So I just bought myself a nice new shiny $20 speedometer from Canadian Tire. When you turn it on, like most speedos, it asks you to enter your wheel circumference in mm. The manual for the speedo included a reference guide for a few common wheels, but not mine. My wheel reads: 37-622(700x35c - 28x1[5/8]x1[3/8]) (it is an adult hybrid bike); None of those values have units marked next to them, so I have no idea what is mm, what is inches.... I just have no idea what any of that means. But the closest reference in the manual was "700x32c", while mine is "700x35c", so I just entered that in, which was 2155 (which works out to 686mm). Well my speedo's odometer was about 0.5km off my GPS's odo on a 10k trip, so its obviously not the right one.

So I'm just going to measure the circumference in mm myself. But where exactly do I measure, which circumference is my speedo asking for? This site says you measure the "bead seat diameter", while my father recommends to just measure the distance one revolution of the wheel moves the bike, which would be a measurement of the tire and not the wheel. Doesn't my speedo need to know the circumference of the tire, like my father suggests, not the bead seat? Otherwise different tires, and different tire pressures, could significantly affect the readouts! But I don't want to enter in the tire circumference and find out that the speedo automatically calculates the difference the average tire would make.

I would just do trial and error here, but each trial requires a 10k bike trip to check for accuracy with my GPS. Any help would be greatly appreciated!

If you want the circumference make a small mark on the edge of the tyre (on the outside) and a mark on the floor, and align the two. Roll the bike forward until the mark on the tyre is against the floor again. Measure the distance between the two marks on the floor.

The speedo has no way of knowing what tyre you've got on the wheel so all it knows is that each time the magnet goes past the sensor you've moved as far forward as the circumference of the wheel. In this context "wheel" means "rim plus tyre".

You don't really need to do a 10k trip to check it against the GPS, to find out if you're about right do a 1k trip and if they look like they align feel free to do a longer trip if you feel the urge.

You should listen to your father.

This will get you really close if a rollout of the wheel measurement seems inconvenient.
Measure from the center of the axle to the outer radius of the tire. Where the rubber hits the road. Maybe reduce that number by a couple of mm to account for tire squish while riding.
Multiply that measurement, times Pie, times 2.
So, here's an example.
(Radius)x(2)x(Pie)
342x2x3.147=2155
Radius=342
Pie=3.147

Okay, that's what my father suggested. I just wasn't sure if this was the right way to do it, since the other website seemed to suggest that you must measure the circumference of the wheel, and the speedo will add the average tire to that value. It makes a lot more sense if the speedo just multiplies circumference by revolutions then divides by time. So I'll just do as you suggested, which will include the tire in my measurement.

Oh but I really do need to do a 10k trip to check against the GPS I set it up yesterday, and the odometers on both sensors were exactly equal, until about 3km and they started to drift apart. Perhaps they actually were equal and then my tire pressure changed dramatically? Who knows.

Thanks for the help, contango!

+1 you want the "roll out" which is the distance the wheel travels in one revolution. I've found the easiest way is to mark the tire with something which will transfer, like chalk, or slow drying paint or ink. Then ride the bike making marks on the pavement as you go along. Now you can measure the distance between the marks for an accurate roll out calculation.

BTW- if your mark lasts long enough to make 3 or 4 dots on the road, measure over the maximum distance, and divide by the number for a slightly more (far more than you need) accuracy.

mmmm.......... Pie.

Do the tape measure thing. Inflate the tires to proper pressure. The result will be more accurate than most car odometers.

Your father is right. Roll it out for the most precise measurement.

To estimate (close enough, in truth), the diameter is 622 plus 2 times tire diameter, times pi for the circumference. Your 35x700 would be about 2174 with that formula.

These numbers will never be exact if thats what you are trying to get it will never happen.

Good thing you checked with us first. Parents can be so dense.

If your first try is off you don't need another test ride All you have to do is multiply the percent difference between what the speedo says to what you think it should be by the initial number you put in. Add or subtract as needed. Example: Setting is 2155, is .5/10 off, or 5%. 2155*5% is 108, so add or subtract that from 2155 (or just take 95/105% of it)

For anyone who saw my first post I plead painkiller fog

I get a 5% error if it's off .5 km on a 10 km trip. That's pretty far off.

Your father is correct, but with the small nitpick that it measures the entire wheel+tire. Do as contango suggests and you will have the most accurate reading for your computer.

measure Roll out .. natural materials approach ..
ride thru some dog poo, and measure the distance between the marks on the sidewalk
when the poo patch on the tire comes around again,
and makes another couple marks on the sidewalk.


if you really want precision lay down a thin piece of tape, put a dab of paint down.
pull up the tape, that leaves a clear line in the paint dab..
then there is that line to measure between.

"WARNING, I'm from New York. Thin skinned people should maintain safe distance."
I'm in Caly, and I've met a lotta NY'ers here, and frankly I like the in-your-facedness of people from NY. Refreshing change.

If you do the roll-out method, be seated on the bike for better accuracy.

Listen to Dad. Also you want to make sure your tires are inflated to normal levels and you do the measurement with your normal weight on the tire.

You want to be accurate but don't obsess. The tire will change circumference on the road due to inflation changes so it will always be "just about right"

-G

Thanks,
question -- does "ride the bike", not clearly enough imply sitting?

Yeah, I never bother measuring with a rollout any more and just use this formula with one tweak: I subtract 1 before multiplying by pi: (622+2*width-1)*pi

The rollout on my bike is 2086 mm using Kenda Kaliantes so I put that in the computer. I always ride with a garmin forerunner gps there are very very close. Did an 81 mile ride labor day and the garmin said I did 81.43 miles. That is about the error ratio I get every ride.

Marking the tyre with oil or something is great outdoors, less great if you live in an apartment and don't want sticky marks on the kitchen floor

If they matched each other exactly for 3k but by 10k were 500m apart that suggests something else might be going on. If the problem were nothing more than your wheel circumference being out then by 3k I'd expect them to be 150m apart.

If you've experienced a bit of GPS drift that could potentially explain a 500m difference. If you rode through an area with dodgy GPS reception that kind of thing can happen. I've seen a couple of places where roads go through cuttings and every single time, whichever GPS I'm using, my GPS shows me drifting about 100 yards off the road as I go into the cutting before snapping back to the road as I emerge from it. There's a clear view of the sky, so I reckon the rock face reflects a strong enough GPS signal to confuse the processor. Whatever is going on, a situation like that adding a 100-yard drift off-road and then a 100-yard drift back onto the road immediately puts circa 200 metres of extra distance onto the log. If you've been through tree cover or a tunnel or some such that will do it as well.

OP: for what it's worth, I never believed anything my dad said either. But there was no internet to second-guess him with back then, so he should be grateful. As for my own kids....

Follow the father rule saved a British flotilla in battle once...can't remember if it was World War I or World War II...

=8-)

If you do a rollout with a bit of weight on the front wheel you will be less than 1% off. For that figure I am assuming max 3mm variance in the distance from axle to road, which is quite a bit. There's going to be some variation no matter what depending on what you are carrying and tire pressure.

Sample distance from axle to road - 343 mm.
Alternate distance from axle to road - 366 mm
Difference is .9%

Absolutely right. I maybe made the mistake of assuming that folks would figure that out and ride the bike outdoors. Of course, if you live an an apartment building with long straight tiled hallways, you can do it there and let the janitor clean up after you.

When I was young I had a part time job delivering groceries. There was another person who rode a bike, and after hours we'd have parimutual track races inside. Riding competitively on a flat linoleum track teaches bike handling, especially the day after they wax the floors.