Coarse-grained molecular dynamics for Martini Forcefield

Hi All
I am trying to run CG by using Martini Forcefield for Octane system.
I did the following:
1. create the 2 beads connected with each other and  set the Mass to 72 and the Radius to 2.35 A for each bead.
2. apply the forcefield types to be C1 from Martini forcefield  for each bead since Martini  mapped the butane as C1.
3 . create a box with 5000 beads and set the density at 0.703 g/cm3.
5. the box was optimized by using Geometry Optimization.
6. the box was run under dynamics task using NPT at 298 K and 1 atm. 
the problem is the density change  to reach to 1.098   g/cm^3. This is not agree with Martini  paper (0.77g/cm^3).
I do not know what is the problem
I used Andersen and Berendsen as Barostat. Both give same result as seen on the Pics.
S.J. Marrink, A.H. de Vries, A.E. Mark, Coarse Grained model for semiquantitative lipid simulations, J. Phys. Chem. B108, 750-760 (2004).