Related
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Large distance discrepancies

allanhvass
over 4 years ago

Hi,

I'm seeing some total distance data with the Edge 520+, that are somewhat difficult to believe. For a recent route, Edge 520+ reports 40,73km whereas importing the .fit file into AllTrails caculate same route as 35,83km, see below screen shot. I believe my old Edge 705 previously measured the distance to be about 38,7km. I do have the speed sensor counting wheel revolution installed. It is a very twisty/turny route with many small turns and ups/downs, but smart recording of data has 3135 points, or about one for each 12m or so, so the recording also has a significant level of detail, so I don't think the difference is due to data point based calculation by AllTrails missing all the cornering.

I've had wheel size set to 2244mm on this trip after a manual measurement with load using 10 revolutions, so it should be quite accurate. The Edge 520+ itself had also reported it using auto calculation to be 2256mm, whereas the 705 had reported it to be 2238mm, so it should be pretty close to correct. Wheel size is 29x2.2in, and that is by Garmins table 2290mm, but that doesn't account for compression under load, and in any case setting it as high as Garmins reference table would only have increased the distance even more.

What could be the reason for such huge difference between Edge 520+ which from my understanding uses wheel revolution as primary measurement for distance / speed versus AllTrails that I believe calculates based on GPS data (I think)?

Additional data point is, that importing this ride into Garmins BaseCamp, while it keeps the distance of 40,71km as a property of the ride, the graphs for the ride shows it as ending at 35,75km, so it seems like data doesn't correspond at all with the totals. How is this even possible?

  • 0 over 4 years ago

    Hi,

    Possible explanations are:

    • GPS calculated distance assumes a straight line between recorded points, so it will be shorter than the actual route if it has a lot of turns close together. Using 1 sec recording rather than smart would improve accuracy.
    • Any GPS dropouts during the ride will give straight lines between the missing points.

     You could check by recording a ride on a reasonably straight route so see if the discrepancies still occur.

  • +1 over 4 years ago in reply to PeterK_55

    After much testing doing about 10 test drives with both an Edge 705 and an Edge 520 Plus and multiple configurations of GPS (pure, Glonass, Galileo) as well as detailed studying of the actual tracks recorded, the conclusions are:

    1. When speed sensor is configured correctly with true precisely measured wheel size under load, distance stated directly on GPS unit is correct. Auto calculation of wheel size is in most cases a good approximation, but can still be off by 1-3%, being a factor in above results.
    2. On open paved roads with not too many turns, Edge 705 and Edge 520 Plus are reasonable close, both with each other as measured on both speed sensor based distance and GPS based calculation afterwards in e.g. Garmin BaseCamp. However there is typically a small consistent difference still with Edge 520 Plus GPS calculated distance being 1% shorter than other measurements, due to GPS tracking having a tendency to cut corners more aggressively in it's tracking.
    3. When doing the same testing as in #2 on MTB tracks in dense forest areas with high amount of turning and twisting and many serpentine tracks, that same shortening of distance on Edge 520 Plus GPS based tracking grows to 7-10% as compared to true speed sensor based distance, and to about 4-7% shorter than Edge 705 GPS based tracking. Studying the maps very closely, the shortening of tight turns is much more aggressive and even generally accuracy somewhat lower, resulting in worse recorded GPS tracks for a modern Edge 520Plus versus a 12 year old Edge 705... This has been validated compared against satellite photos of the tracks. For short sections with many short 180 degree turns, the difference may be as big as 20% lower distance than true distance for Edge 520 Plus.
    4. Interestingly, the results are worst in my geo (Denmark) when using GPs+Glonass (around 10% off true distance), with GPS only being about 9% off, and GPS+Galileo being about 8% off. This compared with the Edge 705 that has GPS only, and which is seldom more than 3% off in total distance.

    Conclusions: On open roads, there's no major issues, expect 1-2% mis-measurement using GPS only with Edge 520 Plus, adding and precisely configuring a speed sensor should make results quite accurate. However for MTB use in dense forest areas, expect significant shorter distance measurement and generally quite inaccurate GPS path tracking (7-10% off) with Edge 520 Plus, distance cannot be trusted at all and supplement with a speed sensor. Test whether GPS+Galileo gives more accurate results than GPS+Glonass in your geo. Consider whether other Garmin models (newer or older) are better at tracking the path under difficult conditions.