Several of those who responded to my earlier notes regarding the Varian
Gemini-2000 requested that I post a summary.
To be fair to Varian, let me first mention that nearly ALL of our problems
with the "turn-key" automation macros can be traced back to ONE "undocumented
feature"...the auto-lock and/or auto-shim routines would try to set the
lock gain to "40", and 30 is the highest number allowed by the Gemini
hardware. Since 40 is an invalid number, the lock would go "flatline",
and the automation macro would abort with an "autolock failure".
The new software HAS fixed this, and I've spent the last two days working
with the Varian service rep., and we've run EVERY automated experiment
on a WIDE VARIETY of samples. The success rate=100%! Unless something
drastic happens, our students will be amazed, excited, and impressed when
they return from spring break the week-after-next!
----------------------------------------------------------------------
As far as other users' experience goes, here's a summary (most people asked
not to be quoted, so I'll paraphrase as best I can):
The general consensus is that the Gemini-2000 is/will be a top-notch,
reliable, turn-key instrument (as it is designed and marketed to be)...
I agree with this assessment. Most of the respondents did, however,
indicate that there was much user-intervention involved in getting the
instrument to reach that point. Getting all of the parameters optimized,
and put in the right places for the automation macros to work, seems to have
been left to the users in most cases. Since most of the installations have
been in facilities with someone like you and I, this hasn't been a HUGE
problem. I our case, the person in charge of the instrument could not be `
expected to perform these tasks, and I couldn't put in the time.
Many of the problems could have been averted at installation time.
---------------------------------
Once the systems are properly setup, and NOT INCLUDING fatal software
oversights in the "beta" (pre 5.1) version, the general feeling seems to
be the throughput is high, training is EASY and PAINLESS for undergraduates,
---------------------------------
More than one respondent suggested the following, which I wholeheartedly
agree with: At installation time, MAKE sure that the instrument not only
performs to the the standard specs, but ensure that the installation
engineer has properly setup ALL parameters required by the automation
software, so that ALL pushbutton macros work smoothly for a variety of
samples. I recommend making up some realistic samples in a variety of
solvents, and radomly inserting them and "clicking" on the "Glide" buttons
to make sure that everything operates smoothly.
In our case, many of the "calibrations" were done at installation, but the
values weren't entered into the appropriate parameter sets which are used
by the automation. There is an "autocalibrate" routine which can be run
under the "Glide" software, but even in 5.1 the engineer had to fiddle around
a little to get everything as it should be.
---------------------------------
For many people, the bottom-line was that we all knew there were some
problems which would be fixed in 5.1...THEY HAVE BEEN FIXED! I will
quote one user, because I think that it reflects how many of us feel about
dealing with Varian (I won't give credit to the "quotee"):
.......taken out of context.......
"One thing I would add is that Varian has essentially fixed ALL valid
problems with all my instruments (not counting things that could only be
fixed by a new generation. Unfortunately, it's like medicine. Users have
to be involved enough to know there is a problem (and, unlike medicine,
probably a fix). I guess I didn't worry about the Gemini-2000 performance
at all at this point, because my comfort level with Varian ultimately
delivering the goods is quite high."
No argument here.
Richard Shoemaker, Ph.D. Phone--(402) 472-6255
Instrumentation Director, Chemistry FAX---- -9402
Research Associate Proffessor, Chemistry
University of Nebraska-Lincoln