[gmx-users] steepest descent problem

[hugo verli]

I'm trying to perform a steepest descent minimization in a solvated protein. I
received the following message:

STEEPEST DESCENTS
   Tolerance         =         1500
Step =     0, Dmax = 1.00e-05 nm, Epot = -1.95525e+05 Fmax = 2.00254e+03
Step =     2, Dmax = 5.00e-06 nm, Epot = -1.95526e+05 Fmax = 2.11981e+03
Step =     5, Dmax = 1.50e-06 nm, Epot = -1.95523e+05 Fmax = 3.15159e+03
Stepsize too small Converged to machine precision,
but not to the requested precision (1500)
You might need to increase your constraint accuracy

writing lowest energy coordinates.
Maximum force:  2.11981e+03

STEEPEST DESCENTS did not converge in 6 steps
  Potential Energy  = -1.95526e+05

This message appear in several conditions, using and not using all heavy atoms
fixed, hbonds constraint. I'm using a emstep of 0.00001 nm.

What it is? There is any way to solv it?

Thanks in advance for any help,

Hugo Verli.