# Setting your tire circumference for cycle computer. Compression measurement?

Just wondering how you guys set the circumference for your cycle computers. Do you pump up to your favorite PSI, and then have someone measure it while you’re sitting on it to factor for compression of the tire? Or do you just measure it normally and ignore the excess?

Ideally you should do a roll-out test; put chalk or water on your tire, ride it, and measure the distance between spots.

I’d imagine that inconsistency in your spin or any veering left or right could offset or compound any minimal change in tire shape based on rider weight.

If what you’re trying to do with your computer is measure your speed and distance over the ground, you want to include wobble in your circumference calculation.

true enough. I wonder if counting full crank revolutions over a few hundred feet, then dividing the distance by number of revolutions would give you a more accurate average. That is, unless you’re super smooth ; unlike myself… :o

To determine the most accurate calibration number for your application (tire choice, body weight, etc.), or if your tire size is not listed in the chart you have, perform a ‘tire roll out’.

To do a tire roll out:

1. Mark a spot on the floor, and line up the tube’s valve stem with it.

2. Running the usual tire pressure and with the rider’s weight on the uni, roll the tire out one full revolution. When the valve stem comes back around to the ground, make a second mark on the floor. Measure the distance on the floor from point to point.

NOTE: This is often best performed parallel to a wall, to ensure the roll out procedure is done in a straight line. Or you can include some “wobble” to make it a bit more “real world” accurate.

1. Record this number in cm if your computer requires a 3-digit calibration # (or multiply inches by 2.54 for cm). Record this number in mm if your computer requires a 4-digit calibration # (or multiply inches by 25.4 for mm).

To confirm my calculation on my strada wireless, I zeroed out my trip meter and rode one mile, according to markers on the beach bike path. It was almost exact, and the difference was quite negligible.

In 2004 I did some statistical work on the actual roll-out value of Coker tyres, which at that moment in time was the only brand offering a 36" tyre. I also made some comments on what rollout actually means, it could be straight rollout (loaded or unloaded), or dynamic road rollout.

My report is on http://www.xs4all.nl/~klaasbil/coker_rollout.htm.

(It includes a download link for a spreadsheet that will estimate your Coker tyre rollout based on rider weight, tread wear and tyre pressure. That spreadsheet has lost some of its applicability since many riders now have other tyres than the original Coker. Those other tyres have probably slightly different dimensions.)

Why would you factor in wobble? That’s still distance that you are traveling, just not straight-line distance.

Because most people are interested in measuring straight-line distance. We measure the marathon course to be 42km in straight-line distance; if you actually ride 43km due to wobble, that doesn’t mean you get to stop 1km before the end.

Say you want to train the exact distance of a marathon for an upcoming race. If you have done 42.195 km according to your cycle computer, you haven’t reached your goal yet. Any race is about going from A to B being some specific distance apart, not about a traveled distance the way you define it.

Note that wobble typically “consumes” a few percent.

Edit: essentially saying the same as Tom.

Ah ok. So probably the best way to check after calibrating would be to measure out a perfect mile on flat ground, and ride it and adjust until it hits the mark?

Now you’re getting into the vagaries of measurement; how do you measure out a perfect mile on flat ground? Don’t use your car, it’s almost certainly wrong. Perhaps you’re best off going around a track four times, but how do you know the track distance is accurate? Or whether a quarter mile is at the inside of the lane or center of the lane?

Ride the bike tour of a 26.2 mile marathon. The event organizers are pretty good about getting the distance right. The last one I rode in came out pretty much on the button with my computer. None of this is an exact science anyway, so I wouldn’t stress too much about it.

I don’t expect it to be perfect, but within range of reasonable error.

Just use the formula in post #6. It will work just fine, but you could also do the same measurement a few times, just to confirm the results, and rule out any possible mistakes. It’s also better to have a second person helping, to mark the spot when/where your stem completes one full revolution, with you ON your uni, of course.

+1

Also mile markers on the road are often estimates. For motorists between any two markers if the distance is off a few % between markers (like 1.05 or 0.95 mi) it doesn’t matter as long as the errors average out over a few markers.

I’ll just take my tape measure to LAX and measure out a mile on the runway.

2800 rule

Or just use “2800 mm”, which will be accurate within a few percent. For a new tire with high pressure and a light load it will be a little bit high. For an old tire under lower pressure with a heavy rider it will be a bit low. From my own measurements I would guess an uncertainty of less than ± 2% for most riders under most conditions.

Scott

MuniAddict’s step-by-step in post #6 is the easiest way to do it and be accurate, though you don’t get a “real-world” wobble that way. I suppose if you want to factor in your wobble, do something like 100 revolutions (in a straight line) at your comfortable cruising speed. Start from a mark on the ground, line up your valve there and walk back a few revs to get started. Then cruise as straight as you can for 95 revs, jumping off as you hit 100 and marking the valve spot again. Then have fun measuring the rollout. After that, simply move the decimal over two places for a single revolution (use the metric system, eh?).

I have a 30m measuring tape, but that’s not nearly enough for 100 revs. Maybe do a shorter rollout and figure out the math from there…