[Pw_forum] Relax error

[Miguel Martínez Canales]

Good afternoon everybody,

I was testing some pseudopotentials, and one of these tests consisted in 
the H-H distance in equilibrium. So I did a relax run with pw.x. It worked 
OK. I copied the final positions into the input and then re-ran the test 
(with higher convergence criteria). However, this time I got an error (also 
appears with the same convergence criteria from the first run). The last 
lines of the output are:



      entering subroutine stress ...

           total   stress  (ryd/bohr**3)                  (kbar)     P=   -0.05
   -0.00000031   0.00000002   0.00000002         -0.05      0.00      0.00
    0.00000002  -0.00000031   0.00000002          0.00     -0.05      0.00
    0.00000002   0.00000002  -0.00000031          0.00      0.00     -0.05

      searching for next position (pslinmin)...
      Eold  =     -2.33845221
      Etot  =     -2.33845221
      DEold =     -0.00000101
      DEtot =     -0.00000158
      linmin: 2nd order interpolation plus 3rd order corrections

 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
      from linmin : error #         2
      unexpected error
 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

      stopping ...

I would be glad if anyone more knowledgeable than me suggested anything. 
 From calculation=md to running CPMD (though it is an USPP...) to dancing 
and singing round the computer while it is running. Thanks in advance,

Miguel

PS: Am I asking too much accuracy? Or is the idea of a second relaxation 
run plain stupid?


-- 
----------------------------------------
Miguel Martínez Canales
    Dto. Física de la Materia Condensada
    UPV/EHU
    Facultad de Ciencia y Tecnología
    Apdo. 644
    48080 Bilbao (Spain)
Fax:  +34 94 601 3500
Tlf:  +34 94 601 5437
----------------------------------------

2005 International Year of Physics

"The only things you can't avoid in life
are death, taxes and Ubuntu reviews."

Robert Storey