|
|
Measurements on speedo- and odometer (long).
Joris Creyghton -- Wednesday, 7 April 1999, at 10:25 a.m. I have been busy checking the accuracy of both speedo- and odometer of my car and found some interesting facts, which I would like to share with you. I measured speed at about 20,80,100,120 and 140 kph and fitted a straight line through these points resulting in: Speed = 0.960 * (dig.speed) - 3.30 kph. In words this reads: lower the speed on the digital speedo display by 4.0% and subtract another 3.3 kph to get your actual speed. I got 10 measuring points (did them all twice to test reproducibility) and they were all very well described by the above *straight* line, none of them more off then 1 kph. In an independent measurement over 50 km I found the odometer to read 4.0% high. The interesting thing now is the equality of these two 4% figures. I suppose that the odo measurement relies on counting pulses from the revolving wheels and that the speedometer uses the same info divided by some time interval. Now since timekeeping in the Boxster seems to be OK to tell from the accuracy of the clock, the error in both speedo and odo stem from the inaccurate translation from a number of pulses to distance. The "offset" of 3.3 kph is, I think, purposely added by PAG, maybe to comply with some silly rule that the speedo may never read low ! I also observed that the speedo apparently is not ment for speeds below a certain value: the lowest non-zero reading I get from mine is 5 kph; so don’t think that I’m driving backwards at 3.3 kph when my speedo reads ‘0’ :-). CFG posted some info (8 March 1999) from a German car magazine about the accuracy of the speedo of a Carrera 4 and the slope of the line through the points he presented is 1.000, with an offset of 4.2 kph (this was correctly signalled by Shawn from DFW in the same thread). From that data I would expect the odo of that car to be accurate within 1 % (Chris ?) Since the speedo appears to perfecty linear (as one would expect from its principle of operation) only two measurements should be sufficient to calibrate it properly: one speed measurement at a convenient speed (high enough to get rid of the one digit uncertainty of the digital speedo) and a calibration of the odometer, which can be easily done very precise because it’s an integrating device. One final warning: in chasing for perfection you should realise that the error in both speedo and odo will increase with about 2% due do tread wear over the life cycle of your tires :-). Joris from Holland. |