Results 1 to 41 of 41
  1. #1
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28

    Garmin 520 distance issue

    I've just got back into riding (intro post to come if I can find where?) and got myself a Garmin Edge 520 to replace a Garmin Forerunner 405 I've had for years.

    The issue I'm having is the 520 consistently records my trail rides as 9% shorter than the Forerunner and the people in my group using strava and Polar on iphones. The iphone recorded distances match the info from my Forerunner almost exactly but the 520 always shows less (I do not have a wheel speed sensor). All the settings are correct, 1 second record, GPS/GLONASS, auto pass off etc.

    I got on to Garmin support who after making sure the unit was up to date and that the setting were as above told me the iphones are all wrong, the 520 is correct and the Forerunner is old so incorrect.

    Has anyone had a similar issue? Is it possible Garmin guy is right and EVERYONE'S phones and my Forerunner are wrong and my 520 is right?

  2. #2
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    If you have a basic bike computer, the type that used a wheel magnet, calibrate it to the actual wheel and tire circumfrance, throw it on and do a ride, comparing the results.

    It’s possible the 520 is correct, especially if assorted cell phones are not picking up GLONASS. The 520 might have a more sensitive and better GPS calculation than older devices.

    I found that even my Garmin 810 was accurate on road rides to about 1/2 mile in 100 as compared to a calibrated Cateye. It was short by about 2 miles in 100 when mt. biking. Thus I started using wheel sensors,

  3. #3
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    To add a brand new Wahoo Elemnt Bolt also matches the other devices almost exactly.

  4. #4
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Curious what the maps from recorded rides look like between devices.

    If you compare, possibly you will see if the 520 is cutting corners, though it obviously might be. Maybe bad signal in woods ? And maybe you will see that on the map.

    Turn off GLONASS ?

  5. #5
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    The 520 tracks look cleaner and seem to cut less corners than the Garmin.

    I’m going to try turning off GLONASS.

  6. #6
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    An accurate GPS will ALWAYS underreport distances. INaccurate GPSes will display extra positional wander that will "add" the distance back, but the track will inaccurately reflect where you went.

    Add a wheel sensor to your edge and you will get that distance back.

    Turning off glonass will not improve anything with distance recording. All that will do is get you a little extra battery life back, at the expense of some accuracy (how much will depend on where you are) and speed of location fix.

    Sent from my VS995 using Tapatalk

  7. #7
    mtbr member
    Reputation: Mudguard's Avatar
    Join Date
    Apr 2009
    Posts
    977
    Quote Originally Posted by Catmandoo View Post
    Curious what the maps from recorded rides look like between devices.

    If you compare, possibly you will see if the 520 is cutting corners, though it obviously might be. Maybe bad signal in woods ? And maybe you will see that on the map.

    Turn off GLONASS ?
    Why would you turn it off? It's reducing the number of satellites to lock on to. Add a wheel sensor.

  8. #8
    Out spokin'
    Reputation: Sparticus's Avatar
    Join Date
    Dec 1999
    Posts
    9,316
    I’ve had two Garmin 520s and they both underreported mileage. In fact, that’s why I returned the first one. I employed a used 820 for a while, then gave it to a friend and bought another 520, hoping my first on was defective. The second 520 was just as bad.

    So I bought a speed sensor and now it’s up to snuff — all good. Except now it seems to be under reporting elevation gain. My GF and I ride together a lot. She has a Garmin 810 which consistently reports 20%ish more gain than my 520. On yesterday’s ride, her 810 gave her 1400’ gain, my 520 said less than 1000’.

    I don’t know what’s up.
    =sParty
    disciplesofdirt.org

    We don't quit riding because we get old.
    We get old because we quit riding.

  9. #9
    mtbr member
    Reputation: Mudguard's Avatar
    Join Date
    Apr 2009
    Posts
    977
    There is a setting in the Garmin app or your online profile that corrects elevation if you allow it to compare it known elevation (ie a topo map or something). It will then tweak it before it goes to Strava. I think. Its been awhile since I've looked.
    As for comparing two devices, well how do you know which is actually accurate? The only way is with a wheel sensor that's calibrated correctly, two GPS devices side my side will still give varying amounts, it's just the nature of them.

  10. #10
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Elevation readings are notoriously inaccurate. Think about it. There is NO WAY to directly measure your climbing totals on a ride. Your Garmin (or whatever device) is making lots of indirect measurements, using a method that gets confounded by changes in the weather...and that is the BEST method available. Plus, it's hard enough to measure the elevation of a single point.

    Sent from my VS995 using Tapatalk

  11. #11
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by Mudguard View Post
    Why would you turn it off? It's reducing the number of satellites to lock on to. Add a wheel sensor.
    Only as it's 2 completely different positioning systems. GLONASS is the Russian version of the US GPS system. It's not the same signal and as such requires the 520 to calculate separately. Wouldn't surprise me if the Garmin software would calculate location differently using these 2 potentially conflicting positions. In fact I know this happens, the article here explains the issues and how the reception and accuracy is dependent on the type of antenna and of course the built in software used to "merge" the positioning of the 2 different systems.

    https://www.electronicdesign.com/tes...ict-gpsglonass

    In truth I don't think we know how the 520 decides which position of the 2 different signals is accurate and it's not the same as receiving from additional US satellites. Thus my thought was to isolate the problem by turning off one system

    Stab in the dark problem solving essentially.

    I turn off GLONASS on my 1000 to save battery life and use speed sensors when in the woods. We assume that's the kind of riding the OP is doing, thus the easy recommendation is to buy a $39 speed sensor, which having just plunked down $250 for the 520 I could see the OP asking "Why should I have to do that ?"

  12. #12
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Whilst I agree that adding a speed sensor is the smart thing to do and will lead to more accurate readings it still doesn’t explain why the 520 reads differently from multiple devices riding the exact trail. And I don’t buy that the iPhone, Forerunner and Wahoo are all exactly the same amount of wrong.

    The other strange thing is both the Forerunner and 520 read exactly the same when taken for a drive. It has to have something to do with coverage in the forest. I read somewhere someone had a similar issue and deleted the training profile and recreated it and the problem went away, but I can’t find that thread.

    This is my fourth Garmin device, should’ve known it wouldn’t have been easy!

  13. #13
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by James80 View Post
    Whilst I agree that adding a speed sensor is the smart thing to do and will lead to more accurate readings it still doesn’t explain why the 520 reads differently from multiple devices riding the exact trail. And I don’t buy that the iPhone, Forerunner and Wahoo are all exactly the same amount of wrong.

    The other strange thing is both the Forerunner and 520 read exactly the same when taken for a drive. It has to have something to do with coverage in the forest. I read somewhere someone had a similar issue and deleted the training profile and recreated it and the problem went away, but I can’t find that thread.

    This is my fourth Garmin device, should’ve known it wouldn’t have been easy!
    The article I linked shows the issues with receiving in urban area's. Wooded rides are essentially the same thing, especially with a lot of leaf cover or conifer type forests. I don't get the issues locally as the woods here are predominantly deciduous and in winter it's a better view of the sky. I still use a speed sensor but as much as the trails I ride are very twisty and tight with a lot of quick turns. GPS has trouble with that,

    As the 520 is now almost a dated device, I would think about a return for a 520 Plus (if not too late), or an Edge Explore, which is larger but the same price as the 520. Or if you are not using pretty maps, maybe get a Bolt or Elemnt.

  14. #14
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Quote Originally Posted by Catmandoo View Post
    The article I linked shows the issues with receiving in urban area's. Wooded rides are essentially the same thing, especially with a lot of leaf cover or conifer type forests. I don't get the issues locally as the woods here are predominantly deciduous and in winter it's a better view of the sky. I still use a speed sensor but as much as the trails I ride are very twisty and tight with a lot of quick turns. GPS has trouble with that,

    As the 520 is now almost a dated device, I would think about a return for a 520 Plus (if not too late), or an Edge Explore, which is larger but the same price as the 520. Or if you are not using pretty maps, maybe get a Bolt or Elemnt.
    Agreed about coverage but any coverage issues in the woods should affect all the devices to the same extent. The distance between the forerunner and 520 is wrist to stem. So we’re back to either

    1: My 520 is faulty (will try to ride with a friend who has one tomorrow to compare)

    2: My settings are wrong and or unit needs resetting

    3: That’s just how 520s work

    4: 520 is right and every other device in my group is wrong.

  15. #15
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by James80 View Post
    Agreed about coverage but any coverage issues in the woods should affect all the devices to the same extent. The distance between the forerunner and 520 is wrist to stem. So we’re back to either

    1: My 520 is faulty (will try to ride with a friend who has one tomorrow to compare)

    2: My settings are wrong and or unit needs resetting

    3: That’s just how 520s work

    4: 520 is right and every other device in my group is wrong.
    You're not listening or comprehending (can't tell which).

    An accurate GPS is always going to shortcut corners. ALWAYS. This is simply due to the way GPS works (connect the dots with straight lines that don't follow the curve). The more frequent the device records, the shorter this discrepancy (and the less frequent, the greater the discrepancy), but it will always exist. Reductions in positional accuracy will add distance back in other areas, like wandering on straight sections, or drifting around when stopped, etc. Plus, each device has a different combination of chipset, antenna, and processing algorithm that affects the end result.

    https://www.jsams.org/article/S1440-2440(09)00182-0/pdf
    GPS Accuracy of Garmin, Polar, and other Running Watches - Fellrnr.com, Running tips

    Note in the 2nd link (which focuses on running models) that the footpods are almost always more accurate AND more precise than GPS-only distance measurements.

    Get a wheel sensor.

  16. #16
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by Harold View Post
    You're not listening or comprehending (can't tell which).

    An accurate GPS is always going to shortcut corners. ALWAYS. This is simply due to the way GPS works (connect the dots with straight lines that don't follow the curve). The more frequent the device records, the shorter this discrepancy (and the less frequent, the greater the discrepancy), but it will always exist. Reductions in positional accuracy will add distance back in other areas, like wandering on straight sections, or drifting around when stopped, etc. Plus, each device has a different combination of chipset, antenna, and processing algorithm that affects the end result.

    https://www.jsams.org/article/S1440-2440(09)00182-0/pdf
    GPS Accuracy of Garmin, Polar, and other Running Watches - Fellrnr.com, Running tips

    Note in the 2nd link (which focuses on running models) that the footpods are almost always more accurate AND more precise than GPS-only distance measurements.

    Get a wheel sensor.
    Harold, the OP is comprehending and talking about something different than why GPS tracks will be short. He’s not understanding why a number of other dissimilar devices are showing the same (potentially) short track while the 520 is significantly shorter, is all.

    For the OP, do a Google of “my Garmin 520 records short” or some such. Tons of posts. Solution is a speed sensor. Or exchange for a different unit. Or both.

  17. #17
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by Catmandoo View Post
    Harold, the OP is comprehending and talking about something different than why GPS tracks will be short. He’s not understanding why a number of other dissimilar devices are showing the same (potentially) short track while the 520 is significantly shorter, is all.

    For the OP, do a Google of “my Garmin 520 records short” or some such. Tons of posts. Solution is a speed sensor. Or exchange for a different unit. Or both.
    So the 520 is shorter than other devices? so what? my point is it doesn't matter "how much" short it is because that's how GPS works. the second link I shared has a nice chart that shows a bunch of devices and guess what - they're ALL different from each other. You cannot tease out the difference between the specific question of "why is x device reporting shorter distance traveled than y and z devices?" without understanding the fundamental reason why GPS measurement of distance in fitness activities is not all that accurate in the first place.

    The nature of GPS means that you simply cannot expect 2 devices to report exactly the same. Ever.

  18. #18
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Quote Originally Posted by Catmandoo View Post
    Harold, the OP is comprehending and talking about something different than why GPS tracks will be short. He’s not understanding why a number of other dissimilar devices are showing the same (potentially) short track while the 520 is significantly shorter, is all.

    For the OP, do a Google of “my Garmin 520 records short” or some such. Tons of posts. Solution is a speed sensor. Or exchange for a different unit. Or both.

    Exactly. And yes I’ve done multiple searches but didn’t find a satisfactory answer hence asking the question here.


    Quote Originally Posted by Harold View Post
    So the 520 is shorter than other devices? so what? my point is it doesn't matter "how much" short it is because that's how GPS works. the second link I shared has a nice chart that shows a bunch of devices and guess what - they're ALL different from each other. You cannot tease out the difference between the specific question of "why is x device reporting shorter distance traveled than y and z devices?" without understanding the fundamental reason why GPS measurement of distance in fitness activities is not all that accurate in the first place.

    The nature of GPS means that you simply cannot expect 2 devices to report exactly the same. Ever.
    I’m not expecting them to report exactly the same. I’m simply saying that every other device is with a “reasonable” range and the 520 is consistently 9% less. And my question is would another 520 strapped next to mine report the same or is mine especially “out”.

    Be interested to hear your thoughts on these:

    Purple is the 520 and orange the Forerunner







    The tracks are identical on the road and start to differ once in the woods interestingly the elevation tracks are identical except the 520 is 9% lower and lags behind by the mileage difference.

  19. #19
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    For one, both devices are set to the "auto" recording frequency.

    Change your Edge 520 to 1 second recording and report back.

  20. #20
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Edge is set to 1 second.

    Any thoughts on Forerunner being GPS only and 520 GPS and GLONASS?

  21. #21
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by James80 View Post
    Exactly. And yes I’ve done multiple searches but didn’t find a satisfactory answer hence asking the question here.

    I’m not expecting them to report exactly the same. I’m simply saying that every other device is with a “reasonable” range and the 520 is consistently 9% less. And my question is would another 520 strapped next to mine report the same or is mine especially “out”.

    Be interested to hear your thoughts on these:

    Purple is the 520 and orange the Forerunner


    The tracks are identical on the road and start to differ once in the woods interestingly the elevation tracks are identical except the 520 is 9% lower and lags behind by the mileage difference.
    I'd be curious as to how the 2 different tracks actually overlay on the actual trail. Can you find any other GPS tracks from other riders ?. MapMyFitness ?, RWGPS ?, TrailForks ?. Maybe compare to what your 520 unit is doing different - if anything.

    You may go nuts trying to figure all this out and as Harold has stated GPS devices have inaccuracies inherent to how they function. It's anybody's guess what Garmin does on the 520 in terms of using both GLONASS and US GPS mixed to determine position.

    I would turn off GLONASS as a start. Then get a speed sensor (good to have in any event) Then and if it was still showing 10% errors on a regular basis, I'd replace with a 520+ or Edge Explore if you still have the option.

  22. #22
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by James80 View Post
    Edge is set to 1 second.

    Any thoughts on Forerunner being GPS only and 520 GPS and GLONASS?
    Were you hauling epic ass on the road segment, then? I only see a saved point where you turned onto singletrack. That is why I assumed auto recording interval (fewer recorded points on straights). That, and the visible portion doesn't look steep.

    This trail layout is kinda worst case scenario for gps recording, anyway. The twistier the trail, the greater the shortcutting errors. and besides, without a known high accuracy baseline, we simply cannot know which track is most accurate, or if they're equally inaccurate but in different ways.

    You can't even really trust the trail data on most mapping websites as being of high accuracy. Most of them heavily simplify uploaded data to save on server storage and processing. And that's assuming that high accuracy data was uploaded in the first place. Which isn't often the case. Many of these sites would rather have low quality data for more trails than high quality data for fewer trails.

    It can be instructive to show multiple passes on the same trail on different days with the same device, too. You are going to see a spectrum of data quality with good quality runs and poor quality runs.

    It's worthwhile to know your latitude and region, too. GLONASS tends to perform better at higher latitudes, as it was designed for that. It actually works better than gps at extreme northern latitudes. Farther south, not so much, and gps will outperform it. Are there terrain impediments like mountains or steep valleys (especially impediments on the southern horizon)? How thick is the veg growth?

    And for that matter, none of what I see there would bother me. I see far worse with some frequency when I check out strava flybys.

    Sent from my VS995 using Tapatalk

  23. #23
    mtbr member
    Reputation:
    Join Date
    Oct 2018
    Posts
    571
    You should ride a staright, long road type of course. See how the results compare to other devices, or known distances.

    My Garmin 405 watch would always report closely to the my computer (not GPS) and Strava when road riding. But not when on the MTB trails.

    Even when trail running, my watch is not as accurate as it is on open-environment road courses.
    The GPS takes a shot and connects the dots. Depending on the frequence and how much distance you have covered in said distance is how much it will affect the curves.

    On my road bike there is a section that always gets off track for me because it's an s-bend. I've going through the curves quickly enough that GPS can't draw the curve and therefore is inaccurate.

  24. #24
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Latitude 10.32.11, riding in a valley, tropical forest at the start of the dry season.

    I checked the track points on a proper viewer and ensured it was every second, the earlier picture was from a track creator so wasn't showing the points. I then overlaid both the Forerunner track and one created with an iphone and they both look to wander quite a bit. The 520 track is much tighter with smooth curves that better resemble the actual trail.

    I then down loaded and overlaid a track from another Garmin 520 that as a speed sensor attached and is using GPS/GLONASS. His tracks are very similar to mine but shows approx 5% longer distance over the same track.

    I tentatively conclude the following:

    1: The Forerunner and iphone suffer quite a bit of wander potentially adding distance
    2: The 520 records a much tighter track and due to the twisty nature and forest cover may be loosing distance
    3: The actual distance is probably in between (the other 520) and will be confirmed with a wheel speed sensor
    4: Yes this is driving me mad but I think we've got somewhere.

    Thanks guys!!

  25. #25
    mtbr member
    Reputation:
    Join Date
    Oct 2018
    Posts
    571
    Now you have the challenge of calibrating the wheel sensor!
    If they are like the older magnet style sensors. I think I've read that the wheel/speed sensor will automatically calibrate one wheel rotation, I can't remember.

    But if it is manual, I feel your pain already.

  26. #26
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by Forest Rider View Post
    Now you have the challenge of calibrating the wheel sensor!
    If they are like the older magnet style sensors. I think I've read that the wheel/speed sensor will automatically calibrate one wheel rotation, I can't remember.

    But if it is manual, I feel your pain already.
    With the old magnet sensors that had a magnet on a spoke and a sensor on the frame, you would do a roll out measurement. Set the wheel valve straight up/down at the bottom, mark the floor, roll the bike so the tire/valve makes one full rotation, mark floor, measure distance between marks.

    Enter that number into the computer.

    If you wanted to be really, really precise you had a helper and would sit on the bike so as to compress the tire.

    The newer Ant+ style sensors used on GPS and other devices, measure the earths magnetic field and can generate wheel circumference from that. It's magic 'ya ask me.

  27. #27
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by Catmandoo View Post
    With the old magnet sensors that had a magnet on a spoke and a sensor on the frame, you would do a roll out measurement. Set the wheel valve straight up/down at the bottom, mark the floor, roll the bike so the tire/valve makes one full rotation, mark floor, measure distance between marks.

    Enter that number into the computer.

    If you wanted to be really, really precise you had a helper and would sit on the bike so as to compress the tire.

    The newer Ant+ style sensors used on GPS and other devices, measure the earths magnetic field and can generate wheel circumference from that. It's magic 'ya ask me.
    Huh? I thought that the magnetless sensors detected rotations with accelerometers. Motion, in short. After connecting one, garmin will still let you do manual setup with the rollout method. You can even see the circumference garmin estimates based on its auto calibration

    Sent from my VS995 using Tapatalk

  28. #28
    mtbr member
    Reputation:
    Join Date
    Oct 2018
    Posts
    571
    That's cool that the newer systems do calibrate for you.

    I've spent a lot of time dialing in my computers in the past.

    I'd usually do about 5 wheel diameters for better accuracy. Setting proper tire pressure, sitting on it and trying to ride as straight a line as possible.

    Then the challenge is if the computer wheel size is entered in centimeters or millimeters.

  29. #29
    mtbr member
    Reputation: Lone Rager's Avatar
    Join Date
    Dec 2013
    Posts
    5,101
    Quote Originally Posted by Forest Rider View Post
    That's cool that the newer systems do calibrate for you...
    Of course it's using GPS to do the calibration over something like 500m or 1/2 mile. It's best to ride a fairly straight course over this distance, and if the GPS distance is off, the wheel size calibration will be off the same amount.

    Tire circumference is roughly 2m, so if you do a single rotation rollout measurement to +/- 1cm, you're already with in 0.5%. Best to do it with the bike weighted as it will change the effective tire circumference.

    For me, distance accuracy is of little importance. If I'm concerned about my physical performance, it's time and intensity that matters as distance can be highly variable depending on equipment and terrain.
    Do the math.

  30. #30
    mtbr member
    Reputation:
    Join Date
    Oct 2018
    Posts
    571
    Quote Originally Posted by Lone Rager View Post
    Of course it's using GPS to do the calibration over something like 500m or 1/2 mile. It's best to ride a fairly straight course over this distance, and if the GPS distance is off, the wheel size calibration will be off the same amount.

    Tire circumference is roughly 2m, so if you do a single rotation rollout measurement to +/- 1cm, you're already with in 0.5%. Best to do it with the bike weighted as it will change the effective tire circumference.

    For me, distance accuracy is of little importance. If I'm concerned about my physical performance, it's time and intensity that matters as distance can be highly variable depending on equipment and terrain.
    Are you saying that the wheel/speed sensor automatic calibration is done by GPS? That doesn't make sense. If so, why even use a wheel sensor since GPS is already calculating (albeit not super accurate) while riding the trail?

    As for wheel rollout, as accurate as possible is important. Consider the rough estimate over the life of the bike, or the ride.

    Maybe you are better than me (sounds like you are for sure), but in no way can I get a single rotation to be within a centimeter. Considering I need to get on the bike, squish the front tire and hold a straight line for only one rotation, while watching the valve stem rotate around.

  31. #31
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by Harold View Post
    Huh? I thought that the magnetless sensors detected rotations with accelerometers. Motion, in short. After connecting one, garmin will still let you do manual setup with the rollout method. You can even see the circumference garmin estimates based on its auto calibration

    Sent from my VS995 using Tapatalk
    As per the Garmin information - "contains a magnetometer that measures the three dimensional components of the ambient magnetic field"

    https://support.garmin.com/en-US/?fa...Jd3Tnyif9jRSy6

    When I'm commuting to/from work I go over a steel decked bridge. The speed sensor goes nuts if I'm on one of my bikes that has a speed sensor installed.

  32. #32
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by Forest Rider View Post
    Are you saying that the wheel/speed sensor automatic calibration is done by GPS? That doesn't make sense. If so, why even use a wheel sensor since GPS is already calculating (albeit not super accurate) while riding the trail?

    As for wheel rollout, as accurate as possible is important. Consider the rough estimate over the life of the bike, or the ride.

    Maybe you are better than me (sounds like you are for sure), but in no way can I get a single rotation to be within a centimeter. Considering I need to get on the bike, squish the front tire and hold a straight line for only one rotation, while watching the valve stem rotate around.
    Yes, Garmin's auto calibration uses GPS to calibrate. This is why you should do that calibration on as straight of a course as possible, to minimize potential error sources. You can tweak it manually to be even better.

    You don't do a wheel rollout for any computer over a single wheel rotation. That's fraught with errors. You do it over the course of multiple rotations (minimum of 3, preferably more) so you can average it out. Yes, you use the valve stem as your marker. Maybe a dab of paint on the tire to mark the pavement (number of rotations). It's true that measuring this to the nearest mm is damn difficult and most people are unlikely to be able to get it that good. I bet most could get it within a cm to .5cm if they made the effort, though.

  33. #33
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by Catmandoo View Post
    As per the Garmin information - "contains a magnetometer that measures the three dimensional components of the ambient magnetic field"

    https://support.garmin.com/en-US/?fa...Jd3Tnyif9jRSy6

    When I'm commuting to/from work I go over a steel decked bridge. The speed sensor goes nuts if I'm on one of my bikes that has a speed sensor installed.
    interesting. I never saw that document before. I always assumed an accelerometer was used. they never mention localized anomalies of the earth's magnetic field in that article, though. I've hiked in areas where those really screwed up compass navigation and I wonder if their checks and calibrations are able to deal with those.

  34. #34
    mtbr member
    Reputation:
    Join Date
    Dec 2018
    Posts
    83
    Quote Originally Posted by Harold View Post
    Yes, Garmin's auto calibration uses GPS to calibrate. This is why you should do that calibration on as straight of a course as possible, to minimize potential error sources. You can tweak it manually to be even better.

    You don't do a wheel rollout for any computer over a single wheel rotation. That's fraught with errors. You do it over the course of multiple rotations (minimum of 3, preferably more) so you can average it out. Yes, you use the valve stem as your marker. Maybe a dab of paint on the tire to mark the pavement (number of rotations). It's true that measuring this to the nearest mm is damn difficult and most people are unlikely to be able to get it that good. I bet most could get it within a cm to .5cm if they made the effort, though.
    I've never noticed if not getting it to within 2 or 3 mm is inaccurate over a 50 mile ride. Not even sure how to calculate that and I sure don't worry about it.

  35. #35
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by Catmandoo View Post
    I've never noticed if not getting it to within 2 or 3 mm is inaccurate over a 50 mile ride. Not even sure how to calculate that and I sure don't worry about it.
    yeah, that's probably not that big of a deal. there are so many other sources of error in distance measurement that it's probably not even a detectable difference. How much error is introduced from airtime? lifting your bike over trees and big ledges? probably more than being 2-3mm off on your wheel circumference calibration.

  36. #36
    mtbr member
    Reputation: Lone Rager's Avatar
    Join Date
    Dec 2013
    Posts
    5,101
    FWIW: To do a rollout measurement, I put a mark on the tire near the ground and mark the ground aligned with the mark on the tire with the bike weighted. Roll forward 1 or more rotations and mark the ground again to coincide with the mark on the tire. I then measure between the two marks on the ground. With some care, the effective tire circumference can be determined to better than +/- 1 cm.

    Yes. During auto calibration, the wheel revolutions are compared to the GPS measurement to determine wheel circumference. The intent of the wheel sensor is to improve distance accuracy over tortuous tracks, and speed indication accuracy at the slow speeds bicycles can be moving. If, for example, you use auto pause, it will be more reliable and less erratic when using a wheel sensor rather than relying purely on GPS.
    Do the math.

  37. #37
    mtbr member
    Reputation:
    Join Date
    Oct 2018
    Posts
    571
    I lay out my tape measure and set my valve stem at a control joint in the sidewalk. Then roll it the length of the sidewalk in front of my house. I stop when the valve stem is again intersecting the ground. I put painters tape down and make that location. And repeat a few times and average the values.

    Seems like we do the same thing, basically. However, if I roll out 5 diameters and do the math, it's not even close to a single diameter.

    The point I meant about mm vs cm is when the bike computer only offers mm or cm values. It's off at max of 5mm doing that depending on how one decides to round off.

    Saying it is off by 1mm or whatever isn't as accutrate as saying what the percent different is. If you measure and are off by 1%, then the total distance is off by 1%. You won't be off 1% if you pick the bike to carry it over a tree.

    Sure, not important for a 20 mile ride, but if you are using it to track gear and you have 1000 miles, it can add up.

    most of the tighter accuracy was for my road bike when I would ride 50 miles or whatever. I remember once I didn't have the computer very well set up, the metric century ride registered something like 54 miles and the others (with road bikes and a properly calibrated computer registered closer to 65.

    In any case, at least it can calibrate it for you then allow tweaking with the Garmins. When I decide if I do purchase a Garmin computer, I'll have to remember to set it up on a nice open piece of our paved trail system.

    Thanks for the chat guys.

  38. #38
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    I haven’t got a speed sensor yet but Ive decided to believe the 520.

    And then I was in Strava and it told me there was an issue with my recorded mileage upon correction it added approx 0.25 miles o each 9 mile ride.

    Whatever.


    Sent from my iPad using Tapatalk

  39. #39
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by James80 View Post
    I haven’t got a speed sensor yet but Ive decided to believe the 520.

    And then I was in Strava and it told me there was an issue with my recorded mileage upon correction it added approx 0.25 miles o each 9 mile ride.

    Whatever.


    Sent from my iPad using Tapatalk
    https://labs.strava.com/slide/

  40. #40
    mtbr member
    Reputation: James80's Avatar
    Join Date
    Jul 2012
    Posts
    28
    Quote Originally Posted by Harold View Post
    Thanks for that. I wonder if it works the other way and reduces the distance for paths that go wide of the heat map.

  41. #41
    since 4/10/2009
    Reputation: Harold's Avatar
    Join Date
    Dec 2003
    Posts
    24,759
    Quote Originally Posted by James80 View Post
    Thanks for that. I wonder if it works the other way and reduces the distance for paths that go wide of the heat map.
    If you browse their help section, you'll find a few articles on their error detection and corrections. They do quite a lot behind the scenes.

    Sent from my VS995 using Tapatalk

Similar Threads

  1. Garmin Edge 520
    By WR304 in forum GPS, HRM and Bike Computer
    Replies: 744
    Last Post: 04-19-2018, 04:58 PM
  2. Garmin Edge 520 or wait for WAHOO ELEMNT?
    By yzedf in forum GPS, HRM and Bike Computer
    Replies: 27
    Last Post: 06-16-2017, 04:45 AM
  3. Garmin 520?
    By Le Duke in forum Where are the Best Deals?
    Replies: 18
    Last Post: 04-09-2016, 06:22 PM
  4. Garmin edge 520
    By mtnmark in forum Where are the Best Deals?
    Replies: 3
    Last Post: 07-05-2015, 09:09 AM
  5. Intensity vs distance in ultra distance training
    By shackleton47 in forum Endurance XC Racing
    Replies: 6
    Last Post: 11-13-2011, 06:57 AM

Members who have read this thread: 43

Posting Permissions

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

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2019 VerticalScope Inc. All rights reserved.