[Frost] target depolarization

Wed Mar 23 12:30:45 EDT 2011

Chris,

The photo shows a coil with two turns. Is it the LF that is on the other 
side?

The web page:
http://www.jlab.org/~ckeith/Frozen/Photos/Photos.html
that includes the photo you link to has in the caption:
"A similar, one-turn coil is located on the opposite side of the mixer and 
is used during the polarization process. FEP heat shrink is used to hold 
the NMR coils in place."

I'm refining the measurement to see if the effect is real.

I have produced the de-polarization plot with background removed using 
side band subtraction. The de-polarization effect is still large but not 
as large as before. The new plot can be found at
http://www.jlab.org/~dugger/g9/g9a/polZ2.gif

The side band subtraction should give a good estimate of the background
subtracted signal, but just to make sure that the plot actually shows the
de-polarization effect, I'll do a polynomial subtraction and see how
that changes things.

In a previous email you wrote:
"It's important to remember that different coils were used for the high
field and low field NMR systems, and it is unlikely that they sampled the
equal parts of the target with equal weight."

and

" The geometry of the LF coil was such that it sampled the downstream part 
of the target a little more heavily than the upstream.  So it "registered" more 
polarization loss following a frozen spin run (type 3 measurement) than did the HF 
coil (type 4).
"

It would be good to figure out how the LF coil sampled the data with 
respect to the z-direction.

Thanks for your time.

Sincerely,
Michael

On Wed, 23 Mar 2011, Christopher Keith wrote:

>
> Hi Michael,
> It's interesting that can see a polarization gradient in the target.  If it's 
> real, I'd expect to see it get worse the longer beam is on target.
>
> Here's a photo showing the LF NMR coil:
> http://www.jlab.org/~ckeith/Frozen/Photos/Mixer.jpg
> You can see that it's arranged to try to cover the entire length of the cell. 
> So it's not correct to ask "where the LF NMR measurements are made".   We try 
> to measure the average polarization along the entire length.  Same with the 
> HF coil, which is placed opposite the LF one.    It looks similar, only one 
> turn, not two, for lower inductance.  Hence, it's unlikely that they sample 
> the polarization exactly equivalently.  For the transverse running, I decided 
> to use the same coil for both HF and LF measurements.  That opened another 
> can of worms though.
>
> I don't think the HF NMR plots are indicative of a polarization gradient, but 
> I could be wrong.  The "ugly" ones come about from thermal drifts of the NMR 
> cable inside the cryostat.  The temperature profile of the cryostat changes a 
> lot when we start/stop polarizing, and we aren't always able to adjust for 
> it.
>
> Hope this helps
> Chris
>
>
>
> Michael Dugger wrote:
>> 
>> Hi,
>> 
>> I have some results regarding the target polarization as a function of 
>> z-vertex that might be of interest. In short: it looks like we can see the 
>> downstream de-polarization of the target using the pi+ n reaction, and this 
>> de-polarization is a large effect.
>> 
>> Method:
>> 1) Isolate regions of the gamma p -> pi+ n reaction where the E observable 
>> appears to be always less than -0.5. By looking at Brian's slides 
>> (http://www.jlab.org/Hall-B/secure/g9/morrison/brian_E_obs_2-10-2011.pdf), 
>> I chose the region in cos(theta_cm) between -0.4 and +0.7 for E_gamma 
>> between 1500 and 1600 MeV, and cos(theta_cm) between -0.7 and +0.8 for 
>> E_gamma between 1600 MeV and 1750 MeV.
>> 
>> 2) Slice the data into 32 z-vertex bins from z-vertex = -4.0 to +4.0 cm.
>> 
>> 3) Find the |E| observable, with no carbon subtraction or other background 
>> subtraction, for the missing mass (assumed reaction gamma p -> pi+ X) 
>> between 0.89 and 0.99 GeV/c^2. This was performed assuming that our current 
>> values of the beam*target polarization are correct.
>> 
>> 4) Plotted the results.
>> 
>> You can see the results here:
>> http://www.jlab.org/~dugger/g9/g9a/polZ1.gif
>> 
>> Top panel is z-vertex of the denominator of E.
>> Bottom panel is the absolute value of E as a function of z-vertex.
>> 
>> This looks to be a huge effect. We need to figure out how to get an average 
>> polarization.
>> 
>> Questions: Can we quantify the region in z where the low field NMR 
>> measurement took place? Is there any pictures, or diagrams the show the 
>> placement of the low-field NMR coils?
>> 
>> We can try and use the high field NMR data but there are a lot of 
>> "anomalous" measurements that would leave gaps in our polarization tables. 
>> See the entries (red implies anomalous measurement) in the list given on 
>> http://clasweb.jlab.org/rungroups/g9/wiki/images/6/66/Clf_spreadsheet3_landscape.jpg 
>> .
>> Moreover, in general, the the high field NMR measurements made previous to 
>> re-polarization might be questionable.
>> 
>> When I look at the baseline subtracted high field NMR signals on the FROST
>> webpage:
>> http://clasweb.jlab.org/rungroups/g9/wiki/images/c/c1/P4_2_areas_edited.jpg 
>> they look bad.
>> 
>> Question: Are these a fair representation of what the baseline subtracted
>> NMR signal looks like for the high field just previous to re-polarization?
>> 
>> All the other baseline subtracted NMR plots look OK:
>> 
>> High field after polarization ->
>> http://clasweb.jlab.org/rungroups/g9/wiki/images/c/ca/P1_1_areas_edited.jpg 
>> 
>> Low field after right after polarization ->
>> http://clasweb.jlab.org/rungroups/g9/wiki/images/d/d4/P2_1_areas_edited.jpg 
>> 
>> Low field just before polarization ->
>> http://clasweb.jlab.org/rungroups/g9/wiki/images/9/90/P3_2_areas_edited.jpg 
>> 
>> Question:
>> Could the poor quality of the baseline subtracted high-field NMR plots be
>> due to the large gradients in the butanol polarization with respect to
>> the beam?
>> 
>> Thanks for your time.
>> 
>> -Michael
>
> -- 
> ______________________________________________________________________
> Christopher D. Keith
> Jefferson Lab, MS 12H                   email: ckeith at jlab.org
> 12000 Jefferson Ave.                    ph:  757-269-5878
> Suite #6				fax: 757-269-5235
> Newport News, VA 23606
> ______________________________________________________________________
>