Below is a summary of responses to my question about the Z1 RT shim going
off scale. Thanks to everyone who responded. Iım dividing this into two
categories: suggestions for tests, and reports/explanations of cryoshim
problems.
Suggestions for tests:
(Note: Iım thinking it is best to proceed in the order shown here.)
1. Make sure that the gradient amplifier is not causing the problem. They
have been known to put out current even when they are not pulsing, and this
can make the Z1 setting change dramatically. On a Varian system, compare
pfgon=ınnyı and pfgon=ınnnı, or better yet, turn off the PFG amp and unplug
it from the probe to see if the problem goes away.
2. Make sure that the RT shim stack has not slid out of position in the
magnet.
3. Make sure (on a Varian system) that the upper barrel is seated properly
so that the probe is positioned properly within the shim stack. Other than
just physically checking this, the Z1 trace on a gradient shim map should go
through zero in the center.
4. Locate the doped water spectrum made by the installer when the cryoshims
were entered, which is taken with all RT shims set to zero. Retake this
spectrum and compare to the original; if the new spectrumıs shape and width
is bad, a problem with the cryoshims is indicated.
5. If test 4 is passed, there should be no problem with the cryoshims; in
that case, consider the possibility that the RT shim stack has developed a
partial short between the windings. Maybe test by measuring connectivity
between the different shim coils with a resistance meter.
Reports/explanations of cryoshim problems:
Since I flunked test 4 above, Iım including myself in this group. There
were 15 reports of having a cryoshim either quench or drift. Of these 6
called this a quench of the shim, some of whom said it happened rapidly, and
9 reported drift. Suggestions for what makes it happen are exposure to He
gas during a He fill, an accident with a ferromagnetic object, ice in the
magnet, a bad magnet, or low helium level causing a switch in the magnet to
become resistive and leak current. Some of these may explain a quench
better than drift. But, it appears that most of us who have actually
suffered with the problem have no clue as to why it happened to us. Those
reporting quenches said they had their cryoshims reenergized without
apparent further problems.
Drift seems to be a peskier problem. In three cases, the drift was reported
as beginning in the year following magnet energization or reenergization.
Of these, 2 saw significantly lower drift after adjustment of the cryoshims
(I am the 3rd case, with no outcome yet). Two of the cases of drift
mentioned that the cryoshim drift was associated with magnet drift. In two
cases, the drift required adjustments of the cryoshims every year or 16
months ouch! Cryoshim cycling may reduce the drift rate, as indicated by
this report from Charlie Fry:
³After a rebuild and re-energization {of an Oxford 500 } nearly ten years
ago, the Z1 cryoshim was drifting very fast (as observed from changes in
normal shimming over the next six months). We were fortunate that Varian
agreed to have Oxford come back in and "cycle" the cryoshim to calm/slow
the drift down. There is a known procedure for this (cycling the amps
above and below the final setting a few times), and in our case it did
work. The drift went from a many thousand units per month to a couple
hundred per month.²
Charlie says that only the Z1 shim was cycled, and this magnet was one of
the first 500ıs built, which may mean it is unique. Still, I have been
hoping that cycling might help in my case, since we cannot afford to adjust
a cryoshim every year. I have tried to become more informed about the
possibility of reducing drift with cycling by talking to magnet service
companies. I hope I donıt get into trouble by naming names here; certainly
everyone should take the opportunity talk to these companies about their own
situation and see whether this is the story. Oxford has a procedure for
cycling all the cryoshims, and this is done routinely at fields above 600
MHz, but only sometimes at 600 MHz. Apparently, if the magnet quenches it
is less likely to have drifting cryoshims, so cryoshim cycling would not be
done on a 600 that quenched when it was energized. Praxair and Varian only
cycle Z0, but their attitudes toward my situation differ: Praxair thinks
cycling Z0 might help reduce my drift, while Varianıs stance is that the
magnet is bad and no cycling will reduce the drift.
As an interim solution before cryoshimming, some AMMRLers suggested doubling
the possible amount of current that can be sent from the RT shim driver for
the drifting shim by altering the shim driver circuit. There might be a
worry about heating up the shim stack too much but it certainly is a less
expensive temporary solution. On the Inova, the resistor to change for Z1
range extension is R30 in the Z1 section of the Z0/Z1 shim driver board.
Varian gave me a 4.7 Ohm resistor to replace the normal 10 Ohm one. This is
a different sort of resistor, which is square and black.
I thought it was time to get this summary out, rather than wait for the
outcome with our magnet, since Iım not sure when it will be worked on. I
may post an addendum later on what happens with our magnet.
Received on Thu Oct 09 2008 - 08:34:16 MST