Wheel Circumference

[Bristolian]

I'm not sure this is the right place for this so MODS feel free to move it.

All of the wheel circumference charts I can find say that a 700 x 28c wheel and tyre should have a rolling circumference of 2136mm and that is what I had set in my Garmin. About a week ago I needed to change the battery in the speed sensor and since then my Edge 530 has taken to dropping the sensor data during a ride, which I find somewhat annoying. I read elsewhere that Garmin bike computers compare the speed readings from the on-bike sensor and that calculated from GPS and if there is a significant difference they stop reading the sensor data and default to the GPS speed. How much of a difference is needed to be considered "significant" appears to be a secret. Yesterday, after another frustrating ride with no speed displayed or recorded, I decided to check the figure in the Garmin and it was something quite a bit smaller than 2136 so I re-entered that but haven't had a chance to try it out yet.

Earlier this afternoon I decided to actually measure my rear wheel - I know, I need to get out more

To my surprise, the actual circumference is 2145mm which is pretty much what the charts say a 700 x 30c wheel and tyre should be (2146mm). I have checked this measurement several times (in several different ways) and get the same result each time (+/- 1mm) so I'm pretty sure it's correct. Is this unusual or not? Am I reading the charts incorrectly? Would a 0.4% speed error be enough to cause the Garmin to play up? Should I enter the actual circumference or leave it at the chart figure?

All comments and advice will be much appreciated

 

[Ajax Bay]

Right, well your/our Garmin normally detects speed (by GPS) and deduces the actual wheel/tyre circumference and uses that to display speed if the GPS fails eg in a long tunnel, provided you have a "speed sensor" fitted. Only if the signal drops out does it revert. You can set it to not auto-calibrate, but why?
https://www8.garmin.com/manuals/webhelp/edge530/EN-US/Edge_530_OM_EN-US.pdf
https://www.gear-calculator.com/?GR...21,24,28&UF=2150&TF=90&SL=2.6&UN=KMH&DV=teeth
suggests 2150mm circumference
Circumference depends on the actual width (and specifically height) of the 28-622 tyre which is never exactly 28mm (as well as make/model, depends on tyre pressure and internal rim width).
Maybe your issue is you have set the Garmin to display the 'speed sensor' speed (actually measures rotations of the wheel per second) and when it fails, speed display fails.
Bike Settings, Bike 1, ANT+ Spd/Cad, connect with working 'speed' sensor, auto-calibrate. Ride on.
I have taken mine off, fwiw. But I don't ride those bikes on the turbo. If you've cycled yours outdoors for a mile or so it will have auto-calibrated: put it on the turbo and it should give a notional speed (you are stationary).
I have highlighted useful google words above

 

[Bristolian]

HI Ajax Bay, and thanks for the response. So the Garmin treats the GPS speed and sensor speed the other way round to what I thought? Interesting. I have noticed that I sometimes get a message on the Garmin that it has auto-calibrated the wheel size but I've never bothered to look and see what it comes up with. If I have manually input 2136 into the head unit would it replace that with the auto-calibrated figure? I think I might take the speed sensor off the bike and put it onto my turbo trainer wheel.

I can't find how or where to see if I have my speed to be measured by gps or sensor. Maybe it's not an option on the 530.

The gear calculator you linked to is very interesting. I'm thinking of changing my front chain rings from 3x to 2x with 50/34 compact rings and this will let me see the effect that will have on speeds and overlapping ratios. Thanks for the link.

 

[Dogtrousers]

Well, if we just do the sums, we know your circumference is 2145mm. We also know your bead diameter is 622mm. That tells us that your tyres are 30.4 mm "thick" ie they stand up about 30mm over your bead. (2145/π) is your actual diameter D; (D-622)/2 is your tyre thickness.

So your tyres would seem to be sticking up 2mm more than the expected 28mm. Is this unusual? - I have no idea. It doesn't seem like a lot as tyres are known to vary from their published dimensions. But you'd need to talk to someone who cares has spent a lot of type measuring tyres.

As @Ajax Bay says

Circumference depends on the actual width (and specifically height) of the 28-622 tyre which is never exactly 28mm (as well as make/model, depends on tyre pressure and internal rim width).

 

[Venod]

So your tyres would seem to be sticking up 2mm more than the expected 28mm. Is this unusual? - I have no idea. It doesn't seem like a lot as tyres are known to vary from their published dimensions.

I find tyres vary wildly from what I expect, different makes of the same size may produce different measurements also the internal rim width has a big influence on tyre width and height, a wider internal width will produce a wider tyre profile, but with less height, a narrow internal width produces a narrower profile but taller tyre, I stopped using a speed sensor and just rely on GPS nowadays, it's correct most of the time, but I did note today under heavy tree cover, my indicated speed was about half of my actual speed.

 

[Ajax Bay]

My 28-622 Conti 4000s come up wide and high on my 15mm internal rim width @90psi (rear).
The 'auto-calibrate' suggests that the actual diameter is 2201mm.
On my wired front sensor, I need to enter 2125mm in to get that display to align within 0.2% of theGPS delivered distance.
All this 'for interest'.
But on the turbo I really don't care: heart rate is my metric for that (plus elapsed time).
Have a read of DC Rainmaker's take on all this: https://www.dcrainmaker.com/2010/01/how-to-train-indoors-with-your-garmin.html
Doesn't include the 530 but it's roughly the same:

 

[presta]

700-35 Marathon measured rolling along the ground: 2187mm
Cateye instructions: 2170mm
Measured over 20 miles using the DoT marker posts at the roadside: 2150mm

It doesn't surprise me that rolling it out along the ground is useless because loading a tyre compresses it, and reduces the radius. The increase in measured distance around a familiar circuit is a very sensitive way of seeing your tyre pressures dropping.

 

[slowmotion]

If you want to get an accurate measure of circumference, ride the for a bit under prevailing ambient temperature, inflate them to a known pressure, and do a roll-out test with you in the saddle. Ignore the clever assumptions built into a piece of consumer electronics.

 

[Ajax Bay]

If you want to get an accurate measure of circumference, ride tyres inflated to one's preferred pressure for a bit, and ride for a mile or two outside. The clever consumer electronic device will derive accurate speed from GPS and corresponding rotational speed from the 'speed sensor' on the rear hub or attached to the LH chain stay (+ magnet on a spoke) and auto-calibrate the actual (and accurate) effective wheel/tyre circumference. To find out that figure, go in to the Garmin's 'settings' to find out.

However this measurement is not useful, in this context. The speed displayed will be per GPS (unless disabled or in a tunnel) or if the bike is ridden on the turbo, speed = 0 and whatever's displayed as speed really doesn't matter.

do a roll-out test with you in the saddle.

Have you actually done this @slowmotion ? How confident are you that your result is accurate within a tolerance of (say) 1% (22mm)?
What did you do with the information?
Temperature change is going to have a negligible effect on tyre height (pressure increase is about 1psi every 3 degrees C) btw.

 

[Donger]

When I buy old-fashioned cheapskate cycle speedo/milometers (the types using magnets attached to a spoke on your front wheel), I have a route that I know from previous experience to be exactly 20 miles ..... my bread and butter local ride. If you pump up your front tyre exactly to your favoured pressure and set your new computer to the recommended circumference, then go for a test ride along your known distance route you can establish the accuracy. (The longer the known distance the better). Any discrepancy between the measured distance and the known distance can then easily be corrected by using simple maths to determine the percentage error and adjusting the circumference settings accordingly. Over the years this has served me well, according exactly to the published distances on audax rides and agreeing with clubmates' computers on group rides.

Once the correct circumference has been calculated and entered into your computer, all you need to do is to ensure a steady tyre pressure in your front wheel , which I check regularly. Visually, you can pretty much see that it is correct anyway, from the width of the dry line on your tyre when riding along. I also keep a note of the correct circumference, so that I can enter the correct one quickly when I have had to change the battery.

 

[Ajax Bay]

. . . cycle speedo/milometers (the types using magnets attached to a spoke on your front wheel),
Once the correct circumference has been calculated and entered into your computer, all you need to do is to ensure a steady tyre pressure in your front wheel , which I check regularly.

I also keep a note of the correct circumference, so that I can enter the correct one quickly when I have had to change the battery.

Absolutely @Donger (both the above).
Use of a GPS (see my posts above) is a more assured way of getting an accurate circumference figure (unless you don't use one). How do you "know from previous experience [your loop is] exactly 20 miles"? Within 1% accuracy? GPS?
This is the only use for the effective circumference of a tyre: inputting in to a simple (non-GPS) wireless or wired 'consumer electronic device' which depends on number of rotations (for distance) and rotational speed (for speed and derivatives).
I observe that, within a few psi, the height of the tyre will vary very little. I assume (maybe wrong) that riders habitually inflate front and rear tyres to the same chosen pressures.

 

[fossyant]

I just let my Garmins sort it, life is too short !

 

[presta]

ride for a mile or two

It takes a lot more than that, which is why I did mine over about 18 miles. To get an accurate measurement the resolution of instrument you're using needs to be significantly smaller than the increment you're trying to resolve. The resolution of a cycle computer is 0.01m, which is only 1 in 200 over two miles, so that's the same size as the increment in the circumference that you're trying to measure if you want your answer to the nearest centimetre, and 10 times too big if you want it to the nearest millimetre. Since my computer doesn't adjust in increments smaller than 1cm (1 in 215) anyway, a distance that gave me a resolution of 1 in 1800 was enough. (I did try asking the DoT how accurate their marker posts are, but they didn't reply.)

Temperature change is going to have a negligible effect

Yup, a range of 0 - 30C is ±5%, which gives a variation in circumference of about ±0.025%.

Being as there's a good 2-3mm of tread on the tyre, that's another 0.5 - 1% error if you don't keep remeasuring as the tyre wears.

 

[Ajax Bay]

The resolution of a cycle computer is 0.01m, which is only 1 in 200 over two miles,

0.01m = 1cm
"Two miles" = 3,219m (4 sig figs)
Resolution = 1:321,900
HTHY with arithmetic
Two miles will allow satisfactory (ie accurate) auto-calibration by GPS device.

 

[C R]

0.01m = 1cm
"Two miles" = 3,219m (4 sig figs)
Resolution = 1:321,900
HTHY with arithmetic
Two miles will allow satisfactory (ie accurate) auto-calibration by GPS device.

Thank you for confirming I haven't lost it yet.