NPT dyn for NVT Nose-hoover thermostat fixed volume

Dear ASE Community,

I am currently using the NPT method with a pressure setting of 0 to construct an NVT Nose-Hoover ensemble for my simulations, using MACE calculator. I have a few questions and would appreciate your insights.

My system is a super cell of NaZr₂(PO₄)₃ crystal under PBC condition, which is an orthorhombic crystal structure up to 400 K. To avoid NPT dynamic gets structure error, I am using Atom.set_cell command by removing so small off diagonal cell elements (10^-10A). I have tested the relaxation time constant (tau) and found it to work effectively around 20 fs, with a time step set to 1-2 fs.
My main concerns are:

  1. The temperature seems to converge around the desired value of 400 K, but I observe fluctuations of about ±20 K, typically around ±10 K, however, approximately every 70-100 fs, there is a sharp jump in temperature to 420 K or down to 380 K. Is this level of fluctuation considered high?
  2. The volume of the system remains completely fixed and shows no fluctuations at all. Is this expected behavior under the conditions I am using?

I appreciate any insights or advice you could provide regarding these observations.

Thank you for your assistance!