AMBER: Problem with modelling ammonium ion?

From: David Cerutti <>
Date: Tue, 4 Sep 2007 18:26:22 -0700 (PDT)


    I modelled ammonium using a tetrahedron and a 1.01A bond length between
N and H atoms; the nitrogen is type N3 and the hydrogen type H. Now, all
this got through tleap with no trouble, and the topology file was created
and seems to have been interpreted properly by PMEMD. However, shortly
after the start of the simulation, I get this message about

"SHAKE failed to converge in 3000 steps"

or something of that nature. Then it bombs.
    Could I have come upon an inherent problem with SHAKE? I know it
has trouble when you get tight rings of constrained bonds, but perhaps
this is also the case when there are so many constrained bonds (4)
arranged so symmetrically about a particular atom.
    Also, are the any tried-and-true parameters out there for ammonium
sulfate? I've been looking, but unable to find these things.


The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Wed Sep 05 2007 - 06:07:42 PDT
Custom Search