Date: Tue, 9 Jul 2002 08:31:59 -0700
From: "David A. Case"
Subject: Re: eedmeth variable in AMBER7

On Tue, Jul 09, 2002, Sarangan Ravichandran wrote:

>
> I have been using sander_classic (AMBER6)
> to do simulated-annealing simulations
> on a non-periodic system in vacuum with
> distance-dependent dielectric(IDIEL=0).
> I am trying to move my calculations to AMBER7.
> If I understand correctly in sander,AMBER7, I should
> use eedmeth (=5) to achieve the same thing.
> But, the eedmeth variable should only be set inside
> the &EWALD namelist. if I do not want to
> use Ewald for my long-range calculations, then
> what option(s) should I choose for my simulations?
>

If you set ntb=0 and igb=0 you will get a non-ewald simulation. If you
then set eedmeth=5 in the &ewald namelist, you get an r-dielectric, but
it is still not Ewald.

I know it is counterintuitive to put eedmeth into the &ewald namelist
section -- that's historical. In the future, we'll add a variable with
a better name to the &cntrl namelist to do the same thing, but for now,
you need to set eedmeth in the way you specify.

..hope this helps....dac