Actually, the more I think about this problem, the more I think: 'oscilator schmoscilator'. My previus monostable solution 'worked' as the width of the pulse by definition is related to the capacitance as modified by the arial, the only trouble with it was that the variaion in capacitance was tiny a...