------------
My original posting:
I've never heard this topic discussed here before, but it's on my mind
lately. Lineshape specs on our new 500 MHz equipment are 0.65/8/16 Hz
nonspinning (50%/0.55%/0.11%). The installer got 0.50/4/8 and I recently
got 0.55/5/9.
Someone who either wants to make me miserable by requiring that I spend my
life shimming or who, perhaps, really believes homogeneity is going bad -
finds my result unsatisfactory and insists I achieve the installer's numbers.
I'm inclined to refuse and direct my efforts to more meaningful projects. I
am also inclined to take the position that I am responsible for occasional
demonstrations of 0.65/8/16, but nothing stricter than that.
What say you?
--------------
Rudi Nunlist, Berkeley: On a practical level, the minor difference is (at
least IMHO)irrelevant.
--------------
Steve Silber, TAMU:
Whether you get .55/5/9 or .50/4/8 is not going to make any difference to
how much shimming you friendly user has to do with his real sample. I use
the chloroform/acetone lineshape as a test to verify that the magnet/probe
are 'right' and doing their thing. If you consistently get .55/5/9 with a
modest amount of shimming, and 'believe in your heart' that it can do better
if pressed, then I wouldn't worry about it. I would worry when you have
to sweat bullets to get .6/7/14 one day....
---------------
Jane Strouse, UCLA:
It sounds to me like you did a wonderful job. If the user doesn't
think it is good enough, I would invite them to do better.
---------------
Quincy Teng, Georgia:
The best nonspin lineshape I have got was 0.45/5/11.5 on our
Inova500, which is specified as 0.65/8/16. The installation numbers
were 0.6/7.1/13. 0.55/5/9 lineshape is better than excellent.
---------------
Chuck Anderson, Houston:
If I get less than double the spec's the service rep. obtained I am happy
and let it go at that. I only concern myself with obtaining better shims
if there is a problem. I almost always shim better and faster than the
normal operators of the spectrometers. I think the person is being
unreasonable in asking you to meet the service reps spec's.
---------------
Jerry Hirschinger, Purdue:
You are achieving performance better than the manufacturer was willing to
commit that the instrument was capable of when it was ordered. That should
be tasty gravy! The fact that the installer got better performance is
irrelevant, because that was another guy on another day. The mfr.
published specs ruled when you accepted the instrument. In other words,
the installer could have stopped at .65/8/16 and you would have had to
accept the instrument, whether better was possible or not. That someone
who is giving you a hard time should not be holding you to any more
stringent rules than the mfr. was playing by when they designed, built, and
you accepted the instrument. If they believe that better is possible,
they're right - that's obvious. But, is better necessary? It wasn't
necessary when you placed the order!
-----------------
Mike McGregor, Rhode Island:
I've come up against this type of people problem many times over
the years. My policy is to exert myself to the maximum to satisfy
the customer - you will never convince them that your results are
good enough, you can only show them that you are working hard for them.
Technical explanations - such as showing how close you are to the
target result or explaining how little impact all this shimming
will have on results when a different sample goes in the magnet just
won't work with such a person.
Once you have done your best, you have to call it quits, tell them
this is as good as its going to get.
Don't expect any thanks.
-------------
Jack Martin Miller, Brock U:
If you can routinely meet specs that should be all that is required - and
at many sites that is probably pushing it - the stickler that wants better
should be welcome to come in after hours and tweak to their heart's content
if they need the lineshape. If you called the installer back with a
complaint that you couldn't match the instalation #s he/she would laugh at
you since they only have to meet spec - anything else is a bonus and a
measure of how easily you might reach spec.
---------------
Jeff Simpson, MIT:
Look, you beat factory specs. Even if the magnet is degrading, it is
still within specs, so there is nothing to be done about it anyways.
---------------
Dan Bearden, NOAA:
I seriously doubt
that ANYONE has ever done anything to 'fix' a system where the best shim was
under specs, but not the best ever obtained.
Bill, you can shim my magnet any time!
---------------
Rich Shoemaker, Nebraska:
I think my approach would be to challenge the person in question to
demonstrate a specific experimental ramification of the difference between
your result and the original installer's result.
----------------
Tom Pratum, Washington:
I think that it is absolutely rediculous to think that 5/9 is worse than
4/8; I personally have found that most people can't tell the difference
between 6/11 and 8/15 (0.55%/0.11%). I absolutely would not waste my time
on this, and I know that you know how much time you can waste getting that
last 1Hz out of a lineshape. I think you should be able to get something
close to the installers numbers (which you have done) and I do think that
if you were way off then maybe something is wrong.
-----------------
Peter Rinaldi, Akron:
I agree with you. Except in certain demanding situations, it is unusual that
NMR of typical samples produces noticably improved results if the lineshape
is improved to the extent they seem to demanding of you.
------------------
Steve McKenna, Amoco:
That's where I'd put our mindless cost recovery system to good use: I'd
offer
to do it and bill him for the time it takes at $195/hr. I suppose that
option
isn't available to you though!
-------------------
Rob Kleps, UIC (already sent you this, but it should be in this archive):
I've heard a story about an electric typewriter salesman who wanted to sell
typewriters that were specified to type at least 3 pages per minute (ppm).
The typewriter was purchased by the business office for our office under
the conditions that this typewriter could meet its specifications.
The saleman came in and because of his years of experience was able to type
at 7 ppm using the standard sentence, "The quick brown dog jumped over the
lazy fox." thus exceeding the machines specifications. The office secretary
came in for the first use of the machine and typed several memos that
needed to go out but these memos were produced at an average output of just
5 ppm.
This demonstrates that the machine was either broken or our secretary was
being very difficult because she kept typing office memos rather than
practicing the the standard sentence of, "The quick brown dog jumped over
the lazy fox." until she was able to type it at 7 ppm.
Esaps Fables.
---------------
Greg Leo, PRI:
Your numbers are excellent. Overlay the spectra and you or your accuser
will see that it's fine.
---------------
Karen Ann Smith, New Mexico:
I like to prove that solids people really can shim by occasionally
spending time really tweaking it up- to the stated spec. We all
know that the instrument manufacturers' specs are (relatively) easy
to reach, so the installer beating it is no surprise. Your mistake
was telling anyone that the installer did better than the spec.
Unless/until someone can demonstrate that their sample needs higher
resolution, the facility is obliged to keep the instruments as close
to spec as possible.
Now comes the real question- why does this "person" demand better
homogeneity? Did they degas their sample? Did they filter it? Did
they refrain from using metal in the sample prep (even a metal
spatula can (theoretically) add enough "stuff" to broaden the
line)? Are they using the highest quality tubes and solvent (and
have they taken care of them)? Exactly why does he/she/it think
better resolution is needed? Is there a multiplet that won't
resolve? Unless/until they can convince you that instrument
inhomogeniety is THE limit on linewidth, politely (or not) tell them
off.
----------------
Jeff Walton, UC Davis:
That last little gnat's ass of shim is immaterial to most experiments. Is
there a reason he needs he needs that last bit of shimming? Tell this
person that real men (or women) do there own shimming.
----------------
Partha Sarathy, Auburn:
Bill: The numbers you got look great to me.
----------------
Ted Barnum, Berkeley:
As a former Bruker Field Service Engineer, I can say that it is
probably worthless to try to dulpicate the installation number when
your numbers are as close to them as you state.
----------------
Bill Kearney, Iowa:
I don't think my students and I are particularly inept, but we've never
been able to do better than get close to the installer's figures. I am
curious to see if this is a common experience.
----------------
Pat Wheeler, Isis:
Tell your tormentor to shim to the installer's numbers him/herself, if
he/she thinks it's really that important.
----------------
George Sukenik:
As long as its in spec, no one should complain.
If you paid the the company to come in, they would probably only guarantee
specs, not original results.
----------------
Elwood Brooks, Cincinatti:
I am surprised that the installer took time to beat specs by that much.
----------------
Bob Dudley, govt:
As a spectroscopist one has to decide whether to spend all the time
shimming or obtaining data.
----------------
AND FINALLY, someone who (as you will see, for good reason) wishes to
remain anonymous:
Lineshape is a function of performance for magnet, probe and shims. In
99% of the examples, the probe is the sole determining factor. You, as the
manager of the facility, should take delight in achieving performance that
equals or betters the performance of the installation engineer.
Yes, the magnet will drift a small amount and the cryoshims will change a
little bit. This will require some changes in the room temperature shim
current to achieve best possible lineshape--so what.
Your director is probably interested in achieving best possible water
suppression. Excellent lineshape is the most important criteria for that
experimental need and his request is well warranted. Best possible
lineshape is also an important measure of probe performance for future
reference. If the room temperature shims ever come into question in the
future, or if the probe declines in performance, the lineshape and flip
angles are important parameters to know.
Remember that installers shim probes regularly, but they generally quit
when they have exceeded specs by some amount. They are paid to work against
the clock. They are not paid to achieve best possible probeperformance. You
should be able to achieve 0.3/2.5/5.0 and possibly better.
You have the advantage of gradient shim techniques for the z shims so the
most
tedious aspect of the shim process is straightforward and fast. [Editors
note: how did this person know this?]
If you don't want to do this then assign the job to a good summer student.
Give him some time on the instrument each week and tell him that you want
to set a new record for performance each week. I have personally done this
in the past with excellent results. I am also speaking from 9 years of
experience in a commercial nmr applications lab and 21 years as a customer.
I could go out into the field and beat the pants off of any lineshape
numbers generated by our
installers. The only people who consistently beat me were the probe design
engineers.
If you are interested in the ultimate possible performance for CHCl3,
including resolution of the Cl35, Cl37 resonances due to varying isotopic
distribution then check with Frank Anet at UCLA Chem Dept. There are also
limitations to best possible performance due to radiation damping from a 1%
sample of CHCl3 on a 500 (consult LeRoy F Johnson in Cupertino CA, retired [
lfjohnson@prodigy.net ] ).
The amount of time that you have invested in soliciting opinions is
roughly equal to the amount of time necessary to achieve better lineshape. In
the words of the Nike commercials..."just do it". Please treat this as an
informal discussion between the two of us. I really don't care to see this
repeated in the ammrl.
---------------
end of summary. Thanks again to all.
Bill
William C. Stevens, Ph.D. Nuclear Magnetic Resonance Facility
Director Southern Illinois University
Carbondale, IL 62901-4405
618-453-6498 voice
618-453-6408 fax wstevens@siu.edu
http://opie.nmr.siu.edu