TIP4P coul/cut long time to equilibrate

I am using LAMMPS-30Jul16 model to simulate a water nanodroplet of size 5nm. I am using TIP4P model. I started with a water box and used lj/cut/tip4p/long first. It gave me a nice spherical water droplet quickly. But, when I tried the same with lj/cut/tip4p/cut, thinking that my droplet is of finite size, it doesn’t appear to go spherical anytime soon. I used the parameters specified in http://lammps.sandia.gov/doc/Section_howto.html. I remember observing similar behaviour last year with spc/e model too.
Am I missing any important “fix” or anything here? Thanks a lot in advance.