relax bunch boundary check, fix continuous pulses (25 ns pulse in two subsequent bunches) -> constraint to the help

not a real issue, the sub-bunch period is not calculated from clock sensing, but from configuration (call to Ffpg.SelectClock("external", 400.789))

initial reset disabled, now both channels can run simultaneously (two call of the script are needed)

added timing closure analysis diagram

added reset of MC100EP140 bad state; better configuration; better debugging; comments; attempt for better trigger latency calibration (not working yet)
Signed-off-by: Michael Betz <Michael.Betz@cern.ch>