The resolution of the Jones Meter is more than adequate for measuring race courses, but much less so for calibration courses particularly if they are short. Users of the Jones would be better to use the rim-reading method instead for wheel calibration with the following advantages:

1.More accurate determination of the calibration factor.
3.Instant comparison between individual rides over the calibration course.
4.Fewer calculations.

For a 36-spoke wheel the spokes should be numbered 0 to 35. (Even though this takes only a few seconds, inexplicably several old timers like Rick Recker continue to count spokes.) In electronic counting it is debatable whether to mark the rim in decimals instead, but there is no debate in this application.

At start, the tire should be pumped up to the same pressure each time, so that a known number of whole revolutions will be obtained on the course. There are several ways of determining this number initially including just measuring the distance covered by one revolution over the ground.

A calibration run is made by placing the zero spoke over the start and taking a reading at the finish in tenths of spoke intervals. In postcalibation the number of whole revolutions should be checked in case there has been a large air leak.

For instance for a calibration course of 300 m requiring 143 rev, the following formula gives the calibration factor with the SCPF for a Jones with gearing of 23.6363:

Jones counts/ km = 11,277.9 + 0.0548 (s1+s2+s3+s4)

where s1-4 are tenths of spoke intervals on a 36-spoke wheel.

Substituting readings of 180, 182, 183, and 181 in the above formula gives a calibration factor of 11,317.7 cts/km.
Original Post

The precision of calibration is indeed improved by using Neville’s technique, but I wonder about accuracy. I think I calibrate as well as most, and I have been tracking the variation of my calibration rides. Over 56 calibration occasions – all with the tire pumped to an initial 100 psi. I have obtained the following variation (I’ve converted revs to spokes):

Variation over 4 rides
***revs ****** spokes
avg 0.0364 *** 1.31
high 0.0800 *** 2.88
low 0.0100 *** 0.36
SD 0.0187 *** 0.67

My variation is about four times that of Neville’s example. Given this, does using spoke-counting actually give me a meaningful increase in accuracy? I don’t know.
In my calibration course length experiments I saw only a small reduction in the scatter of 300m rides when I went from Jones counter readings to rim readings.

For shorter cal course lengths the reduction in scatter was much greater, so that the scatter of 100m cal course rides using rim readings was very similar to the scatter of 300m rides using Jones readings.

https://measure.infopop.cc/eve/forums/a/tpc/f/69510622/m...401090642#7401090642
Pete:
My example above was only a fictitous, but typical of what I can get on a good day. On March 14 I actually got rim readings of 236, 237,237,and 236. In other words the total spread throughout the four rides was only one tenth of a spoke interval! I note that your spread of four tenths on a good day approaches this.
The error in calibration is not the limiting factor in the overall mesurement of the race course no matter what method is used. However I think the rim-reading method is easier to use and allows accurate calibration over 25 meters as I will show in a subsequent post.
Last edited by neville