I bought two new Armadillo tires on January 12, 2006. I pumped the front tire up to 120 psi before each precalibration. It is rated at 115 to 125 psi. I did this until July 1, when I was out measuring on a very hot day and the bead popped off the rim and the inner tube exploded. Fortunately I was not far into the measurement and came back next day.

I decided to use 100 psi as an initial pump-up pressure, and have done so since.

Here’s how my constant was affected:
Original Post
Pete:
Until about two years ago I was very much a high-pressure man and with my sew-ups in multisport competition used pressures as high as 170 psi.

However when I started pressure-monitoring in course measurement, I decided to use about 80 psi for the front wheel because I thought I would more likely get predictable results when the tire was not operating near its pressure limit as recommended by the RRTC manual. Indeed the calibration factor has proved to be remarkably predictable from the pressure reading of the tire. Evidently measurement results are not in the least degraded by using this low pressure.

When you verified my results with the pressure-monitoring gauge, you used 120 psi rather than 80 psi because you were not willing to degrade your performance by increasing rolling resistance. Actually I have not noticed any degradation with 80 psi on the front tire in the relatively short distances of course measurement especially as I keep 110 psi on the rear.

The change in calibration factor at constant temperature you observe is quite predictable from my recent findings. I have found that at constant temperature the pressure coefficient is negative 1.00 plus or minus 0.05 rev/km/bar. This does not seem to vary going from 700 to 27-in tires or from high to low tech tires. Assuming 1 rev = 23.63 counts and 1 bar = 14.50 psi, the coefficient in the units you use above is negative 1.63 plus or minus 0.08 counts/km/psi. Thus your reduction in pressure of 20 psi would be expected to give an increase of 32.6 plus or minus 1.6 counts/km. Inspection of your results above shows 20-30 counts/km.
Last edited by neville