Topspin and 32 bit libraries in a 64 bit system

On and off we have touched upon this topic, either in the context of making Topspin play well with Fedora or RHEL5. In particular, this problem is related to the inability of Topspin not able to source the 64 bit libraries.   This is a small snippet in that list. Hopefully, I will summarize the required libraries and rpm-s in a comprehensive post later.

System :  RHEL 6.3 on a Intel 64 bit machine (Dell Optiplex 390)

After the install, I needed to install the following 32 bit rpm-s that essentially provide X11 related shared object libraries.  Please remember that, in a 64 bit system like the above, the 64 bit versions of these shared libraries are already installed.  Since Topspin takes a minimalist approach, it only looks for files under /usr/lib instead of /usr/lib64, to give an example.

The following 32 bit rpm-s are needed even for TS install to work properly, via the graphical interface. 

glibc.i686             : 
glibc-devel.i686    : 
libX11.i686 (this gets pulled in as a dependency if you install libX11-devel)
libX11-devel.i686  : 

After install, once your flexlm license manager is started

correctly, the TS startup halts at the point where the

welcome screen is supposed to come up.

One or more of the shared library files will be reported

to be missing in the console window, at this stage. The rpm-s shown above are the ones that provide those missing shared object libraries.

The following are needed for the TS gui to start correctly.


libXext-devel.i686 

libXtst.i686

 After the above rpm-s are installed,  the GUI comes up without any issues.

High T served..

How high can I go with my VT setup ?  This is a nagging but relevant question that needs a definitive answer for many users.  I ran  the following test to figure this out.

Here is some useful info first, as to what we already know.

• The recommended temperature extremes that the room temperature bore can tolerate is between 0 C to 50 C.
• The shim stack itself can apparently handle +80C (this needs clear confirmation from Bruker and I am working on it).
• There is no active thermal shielding setup between the probe head sample space and what surrounds it.  In non-gradient model probes, there used to be a vacuum dewar that provided effective insulation. 
• The gradient coil takes up the space occupied by the vacuum jacket and this means that the heat will transfer from the coil space to the probe sleeve and then onto the RT shim stack and eventually to the RT bore itself (heat will flow in the opposite direction for low VT experiments)
• Bruker provides an active cooling mechanism by providing either a barbed connector or a compression coupling at the top of the BST (Bruker Sample Transport) assembly.  We are asked to provide an air flow of 200 to 300 L/hr through this to cool the shim stack. We don't need much pressure since this is supposed to be a free flow, where the cooling air exits out of the bottom of the shim stack.
• When we jacked up the flow beyond 600 L/hr., the sample is getting lifted from its seated position and so that should give us an upper cutoff.

We employed the temperature and humidity logger from a company called 'Onset'.  They have a windows based $$ware called HOBOware that controls the logger, transfers data to the PC and plots it.