| 700 X 44 | 236 |
| 700 X 38 | 231 |
| 700 X 35 | 230 |
| 700 X 32 | 227 |
| 700 X 28 | 225 |
| 700 X 25 | 223 |
| 700 X 23 | 222 |
| 700 X 20 | 221 |
| 27 X 1 3/8 | 232 |
| 27 X 1 1/4 | 230 |
| 27 X 1 1/8 | 228 |
| 27 X 1 | 226 |
| 26 X 2.125 | 225 |
| 26 X 1.9 | 220 |
| 26 X 1.5 | 212 |
| 26 X 1.25 | 206 |
| 26 X 1.0 (559 mm) | 205 |
| 26 x 1 (650C) | 206 |
| Wide Tubular | 224 |
| Narrow Tubular | 223 |
| 26 X 1 3/8 | 222 |
| 24 | 205 |
| 24 x 1 | 188 |
| 20 X 1.75 (406) | 158 |
| 20 X 1 1/4 (451) | 173 |
| 17 x 1 1/4 | 142 |
| 16 x 1 3/8 | 137 |
If calibrating these Avocet computers using a rollout test:
The numbers are from an original Avocet 20 manual. The two results are slightly different. Use the appropriate result for greatest accuracy.
Why are the numbers different? Warning: math ahead!
Avocet explains that 2.7273 is 30 / 11 (actually the repeating decimal, 2.727272727...)
30/11 is 4 times the ratio of miles per hour to feet per second, 15/22.
These Avocet computers are unusual, then, in being calibrated directly to miles rather than to miles by reference to kilometers.
But for kilometers, the same computers ask for the circumference in millimeters x 0.108.
Now let's see how the mile and kilometer calibrations compare:
25.4 is the number of millimeters in an inch.
25.4 * 0.108 is 2.7432 -- slightly but significantly different from 2.7272....
2.7432 *11/30 is exactly 1.00584
There are exactly 1.609344 kilometers in a mile, and 1.609344 / 1.600000 is also exactly 1.00584.
Therefore, the calibration numbers for kilometers will be slightly different, because the computer uses the rounded factor of 1.6 when converting miles to kilometers. The error if not adjusting for this is over 1/2 of 1 percent, which is significant, as bicycle computers can be calibrated by rollout to greater accuracy. The Avocet manual says to add 1 to the calibration number for kilometers, but this is only an approximation.
Also, the inductive sensor generates a stronger electrical signal, the faster the wheel is turning. For this reason, the computer does not respond at very low speeds. If the sensor is placed as close as possible to the magnet ring, response will begin at 1 mph or less, and the resulting error is negligeable.
