CG-CD1 TRP bond oscillational less than 0.004 fs

GROMACS version: 2023.2-plumed_2.10.0_dev
GROMACS modification: Yes with plumed
Here post your question

I am getting a bond oscillation Note - see below. I am using the Amber19SB with OPC forcefields. I used amb2gro_top_gro.py to convert the amber parm and inpcrd into gro files. I am also using hydrogen-mass-repartioning via parmed so that I can use dt = 0.004 fs. I suspect this bond oscillation is the cause of my system frequently blowing-up.

My question is why is this occurring within a TRP residue?

38 C* 3 TRP CG 38 -0.14150000 12.010000 ; qtot 0.862300
39 CW 3 TRP CD1 39 -0.16380000 9.994000 ; qtot 0.698500

NOTE 1 [file topol.top, line 3771]:
The bond in molecule-type system1 between atoms 38 CG and 39 CD1 has an
estimated oscillational period of 2.2e-02 ps, which is less than 10 times
the time step of 4.0e-03 ps.
Maybe you forgot to change the constraints mdp option.

My mdp file:

integrator = md
dt = 0.004
nsteps = 250000
nstxtcout = 250
nstvout = 250
nstfout = 250
nstcalcenergy = 100
nstenergy = 100
nstlog = 1000
;
cutoff-scheme = Verlet
nstlist = 20
vdwtype = Cut-off
vdw-modifier = Force-switch
rvdw_switch = 1.0
rvdw = 1.2
rlist = 1.2
rcoulomb = 1.2
coulombtype = PME
;
tcoupl = v-rescale
tc_grps = SOLU SOLV
tau_t = 1.0 1.0
ref_t = 310.15 310.15
;
pcoupl = C-rescale
pcoupltype = isotropic
tau_p = 5.0
compressibility = 4.5e-5
ref_p = 1.0
;
constraints = h-bonds
constraint_algorithm = LINCS
continuation = yes
;mass-repartitioning-factor = 3

;
nstcomm = 100
comm_mode = linear
comm_grps = SOLU SOLV

This paper suggests to use a factor of 2 for HMR for TRP residues. 8 ns into the 100 ns sim and so far so good.

https://pubs.acs.org/doi/10.1021/acs.jctc.1c00185

I am having a similar Note with CG-OD1 for aspartic acid, which is a double bond. Not sure what the issue is.