----
Richard Shoemaker, Ph.D. Phone--(402) 472-6255
Instrumentation Director, Chemistry FAX---- -6964
Research Associate Professor, Chemistry
University of Nebraska-Lincoln
URL: http://www.chem.unl.edu/nmrlab.html
>>>> Dear AMMRL Members:
Miriam sent this summary over a week ago, and I made a mistake which
prevented it from going to the group. My apologies to Miriam.
Sincerely,
<<<<<<<Rich Shoemaker>>>>>
Dear members,
Thank your for everybody that send me e-mail. Follow you can read
some of answer that I received.
Regards, Miriam
****************************************************************************
*************
We put a Windows-NT based DRX-Avance console on our 500 in July of 1999, so
I have a little over a year with it. This was one of the earlier
PC-platform systems delivered, but certainly not the first. We are,
overall, extremely happy with the performance and (especially) the stability
of the system. The system has never crashed once in over a year. XwinNMR
does crash (briefly) on rare occasions when users push it into doing
abnormal things; however, the X-Windows manager (Exceed) has always
restarted things automatically, without intervention.
The only computer-related downtime we have had was when the video-card (a
"gamer"-quality Voodoo-2000 AGP video card that is way-overkill for
NMR...the card was so over-clocked that it cooked itself). Gateway-2000
replaced the card for free, but it was still flaky. I finally replaced it
with a very cheap XVGA card ($25), and it has been perfect. Cheap computer
repair is one advantage of the PC, as I see it.
We recently had a visiting student from overseas doing some collaboration
with one of our faculty, and he FTP'd several large 2D datasets from his
instrument (SGI-based Bruker). We did all kinds of processing, and there is
no real difference...the data files and software have proven to be 100%
compatible w/ SGI. I don't think there are any tradeoffs in using an
NT-workstation for data processing. Besides, I use Nuts(r) on the NT
workstations for most of my data processing anyway. The same computer can
have both programs available.
I can only think of two real PC-related issues with which we have dealt that
might be considered to be negatives:
1) We can't telnet to "spect", to connect to the CCU in the instrument
console. This is a permissions problem in the NFS-server software
(Hummingbird, which serves the filesystem for the CCU computer).
Instead,
there exist iconized batch programs on the PC that must be used to run
diagnostics, which would normally be executed via "telnetting" to spect on
an SGI. All of the CCU diagnostic programs seem to function fine...they
simply are accessed differently.
2) If you do lots of DOSY-type experiments, not all of the programs (i.e.
"ilt") were ported over/compiled properly for the PC. This is a temporary
problem because they're doing DOSY differently now (and better, in my
opinion) in the upcoming XwinNMR version for the Avance.
This was our first Bruker instrument at Nebraska, so I had no history, nor
any experience with the nuances of their instruments. I must admit that as
we worked through the minor "infancy" problems with the console, I had to
deal with the problem of determining when something was a problem with using
an NT-based PC host, or when it was it a real instrument problem. There
were some software-related instrument problems in the earliest PC versions
of Xwin, but with the current patch level of XwinNMR V 2.6 these seem to be
resolved. Now, with XwinNMR version 3.x coming I expect PC related bugs
won't exceed those for SGI users.
If this sounds like a Bruker sales pitch, I apologize. Adapting to the
Bruker way of doing things has been somewhat of an adventure for me. Our
500 has been running at nearly 90% capacity, 24/7 for the past year, and
having a Windows-NT PC as the host has, thus far, not been a problem.
Of
course, your mileage may vary.
****************************************************************************
*********
I've been using the program for about a year now, and if you are used to
UXNMR (i.e. the SGI version) then you should have no problems. It is,
however, a resource hog (i.e. make sure your PC has plenty of RAM!). Also,
not everything is as straightforward as I would like.
****************************************************************************
**********
we have on O2 for acquisition and 3 pc for post processing we just began using the PC's
our engineer used to work in a place with two O2's and was more satified
with unix stability,
no opbjective bugs for the moment.
We use a NFS shared disk space common to PC's and O2 on a Sun machine for data exchange.
I would stick to the O2, the NT was a customer demand produced by Bill's
marketing policy but bruker's staff prefers Unix in general.
negociate a return to Unix at the end once NT has been used for
bargaining
****************************************************************************
*************
It works well for me here at MIT for offline processing. I do not like it
for the host computer. For instance, if you log out incorrectly you have to
run shmrm to fix the 'hangup.'
Also, CORTABs are not yet supported on the pc version, so if you want to
have 'linear' amplification, then you have no choice but to go with the SGI.
If you are feeling devious, get them to offer you an SGI and 'linear'
power
amplification and then you can demand an SGI. Not that that is entirely
ethical...but some people enjoy playing those games.
Bottom line: if you have the option, stick with the SGI. The code is less
buggy and what you see on the screen corresponds more closely to what is in
the manuals. Nothing is more frustrating than going through the manual and
having them tell you to press a button or a pull-down menu that doesn't
exist.
****************************************************************************
**********
Received on Fri Oct 20 2000 - 13:43:36 MST
This archive was generated by hypermail 2.4.0 : Sat Jun 03 2023 - 17:40:01 MST