conservation of Cons Qty in NVT ensemble

[Anna Pavlova]

Dear cp2k users,

I have a problem with a large drift in the Cons Qty when i turn on
Nose-Hoover thermostat. I have equilibrated the system for 6 ps with
velocity scaling, prior to using the thermostat, and I didnt see any
energy drift then. I have been using the default settings for Nose-
Hoover thermostat. If anyone has an idea on what could be going wrong,
I am grateful for suggestions. The relevant input part is included
below.

Regards,
Anna


&MOTION
  &MD
    ENSEMBLE NVT
    STEPS 3000
    TIMESTEP 1
    TEMPERATURE 350.0
    &THERMOSTAT
     TYPE NOSE
     &NOSE
     LENGTH 3
     MTS 2
     YOSHIDA 3
     &END NOSE
    &END THERMOSTAT
  &END MD