possible bug in Ewald routine ?

[vish]

Dear CP2K users,

EWALD section requires one to input ALPHA (ang -1) parameter. In general, I 
would expect changing this value should not have any affect on the total 
potential energy of the system. However, when I change ALPHA in the tests 
files in CP2K tests files, the total potential energy changes a lot (for 
example water512.inp in tests/Fist folder). I was wondering if there is a 
bug in ewald routine or I am understanding this wrongly?

Any help on this is greatly appreciated. 

Thanks in advance,
Vishal
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20130325/f516858c/attachment.htm>