multiple auxiliary basis sets - $trunc option

[klickklack]

Hello,

I tried the $trunc option in order to invoke a multiple auxiliary basis set calculation in the ridft module by just adding $trunc to the control file.

Somewhat surprisingly, my test calculation without the $trunc option was faster than the one with the $trunc option. 
What did I do wrong?

Is there more documentation for the $trunc option than the short note in the manual that it can be invoked?

Thanks for your help.

[Arnim]

Hello,

if you switch on trunc in the last define menu, a data group
$truncated RI
is added to control and the $atoms data group is modified with litmax, e.g. like this
$atoms
o  1                                                                           \
   basis =o def-SV(P)                                                          \
   jbas  =o def-SV(P)                                                          \
   litmax=*
h  2-3                                                                         \
   basis =h def-SV(P)                                                          \
   jbas  =h def-SV(P)                                                          \
   litmax=*
For litmax you would have to insert the highest angular momentum from the auxiliar basis set, that should be used during the SCF cycles, e.g.
$atoms
o  1                                                                           \
   basis =o def-SV(P)                                                          \
   jbas  =o def-SV(P)                                                          \
   litmax=2
h  2-3                                                                         \
   basis =h def-SV(P)                                                          \
   jbas  =h def-SV(P)                                                          \
   litmax=1

With this it should run, but I am not sure, if this is really a useful.

Cheers,
Arnim

[christof.haettig]

Can you report your problem (together with the input and output files) to your TURBOMOLE seller and ask them to forward it to the developer responsible for the $trunc option?

Regards,
Christof