VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Fri Oct 26 2012 - 08:22:00 CDT

On Fri, Oct 26, 2012 at 1:01 PM, George Patargias <gpat_at_bioacademy.gr> wrote:
> Hello
>
> Apologies if this is too trivial. I am trying to figure out how to
> connect/use the Novint Falcon device for IMD with VMD/NAMD.
>
> As far as I understand, the VRPN-ICMS software (from Dr Kohlmeyer's page)
> provides an interface between Falcon and VMD; is this correct? If so it

yes. the regular VRPN distribution should work as well,
since the code has been folded into VRPN quite a while back.

> needs to to be installed in the the machine where the IMD will run?

no. please keep in mind that you have two connections (and
both use TCP/IP sockets). the IMD connection between the
ongoing simulation and VMD over which VMD receives updated
coordinate information and tells the simulation about adding
forces to atoms that are communicated via IMD.

and then you have the connection from VMD to the haptic device
via VRPN which tells VMD the position of the effector or "3d mouse"
(which is used to compute the forces to be send to the simulation
via the IMD connection) and through which VMD tells VRPN what
force feedback to generate on the haptic device.

i know about a 4-way setup that was run at a supecomputing
conference:
1: supercomputer to run the MD simulation on
2: remote visualization facility to run VMD on
3: laptop machine to run VRPN on
4: multiple desktops to run the viewer for the remote viz on.

it is also possible to connect multiple haptic devices
via separate VPRN instances, eg. on multiple laptops
and then have a tug-o-war between them. theoretically,
the falcon support in VRPN would handle multiple falcons
from the same instance, but there are some nasty issues
with initializing the device, that can be avoided by using
a separate host for each falcon.

> I am asking because I have seen pictures in which haptic devices are
> connected/controlled by a separate machine (e.g. a laptop).

yes. that is how it works. with the way how the IMD
support is implemented in LAMMPS, you can also
easily have multiple IMD connections to the same simulation.

> Apart from VMD, NAMD and VRPN-ICMS no other additional package is needed
> to run an IMD?

you don't necessarily need NAMD as MD code. several other MD codes
(LAMMPS, HOOMD, Gromacs) also support the IMD protocol.

hope that helps,
    axel.

> Thanks in advance.
>
> George
>
>
>
> Dr. George Patargias
> Postdoctoral Researcher
> Biomedical Research Foundation
> Academy of Athens
> 4, Soranou Ephessiou
> 115 27
> Athens
> Greece
>
> Office: +302106597568
> 

-- 
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com  http://goo.gl/1wk0
International Centre for Theoretical Physics, Trieste. Italy.