Using improved , up–to–date stellar input physics tested against observations of low–mass stars and brown dwarfs we calculate the secular evolution of low–mass donor CVs , including those which form with a brown dwarf donor star .
Our models confirm the mismatch between the calculated minimum period ( P _ { min } \simeq 70 min ) and the observed short–period cut–off ( \simeq 80 min ) in the CV period histogram .
Theoretical period distributions synthesized from our model sequences always show an accumulation of systems at the minimum period , a feature absent in the observed distribution .
We suggest that non–magnetic CVs become unobservable as they are effectively trapped in permanent quiescence before they reach P _ { min } , and that small–number statistics may hide the period spike for magnetic CVs .
We calculate the minimum period for high mass transfer rate sequences and discuss the relevance of these for explaining the location of CV secondaries in the orbital period - spectral type diagram .
We also show that a recently suggested revised mass–radius relation for low–mass main–sequence stars can not explain the CV period gap .