I would =
like to
thank Andy Soper, Alex Kurochkin,
Kelly Moran, Dave Scott, Jerry Hirschinger, =
Joe
Yeager (Oxford Magnets), and Paul =
Cope
(Bruker) for responding to my posting. It seems 1 Gauss line is a good =
guideline,
although I feel that I still do not have a good grasp on the effect of =
the
ramping speed. I apologize =
if I
missed anyone.
The =
original
posting was:
>
We are asked to participate in the planning of a new science building =
>
and to consider the proper distances between =
NMR, EM,
and cryophysics =
>
facilities. All of us want to be on the =
ground level
and the space is
>
limited.
We learned that our cryophysics =
colleagues
ramp their
>
bipolar superconducting magnets up and =
down.
>
What specs of their stray field should we consider to ensure our =
>
operation will not be affected by =
them?
Depending
on whether your =
NMR
spectrometers are made =
by Bruker, Varian, Jeol =
or some
other vendor you should probably contact them for a definitive =
answer. The answer for each instrument =
will
probably be different.
Modern
magnets employ active shielding and the external fields are quite =
small. Older machines are not shielded =
and
their stray field extends much further from the magnet. =
Generally
the 5 Gauss line is used;
occasionally the 1 =
Gauss line is
used. 1 Gauss is about =
60-70% of
the strength of the earth's magnetic field =
strength.
So
possibly you are looking for the unshielded magnet which has the highest
fundamental field strength.
If all are shielded =
the one
with the highest field is probably the greatest threat to your =
instruments.
For
constant stray fields, it is commonly accepted that the 5 Gauss lines of =
the
magnets should not cross - you can always shim it out. =
In
your case, you want to make sure that the range of the bipolar magnet =
stray
field changes does not exceed your desired resolution which in turn =
depends on
the field strength of your magnets.
Also take into account the slope of the field ramp: the steeper =
the
ramp, the more it may affect the desired resolution of your =
system.
For example, if the =
bipolar
and constant field magnets touch their 1 Gauss lines, then the effect =
will be
the following. 1 G for say =
400 MHz
machine is equivalent to ~4500 Hz in protons. Upon reduction by a factor =
of
9*10**4 (which is the ratio of the field in the sample to 1G), it would =
result
in the influence of ~0.05 Hz. =
You
decide if this stability is really =
necessary.
At
the Univeristy of Warwick Physics Dept, =
ramped
bipolar supercons and high field NMR supercons lived in harmony on the first floor. =
They
were separated by space. Find out what the stray fields are on =
every
magnet and position them accordingly. We even had an unshielded =
600 wide
bore NMR magnet, which had a very large stray field.
The biggest problem we had was with the cleaning staff. They knew =
enough
to ask if the magnets were on or off before they entered the rooms with =
vacuum
cleaners. They asked an NMR user if the 600WB was off and the =
answer was
"yes" for some reason, and the vacuum cleaner quickly became =
engaged
with the magnet. Although the field had to be lowered to remove =
the
appliance, and the motor on the vacuum was fantastically destroyed, no =
one was
injured and the magnet was brought back to field =
successfully.
NMR magnet manufacturers can tell you the stray fields of their magnet =
and the
other magnet manufacturers should be able to as well. It's a =
safety
issue, after all.
--
I no longer work =
there.
For us the unshielded 600WB NMR magnet was the biggest problem. =
We
clad the walls and ceiling with stainless steel to contain the field, =
but it
still got through the windows.
All
of the magnets were in different rooms, some closer than others =
depending on
their stray fields. You really have to find out what the stray =
fields
are-- the magnet manufacturers can tell you, or you can measure it =
yourself
with a Gauss meter. Comparing it to someone else's magnets won't =
help
because they will be different.
We
once had an epr magnet at the 5 gauss line =
of a
DRX-400.
Sweeping
the epr from 40 to 4540 gauss would move the =
unlocked
nmr spectra about =
.9Hz. It didn't have any effect on 2H =
locked
spectra.
Our
most recent Bruker installation guide =
suggests that
field changes between 0-5mG regardless of gradient are generally =
considered
harmless for standard NMR work.
If
the change is larger than 5mG, the lock system will compensate if the =
gradient
is less than 5mG/sec.
I
would try to avoid the stray fields crossing at any value greater than 5 =
gauss
for a static stray field.
Especially, I would be sure none of the magnets are positioned =
with
their bores near a common axis. =
The
dynamic stray field you describe would best be positioned radially
from the NMRs, so major effects are not in =
the axial
gradients.
I am attaching a typical stray =
field plot
for a 14Tesla magnet. You could extropolate
from 5 G line to estimate 1 G point. I am also asking our =
engineers for
estimate of 1 G line.
In general the siting
of a NMR magnet near a 13T physical science magnet is no different than =
siting next to another NMR magnet. In =
this case
one of the magnets will be changin field
often. The speed of ramping is not the issue, only =
that it is =
changing.
So in the case of two NMR magnets =
(or any
two magnets), where one is changing fields, you should positon
the static system so that it is well outside the 1G line of the =
other.
When you are at this level, then =
variation of the other magnet are down in the noise =
of other
items (subway trains, elevators...)
Couple of other notes.
If your NMR magnet is an actively sheidled =
design,
then this has added advantage not only minimzing stary fields, but also shields your magnet from =
outside
fields.
Generally speaking, you want to try =
to
determine the magnetic field perturbation at the location of your NMR
magnet. This will be the fringe field at the highest field =
strength
generated by the magnetic field ramp. Lower fields during the ramp =
will,
necessarily, have a lesser effect on your NMR magnet. Then, you need to
evaluate the inherent sheilding factor =
provided by
the NMR magnet design. This can be provided by the manufacturer of the =
magnet.
Of course, the Bruker magnets have =
proprietary
designs that minimize the effect of magnetic field perturbations, with =
the UltraShield magnets providing =
one-order-of-magnitude
reduction in external field disturbance and the UltraShield
Plus magnets providing two-orders-of-magnitude reduction. Are the =
physics
magnets shielded in any way?
Examples of the most challenging =
sites in
which customers have taken advantage of this are Northeastern U, where =
magnetic
field disturbances from a nearby (~75 feet) electric train track were as =
high
as 87mG, which corresponds to approximately 370 Hz shift in the magnetic =
field.
Of course, in high resolution, there is also a digital lock circuit that =
helps
the NMR chase any field shift. But at Northeastern U, like you, they =
were intersted in solid state NMR, which generally =
doesn't use a
lock (although an external lock is available). With our UltraShield
Plus magnets, the shift in magnetic field inside the magnets measured in =
the
presence of 65mG field (275Hz shift) was <2Hz, pretty good for =
standard
solid state NMR and better than the two-orders-of-magnitude promised by =
the
technology. You can expect that we will highlight this capability when =
you
arrive at the evaluation period.
From the physics lab standpoint, =
the NMR
magnet is a static field, so again, determine the fringe field at the =
location
of the physics magnet and determine if this fringe field will create a
distortion to their experiment. Since the fringe field of an 800MHz =
UltraShield Plus drops off pretty rapidly with =
distance, I
would guess that 20-25 feet will probably be more than sufficent
separation to convince them that you aren't an issue to their =
experiments.
EMs will be the most sensitive instrument to both static and =
dynamic
external magnetic fields. Thus, they may need the greatest distance from =
the
NMR and physics magnets.
The effect on EM is less clear.
Hsin
--
Hsin Wang, Ph.D.
NMR Facility Manager
Department of Chemistry
CUNY - City =
College
of New =
York
hsin.wang@sci.ccny.cuny.edu
212-650-8314 (Lab)
------=_NextPart_000_002F_01C86E6A.35A2F580--