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

Automatic wheel size accuracy...? (speed sensor)

I have my new magnet-less speed sensor on the rear wheel and have set the Edge 510 to automatic wheel circumference. After a few rides it has settled on 2065mm for my 700c/23mm wheels. According to http://www.bikecalc.com/wheel_size_math and other calculators the circumference should be more like 2100mm. Of course this depends on the tire and tire pressure, but a 35mm difference seems a lot.

What is your experience with the automatic wheel size calculation? One should think it gets very accurate after a while.. I just have doubts when the difference is so large relative to the expected value.
  • That does sound small for 700 x 23 (it would mean that the tyres are actually a bit less than 622/18!).

    But why not just check? Roll the bike along the ground and measure. Putting your weight on the bike will take some of the tyre deflection into account, but if your tyres are inflated correctly it shouldn't make much difference.
  • Is it not the case that the rear wheel of a bike will also take a shorter path than the front wheel - but that should tend to lengthen the apparent size of the rear wheel compared to the bars where the GPS is mounted... regardless you are not measuring like-for-like.

    Plus there's always the inherent error in GPS position in determining the distance travelled (then dividing by the number of wheel revolutions - which may not be an "easy" division... even a 1km test would only be 500 revolutions and if you happened to get the most possible wheel revolutions (501) versus 500 (assuming a nominal 2m circumference) or the fewest possible (499) both introduce errors... add in the uncertainty of the GPS position (±3m at absolute best - so your 1000 could be 994-1006 *at best* since the GPS start and end positions are both uncertain to ±3m) assuming that the path in between is taken and measured perfectly and you can easily count up a cumulative error of ±5% or more in the automatically determined wheel circumference.

    For an abstract example, there are many texts available - https://www.lhup.edu/~dsimanek/scenario/errorman/propagat.htm is one such.

    Agreed that the way to find out for sure is to measure it... it's still not going to be perfect - those darned errors again - but 10-20mm in 2000 is more attractive than 5% :)
  • MCHUNNYMONSTER: That's a good explanation of inaccuracies in the calculations. I had not thought of those details. If the Garmin keeps calculating the wheel size over shorter distances, which it appears to be doing, it is limited in accuracy. I usually run 95-100psi for rough roads on my commute, but I tried 110psi this morning, and it changed from 2065 to 2081mm over a 9km ride. Previously it had been 2085mm, then 2065mm after the last ride, without changing the tire pressure. I guess I will have to go back to measuring and compare.
  • Previously it had been 2085mm, then 2065mm after the last ride, without changing the tire pressure.
    The wheel size setting is always being calculated and can be updated at any time. So when you check the setting there is no way of knowing for what percent of the ride it was at that setting. The setting does jump around some but overall I think it settles on a good number and is accurate.
  • After a 3rd ride today it went up to 2091mm, which is pretty close to the expected value, with tire pressure 110psi. I think it started low due to lower tire pressure. I will keep an eye on it, but it seems to be working well... It's an easy calculation, except the inaccuracies mentioned, but over time it should get to a relatively accurate average.
  • Hi,

    Forgive my ignorance, but, why not just manually enter it based on the chart in the manual? What are the advantages and disadvantages?

    Thanks in advance.

    Moose.
  • I agree, enter it manually. Although 99999 out of 100000 times it will be spot on with the auto calibration, why risk the chance that it'll choose the most important ride of your life to mess up.

    There is the argument that different tire pressures will affect the size of the wheel. But I check mine frequently.

    A rollout test is more accurate than the tables and very easy to do.
  • +1 to entering manually. Earlier this year I was having serious discrepancies in speed and distance between Garmin and Strava, turns out my "automatic" wheel calibration had changed and was using a value that was WAY off. Now I measure my actual wheel circumference (lines drawn on the floor of my garage) and enter my data manually.

    Note that things like PSI and weight (if you're carrying loads of various sizes) will change your wheel circumference so it's best to measure often when things change about the bike.
  • I measured my Trek DS 8.3's 700c/38mm tires with chalk lines on the driveway, but then also let the sensor/Edge500 do the auto-setting. I measured 216.5cm on the driveway and the Garmin measured 216.6 by GPS ... I was impressed! :) And, it hasn't changed in about 10 rides....
  • Seems to be the 1 time in 500 rides that ticks people off. That's always the most important ride of their life.

    As all my rides are relatively important to me, I choose to set my wheel size myself. That way I won't have any auto-calibrate issues.