Results 1 to 20 of 20
  1. #1
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273

    Garmin 510 consistently reading low mileage.

    I have been riding with my 510 for the last few weeks and love it. Prior to procuring it I was using the Map my ride app on my phone. I have noticed that during my rides the Garmin was consistently lower in mileage than my phone. Inter unit reliability is very consistent with the Garmin, more so than my phone. I just figured that the Garmin was more accurate than my phone. Over the last few weeks I have hit four races and while the phone was relatively accurate concerning the course mileage the Garmin was always short. For instance I had a race this past weekend that was 13.2 miles. The phone measured it at 13.5 and the Garmin measured it at 11.9. It was not a heavily wooded area and there were no gaps in the map. I have GLONASS enabled. Any ideas?
    I'm not a vegetarian because I love animals, I just hate vegetables.

  2. #2
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    what is your sampling interval? do you use a wheel sensor?

    An accurate GPS will ALWAYS read a little bit low. The more frequent the sampling interval, the closer it will be, but it will ALWAYS be low. an inaccurate GPS will throw points in places you haven't been and may or may not inflate distances. a wheel sensor for your Edge 510 will help with the distance display on the unit but the location points will not be changed. any program or website that calculates distance based on the GPS location points will show the shorter distance.

  3. #3
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    Thanks for your response Nate, it makes sense.

    It is set to auto sample and I am not using a wheel sensor. What should I change the sample rate to be more accurate.
    I'm not a vegetarian because I love animals, I just hate vegetables.

  4. #4
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    Quote Originally Posted by Orthoguy View Post
    Thanks for your response Nate, it makes sense.

    It is set to auto sample and I am not using a wheel sensor. What should I change the sample rate to be more accurate.
    auto is probably fine for relatively straight trails and definitely fine for road riding. mtb on twisty trails benefits greatly from 1 second sampling. an all day ride at 1sec sampling can have tens of thousands of points but that captures the curves in the trail much better.

  5. #5
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    I'll try it out this weekend. Thanks for your help.
    I'm not a vegetarian because I love animals, I just hate vegetables.

  6. #6
    mtbr member
    Reputation: EBrider's Avatar
    Join Date
    Aug 2004
    Posts
    3,356
    FWIW, the distance on my Edge 500 and 510 are consistent.
    Are we putting air in the tires today?

  7. #7
    mtbr member
    Reputation:
    Join Date
    Sep 2012
    Posts
    31
    I'm glad you posted this question because I am having the same issue with my Edge 810. I went to my typical riding spot on Tuesday, did two of the three trails but took a shortcut. The distance should have been around 9 miles but my Garmin registered 6.6 miles.

    Does the Garmin still use the auto or 1 second sampling when the GSC10 cadence/speed sensor is installed? If not, is there another reason that my mileage was that far off? It has been accurate in the past so this has me a bit stumped.
    2009 Specialized Stumpjumper FSR Elite

  8. #8
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620

    Garmin 510 consistently reading low mileage.

    Smart recording is the default. You have to change it yourself

  9. #9
    mtbr member
    Reputation:
    Join Date
    Sep 2012
    Posts
    31
    I've made that change and will test it today on my ride at the same place.

    I will report back later on.
    2009 Specialized Stumpjumper FSR Elite

  10. #10
    mtbr member
    Reputation: pacman's Avatar
    Join Date
    Jan 2004
    Posts
    3,928
    Quote Originally Posted by Orthoguy View Post
    Over the last few weeks I have hit four races and while the phone was relatively accurate concerning the course mileage the Garmin was always short. For instance I had a race this past weekend that was 13.2 miles. The phone measured it at 13.5 and the Garmin measured it at 11.9.
    I'm not convinced the Garmin is off. Who guarantees the accuracy of the race course distance? The course may have been measured with a sub-standard GPS that exaggerates the distance. Every wiggle off the true line increases the so-called distance. Last year the ascent for the Sea Otter XC course was over-stated. It was obvious from the altitude profile that they had used a non-barometric GPS for the numbers. Lesson - don't rely on official numbers.

    When you say the "Garmin" is that the value in the GPS history or is that the summary from software like Garmin Connect?

    BTW My 510 and 800 tend to agree.

    Garmin 510 consistently reading low mileage.-summary.jpg

  11. #11
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    True. I was referring to the unit history. As an interesting aside I often see variability between the unit and Garmin connect and it doesn't appear to be simple rounding. I figured that it would simply upload acquired data. I guess there is more going on behind the scenes. I did a ride at Tsali yesterday after changing the sampling to 1 Sec and it was moderately closer to the published distances.
    I'm not a vegetarian because I love animals, I just hate vegetables.

  12. #12
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620

    Garmin 510 consistently reading low mileage.

    There is behind the scenes stuff occurring. Unless you have a wheel sensor, your gps itself will not adjust distances for the slope of the trail. This is another reason that an accurate gps will underestimate distances. The corrections that are applied after the fact use elevation data that is imperfect and might only moderately improve it. Many programs and websites will make those corrections. Some do not.

  13. #13
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    Again, that makes sense. It's not a huge deal that my mileage is hyper accurate as I only log it to get a history of my mileage for the year. I just wanted to understand why it was not and to address it if I could. I will look into a wheel sensor. Thanks again!
    I'm not a vegetarian because I love animals, I just hate vegetables.

  14. #14
    mtbr member
    Reputation: Noclutch's Avatar
    Join Date
    Jun 2010
    Posts
    792
    This weekend my 510(one sec sampling and paired gsc 10) recorded 14.3 miles, while simultaneously my Samsung S3's gps and Strava said 13.1, and what was auto loaded from Garmin connect to Strava via 3rd party site was 13.1 (also ), while riding partners Droid said 11.5 miles.... Now how's THAT for a spread!!
    Many of these TBD miles were short ups and downs on single track quarry trails. Did the wheel sensor make the difference?? That would makes sense, but no one seems to know if/how for sure a gsc 10 can be used for distance logging, at least with the 510.

  15. #15
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    with the 500, the GSC 10 is used for speed/distance on the unit at least when it's available. like I have mentioned before, that varies once you put your GPS file on your computer or online. that method seems to work well, so I don't see why the 510 would be any different.

    frequent ups/downs will affect the spread more than a long, consistent climb/descent especially when comparing elevation corrections many sites or programs use that are based on an elevation dataset with a 30m pixel resolution. A lot can happen in a 900m^2 area and the wheel sensor will pick it up where other methods won't.

  16. #16
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    I did a ride this weekend at tsali. The phone had the mileage at 20.1 and the 510 had it at 17.5 but the interesting thing to me is that the phone had the elevation gain at 670 feet whereas the 510 had the elevation at 2100 feet. Is this what you were talking about Nate as far as linear distance computations in regards to elevation?
    I'm not a vegetarian because I love animals, I just hate vegetables.

  17. #17
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    Quote Originally Posted by Orthoguy View Post
    I did a ride this weekend at tsali. The phone had the mileage at 20.1 and the 510 had it at 17.5 but the interesting thing to me is that the phone had the elevation gain at 670 feet whereas the 510 had the elevation at 2100 feet. Is this what you were talking about Nate as far as linear distance computations in regards to elevation?
    not exactly. when accounting for slope of the trail with distance calculations, a GPS with a wheel sensor is simply reporting distance from the wheel sensor. your phone has no way of accounting for slope, so it is giving you simply a linear distance between the GPS points. Any GPS without a wheel sensor will report distance this way.

    the elevation readouts you get on different devices are not used to adjust linear distances. if they were, it might improve speed/distance accuracy somewhat, but they are not used that way. the differences you see in the elevation readout are due to the way the elevation is calculated. your phone is probably using GPS-based elevation calculation, which has an accuracy that is worse than the horizontal accuracy of the GPS location. There's a way to calculate it more specifically, but suffice to say that the accuracy varies depending on your GPS signal lock. the accuracy of the barometric altimeter is much better, does not vary as much, and is much more precise (usually just a few feet per reading). so every foot or two gets recorded with a barometric altimeter, hence the larger number.

  18. #18
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    take the following simplified hill as an example. The hill is the brown slope, the red dots are the track points you GPS lays down. The black steps are a simplified example of a digital elevation model. Each step is one pixel of the DEM that is assigned the same elevation. As you see, some GPS points fall within one pixel even though the elevations were different. This example shows all of the elevations as overestimating the true elevation where the reality is more complex. Some overestimate, some underestimate, and some might be spot on. It all just depends. But the concept of it messing with GPS track elevation totals is the same basic idea.

    If you were to project all of the points to the bottom of the image and measure the horizontal distance between them, that is how your GPS calculates the distance you traveled. Adding a wheel sensor that reports speed and distance on the head unit live will account for the extra distance traveled on slopes.
    Attached Thumbnails Attached Thumbnails Garmin 510 consistently reading low mileage.-gps_hill.png  


  19. #19
    mtbr member
    Reputation: Orthoguy's Avatar
    Join Date
    Dec 2011
    Posts
    273
    I see so error will increase with increasing grade.
    I'm not a vegetarian because I love animals, I just hate vegetables.

  20. #20
    Rep Power: Infinity
    Reputation: NateHawk's Avatar
    Join Date
    Dec 2003
    Posts
    12,620
    Quote Originally Posted by Orthoguy View Post
    I see so error will increase with increasing grade.
    it's more like error increases when grades are short and reverse quickly. If they are long and consistent regardless of steepness, there will be less error. Short and steep will increase error, but generally speaking in MOST cases, the over-estimates and under-estimates will start to balance each other out so you cannot say with certainty that a number will be definitely an overestimate or definitely an underestimate. Keeping in mind that most elevation models assign the same elevation to each 30m x 30m pixel, anything that happens within that pixel (up or down) won't be captured by elevation corrections applied after the fact.

Similar Threads

  1. Garmin 500 & Garmin External Power Pack test
    By xxxxx in forum GPS, HRM and Bike Computer
    Replies: 6
    Last Post: 04-02-2013, 07:17 PM
  2. Garmin Edge 800 mileage tracker
    By ZTallboy in forum GPS, HRM and Bike Computer
    Replies: 1
    Last Post: 05-31-2012, 08:39 PM
  3. Powertap calibration reading on Garmin Edge
    By yampydog in forum GPS, HRM and Bike Computer
    Replies: 0
    Last Post: 12-04-2011, 02:36 PM
  4. Contact Seller in classified not working consistently
    By Swami Scott in forum Site Feedback/Issues
    Replies: 3
    Last Post: 08-24-2011, 04:00 PM
  5. Not Shifting Consistently
    By Adampff in forum Drivetrain - shifters, derailleurs, cranks
    Replies: 5
    Last Post: 06-04-2011, 04:04 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •