The next step in my modding of my home bike trainer is to completely replace the control computer by my laptop with USB-IO-card.
I have shown how to record rotational speed and control voltage for the eddy-current brake in a previous post (https://technikgebastel.blogspot.de/2016/02/bike-home-trainer-i.html).
Since these are the only input quantities for the control computer, I now have to reverse engineer the formula for the dependence of the performance P in watts on the rotational speed f_rot and brake voltage U_brake.
Fortunately the bike computer has a program, where it adjusts the brake voltage depending on the rotation speed, so that a constant work has to be performed. This is very helpful to explore and measure the dependence of the performance on the brake voltage and rotation speed.
The following graph shows the measured brake voltage versus the rotations speed for a fixed performance program (50 watts and 55 watss):
Even though there seems to be a lot of noise (not sure if it is because of the mesaurement setup or something else), there is a clear difference between 50 watts and 55 watts. Now we only have to record enough data during different constant performance runs and interpolate.
Another and probably faster way would be to simulate the input for the bike computer, i.e. the pulses at every full rotation, and measure the corresponding brake voltage without the brake actually connected to the computer to avoid damage.
