Gradient Shimming

Note: if the shims are really messed up, you may have to load in shims from the last lineshape test. To do this, select menu File->Open and navigate to
and open the last dated lineshape fid. Then immediately type in “load=‘y’ su”

Gradient Shimming Procedure

To start gradient shimming, use VnmrJ tabs/pages to select Setup->Lock->SetupGradientShimming (or type in ‘gmapsys’). Choose a shimmap from the shimmap listbox. Use H_map for water and H2_map for samples in D2O. Select TrialAcquisision button; you should then see 2 camel humpback profiles with the 2nd profile between 50% and 80% of the first’s magnitude. If it looks good then press DoGradientShimming button.

Note: if you are using a Shigemi tube, make sure you are only using z1-z4 gradients. Do not try to use gradient shimming on z5 in a Shagehmi tube. It will drive z4 or z5 to its end value and you will not be able to get good shims.

How to Tell if You Have Good Shims

For protein samples in H2O (10% D2O), you need to pass the “below 100” test if you expect to get good spectra for assignment purposes. Results of the “below 100” test depend on both sample concentration and shims.
  • collect a 1D proton water presat scan (like when finding tof)
  • phase spectrum and adjust so amplitude of aliphatic peaks are at about 75% full scale
  • put vertical cursor on top of strongest aliphatic peaks
  • place horizontal on either side of water peak and expand
  • put horizontal cursors on edges of H2O peak where they intersect vertical cursor
  • type “delta?”
  • returned value should be less than 100; if not you may need better shims

Making a New Shimmap

VnmrJ can automatically make a new shimmap for you by selecting the AutoMakeShimmap button. If you want to make one manually, here is how I (S.E.) have always done it:
First collect TrialSpectra. Display 1st profile by typing in “ds(1) f full”. Press horizontal cursor button (yellow button on left side of graphics) and put yellow line (threshold) so it bisects the 2 peaks of the humps. Decrease the threshold by typing “th=th*0.18” Now put both vertical cursors and position them at intersections of yellow and red lines. Now select WindowFromCursor button. Make sure all parameters are correct. For H2O samples use d1=1 (500) or d1=3 (800) and nt=2, For shimming on D2O use d1=3 (500 or d1=6 (800) and nt=2 or nt=4. Finally press MakeShimmapUsingCurrentValue button.

Varian says it is not necessary to make a new shimmap for different samples. Some users say otherwise. If you are collecting data for several days, the 5 minutes it takes to make a new shimmap is small.


The system administrator (user vnmr1) should periodically adjust the shims using the lineshape standard and then save the data in /home/vnmr1/vnmrsys/data/lineshapes. General users should then only need to touch up the z1, z2, z3 (and maybe z4/z5) for their specific sample.

Insert the lineshape standard sample in the magnet. Allow the temperature to equilibrate for 5 minutes (at 25C?), find lock, and tune the probe. Load a recent lineshape spectrum, which should set all the parameters for you, including the proper spectral window. If you want to start over with these saved shim values then type in “load=‘y’ su” immediately after reading in the lineshape file. Note the acquisition time is around 16 seconds in order to accurately describe the shape of the narrow chloroform line. Manually touch up the z1 and z2 shims and maybe x1 and y1 shims. Perform gradient shimming using the H2_map shimmap (or make another shimmap using H2_map as the starting map). The administrator may then touch up the higher order nonspinning shims. General users should NOT change any of the nonspinning shims unless they know what they are doing.

Place the cursor on the single chloroform line and type 'nl' to center the cursor on the line. Type 'dres' to obtain the linewidth at half height, at 0.55% intensity, and at 0.1 % intensity. The three values characterize the lineshape and are expressed as the ratio of the three values in the order 50% / 0.55% / 0.11% (e.g. 0.5 / 3 / 6).

Optimal lineshape on the 500 and 800 differ. Check the log books to see what has been obtained recently. On the 800 we typically obtain around 0.8 / 5 / 10. On the 500 we obtain around 0.45 / 4 / 8.

Normally you should not have to touch the higher order non-spinning shims (i.e. beyond x and y). The z-shims (the so-called spinning shims) are optimized using gradient shimming. (Note: we never spin samples.)