't calibrate them. What you can do, is find somewhere right near the start of your ride, go there, let it settle down and create a waypoint. Next time you cycle away, you go through that waypoint, and the BP elevation flips instantaneously to what the waypoint is.