[dsg-hdice] minimum requirements on Rack #2

[Dr. A.M. Sandorfi]

Dear Amanda,

The minutes of your 12/8/16 meeting requested my input (1.3.ii.1) regarding
³What will need to be on the rack by mid-January?² Since we still have rack
#1 in its old g14-configuration as a fall back, let me make your question a
little more explicit and address what are the MINIMUM (albeit incomplete)
requirements of a new rack #2 that would make it worthwhile to start using
it in January.

To set the context, barring any cryogenic issues, an NMR rack will be needed
by about the third week of January. Once it is used to establish an NMR
calibration for a set of targets, it will need to be temporarily "frozen"
until those targets are used, typically about 3 months later. If rack #2 can
be demonstrated to be functioning reliably (which will take time to
validate) and so used in January, then development work on rack #1 (to bring
it up to the level of the new rack #2 and beyond) could begin in parallel.
If a new rack #1 were then further developed (and thoroughly tested) by the
time the polarized targets are ready (~April-May/17), both racks could be
used to obtain an NMR signal from the same highly polarized target, thereby,
IN PRINCIPLE, transferring the calibration from rack #2 to rack #1. This
might free up rack #2 and allow it to be brought up to a level equivalent to
rack #1. However there are some points of caution that need to be
investigated ­ the absolute calibration is needed to track the polarization
as it decays. The cross calibration provides only a single reference point,
so the non-linearities (typically about 5%) have to be determined, and this
process is not trivial.

For the January/17 NMR calibration, a new rack #2 must function reliably
beyond the level of the old g14-era rack #1. In particular, the following
minimum functionality is required:
- a working Fast Resonance Scan program;
- a working NMR scan program, which controls the Oxford magnet power supply
and causes the field to follow the user input ramp sequence for a range of
choices in selected input parameters ­ RF power, attenuation, ramp ranges,
ramp times, hold times, etc. (The g14-era rack #1 functions properly for
only one peculiar set of conditions - a 31 s ramp down and 31 s ramp up,
with a fixed 2 s hold in between. This limitation is one of the high
priority things we had requested be overcome in the new racks.)
- in lieu of sweep synchronization via the CAEN shunt, it must be
demonstrated that multiple sweeps can be added together to reduce noise
without broadening the averaged NMR signal by more than an additional 4
gauss. For this the NMR signal and the magnet current must remain
synchronized to better than 10 ms. (Note the hidden issue here: there is a
potentially variable time delay between sweep commands generated by LabView
and the corresponding signals received by the Oxford supply. This can depend
on the LabView code and possibly even on other programs running on the PC.
Delays larger than 10 ms can washout the average and deteriorate the
signal-to-noise ratio.)

 I should stress that this is NOT a complete set of the ultimate
requirements, but rather the minimum level at which a new rack becomes an
³improvement² over the old. There is a definite value in starting to use a
new system in a real application, since it reveals issues better than simple
bench tests. It is in that spirit that we hope to be able to move over in
January to a new rack #2, even though its functionality may still be
limited.

Please continue to work with Xiangdong Wei to quantify outstanding issues
and validate functionality. We're optimistic that the new DSG racks will
ultimately provide important expanded capabilities, even if the road to that
goal is somewhat bumpy.

Andy






On 12/14/16, 10:05 AM, "Amanda Hoebel" <amandah at jlab.org> wrote:

> Hi,
> 
> Mary Ann, Peter, Xiangdong, Amrit, and I met last week to discuss progress for
> the NMR program. Attached is the meeting minutes.
> 
> Regards,
> Amanda
> _______________________________________________
> dsg-hdice mailing list
> dsg-hdice at jlab.org
> https://mailman.jlab.org/mailman/listinfo/dsg-hdice