Unreasonable pop_frequency when considering NAC

AMSET
1

Dear Expert,
I’m currently attempting to calculate the pop_frequency of BAs using NAC, but the results are unreasonable. The issue lies in the improper allocation of weights, as detailed below:
No NAC corrected file provided. Skipping calculation requiring NAC

Frequency Weight

  -0.00      0.00
  -0.00      0.00
  -0.00      0.00
  20.37      0.33
  20.37      0.33
  20.37      0.33

pop_frequency: 20.37 THz
and for Found NAC corrected phonon frequencies

Frequency Weight

   0.00      0.25
   0.00      0.34
   0.00      0.41
  20.37      0.00
  20.37      0.00
  20.46      0.00

pop_frequency: 0.01 THz
My POSCAR is
c-BAs
1.0
0.0000000000000000 2.4084714529927198 2.4084714529927198
2.4084714529927198 0.0000000000000000 2.4084714529927198
2.4084714529927198 2.4084714529927198 0.0000000000000000
B As
1 1
Direct
0.7500000000000000 0.7500000000000000 0.7500000000000000
0.0000000000000000 0.0000000000000000 0.0000000000000000
and the INCAR is
ISTART = 0
ISMEAR = 0
SIGMA = 0.05
POTIM = 0.01
ADDGRID = True
EDIFF = 1E-8
PREC = Accurate
NSW = 1
ISIF = 3
NELMIN = 5
ENCUT = 550
LWAVE = .FALSE.
LCHARG = .FALSE.
LREAL = .FALSE.
KPAR = 4
IBRION = 6
LPHON_DISPERSION = .TRUE.
PHON_NWRITE = -3
LPHON_POLAR = .TRUE.
PHON_DIELECTRIC = 9.841758 -0.000000 0.000000 0.000000 9.841758 -0.000000 0.000000 0.000000 9.841758
PHON_BORN_CHARGES = -0.45953 0.00000 -0.00000 0.00000 -0.45953 -0.00000 -0.00000 0.00000 -0.45953 0.45953 -0.00000 0.00000 -0.00000 0.45953 0.00000 0.00000 -0.00000 0.45953

I completely understand that your schedule must be busy, but I would truly like to know how to solve this problem.
Looking forward to your response.

Best regards,
Shicong Ding

2

Dear Schicong Ding,

Can you share the NAC corrected frequencies written in the OUTCAR file? First three frequencies should be zero. If there are positive frequencies, this issue might arise.

Regards,
Shiva