open topic, for anything cycling related.
17 posts • Page 1 of 1
Im having troubles with my elevation data on my 800.
It was working great for ages then all of a sudden it stopped reading properly altogether. At the start of my ride it will go from 0m and shoot straight up to 600m straight away, then slowly come down over the course of my ride.
I have set the elevation of where my house is, coincidentally it is about 600m.
Has anyone else had this problem or know how to solve it?
Heres a pic of what it looks like on strava.
Uploaded with ImageShack.us
http://www.daftlogic.com/sandbox-google ... titude.htm
Or alternatively there are plenty of GPS apps for smart phones that will give your altitude
You can resolve it in Strava by clicking on the Elevation (?) bit and it will re-calculate you elevation data based on maps. With the Garmin, I have found you need to get a good number of GPS locks before you start. What sort of mount are you using? I've found my aluminium mount affects to GPS signal and I need to hold it in my hand to get a better lock.
My only elevation problems have been when water droplets have covered that hole riding in the wet.
Have you set an elevation point on the garmin? Let it get a good lock while you are outside your garage door or wherever you normally start your ride then set it as an elevation point with your known elevation. As long as you start your ride within 35 meters of that spot, it will use the elevation point info and use that. Mine lets me know once I start moving and hit start that its using an elevation point, by popping up a message on the screen saying "elevation point found" or something like that.
Well thats the way it works with the 500 and I assume the 800 will be very similar, they both have barometric altimeters.
Every one of the garmin Edge computers (705, 800, 500) uses a barometric altimeter which is prone to massive inaccuracy (+/- 100m from the truth), as well as the one-sided decline in readings that you've now found. What is surprising is that
it seemed to be accurate for a while - beginners luck maybe.
Their notorious inaccuracy is why most logging tools and web-sites have an elevation correction tool that uses an estimated elevation from a web lookup. That's a pretty rough estimate only - based on sampling satellite imaging at a 100m grid, so it often misses the truth where you ride through a cutting or on a filled section of road. The elevation correction tools can also get it wrong where there are cliff-faces or steep mountains near to your route.
If you want accurate elevation data, get another GPS logger that doesn't have a barometric altimeter. Most phones
will do the job well (except for the battery drain trouble !).
WombatK - Jerry Garcia, Grateful Dead
I've used a 705 (for over 4 years now), and more recently the 800, and with both have found that only on rare occasions (perhaps a handful of times a year) does it give elevation readings that are incorrect by +/- 100m. For 99% of the time, the elevation readings are fairly good. I've done rides anywhere from around Port Philip Bay to mountains in the Alps, and have been very happy with the elevation readings from my 2 Garmins.
If your doing a long ride as well there will be changes in the atmospheric barometric pressure as well which has an impact on reading from barometric pressure. Most time you will get a good reading
But I do know on my edge 705 I get different elevation to a rider using an edge 800, someone else on same ride has edge 800 and their values are close but out compared to mine. Most of the time out by 100m elevation gained etc
Masi Speciale CX 2008 - Brooks B17 special saddle, Garmin Edge 810
For the while I've had my 500, its been pretty accurate when I use an elevation point at the start of a ride. It starts a ride knowing its exact altitude at the start and any drift after that is down to changes in the weather which aren't all that huge for my length rides. 10km after leaving home and climbing Mt Cootha its showing pretty much the right altitude at the top so its working fairly well for me
It doesn't have to be a long ride, and my experience with both a 705 and a 500 is that you rarely get good performance if you compared it with something that was accurate.
Here is a typical example from my 705 where I rode laps around the Dunc Gray Velodrome crit track (and to & from)...
The real elevation at the start and end is 53m above sea level - and you can see a steady gain in elevation every
lap (800 m) - in this case about 1 m per lap.
The elevation change is always one-way - but sometimes it will fall, others it will rise - and it bears
no relationship to the change in atmospheric pressure, as measured at the very close Bankstown Airport weather
5 km laps of Sydney Olympic Park produce a drift of approximately 5m per lap.
In the case of the 500, setting the correct elevation at the start is soon enough undone by the usual drift - I've probably never been for a ride where the finishing elevation of my home was within 25m of where it started, and had quite a few where it's more than 100m out.
WombatK - Jerry Garcia, Grateful Dead
It’s all the heavy breathing you guys are doing just above the poor thing..........no wonder it gets confused about elevation with all that hot air being blown into the holes on the underside.
My understanding of the barometric unit in Garmin GPS' is that it is used to 'smooth' the height information.
The GPS only elevation information is significantly less accurate than horizontal position. However if you analyse this data over a long time (hours) you get pretty accurate elevation information. However this isn't too useful if your actually going up a hill.
If you assume that relative elevation changes calculated from the barometer are pretty good in the short term (pressure changes normally aren't that fast),but dodgy in the long term, you can use the short term poor, long term better GPS elevation to correct the barometric information with some nifty statistical analysis. Of course this isn't perfect, a low coming though will effect the elevation - its a trade off from almost useless raw GPS elevation data to elevation data with a shift in elevation.
The other issue is that if you start from a point where the GPS thinks you an elevation of 80m when you turn the GPS on but your really at 10m , the algorithim takes some time tweak the elevation readings using the GPS elevation info. I'd says that's pretty much what is seen in wombatK's track.
The best idea is to set a known elevation point. If you're in Victoria you can rummage around:
http://services.land.vic.gov.au/landcha ... ressSearch
and find a local survey mark with a value for AHD (Australian Height Datum). Of course that isn't the height you want. You'll
need to convert to WGS84 for GPS elevation. GDA94 is nearly the same as WGS84 (It was the same in 1994 but the Australian
tectonic plate has moved a bit since then!) You can convert from AHD to GDA94 elevation here :
At my location ( 138S 145E ) the local survey marker at 18M AHD is 22M GDA94 - a 4 M difference.
1.370" x 24 tpi - what sort of stupid standard is that?
17 posts • Page 1 of 1
Who is online