[gmx-users] how to deal with large pressure fluctuation

[Q733]

Dear gmx-users:

I made a lipid bilayer fully hydrated by water,without water in the
lipid hydrophobic tails.Then I minimized the energy to Fmax=132.5,then I
ran NVT md for 200ps.However when
I wanted to run NPT simulation, the pressure fluctuation was as high as 
1.33*10E+05,it crashed very easily.
Does anyone know how to deal with large pressure fluctuation?I tried to
enlarge tp,however,it was of no use.