[Frost] Moeller measurements for g9a

[Michael Dugger]

Hi,

Sorry if I sounded like a jerk :(

I am becoming very frustrated with the measurement of E for eta. There is 
a problem with either the code or the polarization values. So far, I have 
not found a problem with the code, so I have started to look at possible 
problems with the polarization values.

We have been in the process of fine tuning the code. We now have estimates 
for the TBTR vertex leakage that allow us to calculate the numerator of E 
without performing a carbon subtraction. The ability to neglect the carbon 
subtraction for the numerator greatly reduces the error bars, and we are 
happy about that. We also compensate for the TOF trigger inefficiency in 
the calculation of the scale factors and this makes the final E results 
smoother. However, after all of this, we are still getting E values at 
W=1525 MeV that are too large.

If we renormalize the polarization such that P_new = P*1.082 we get the 
following results:

W = 1525 -> http://www.jlab.org/~dugger/tmp/w1525.gif
W = 1575 -> http://www.jlab.org/~dugger/tmp/w1575.gif
W = 1625 -> http://www.jlab.org/~dugger/tmp/w1625.gif
W = 1675 -> http://www.jlab.org/~dugger/tmp/w1675.gif
W = 1725 -> http://www.jlab.org/~dugger/tmp/w1725.gif
W = 1775 -> http://www.jlab.org/~dugger/tmp/w1775.gif
W = 1825 -> http://www.jlab.org/~dugger/tmp/w1825.gif
W = 1875 -> http://www.jlab.org/~dugger/tmp/w1875.gif

Notes: The red lines are SAID and the blue lines are eta-MAID. At W=1775 
we begin to distinguish between SAID and MAID (in favor to SAID). The 
negative value on the W = 1825 MeV looks to be real and something that 
SAID does not predict.

I think that the current results would be potentially interesting to 
theorists. I just wish that I understood why we see the need to normalize 
the polarization.

We will continue to try and figure out what is going wrong.

Thanks for your time.

Sincerely,
Michael


On Thu, 10 Mar 2011, Eugene Pasyuk wrote:

> Mike,
>
> I don't say they are not important. They are. At the same time it is
> reasonably well established procedure. The good thing about it is that
> the analysis is done on the fly and we get instant feedback. If there
> are reasons to suspect that something is wrong with the measurements we
> always redo it right away.
> Yes it happened in the past that for some reason beam polarization
> disappeared. We did not have such incidents in our run. That was
> verified by Steffen who extracted E for pi+ for every single run and did
> not see any anomalies.
> However in that analysis it was assumed that target polarization is a
> constant. Since now we know target polarization on run-by-run basis we
> should revisit this.
> Having said that still we have to use other means to verify that the
> number is correct. In electroproduction experiments they use elastic ep
> to get beam polarization from the data because for this reaction
> asymmetry is known. They can do it more or less on-line too. We did not
> have this on-line opportunity with photon beam.
> In our case we we can use some reactions where we know the asymmetry and
> use this to cross check normalization. One of the spots where we can do
> that is eta at threshold where we have essentially only A(1/2) present.
> The other spot is pi+ production at low energy where there are previous
> measurements.
>
> -Eugene
>
> Michael Dugger wrote, On 03/10/11 01:11:
>>
>> Eugene,
>>
>> Why bother making the measurement at all. It sounds like a waste of time.
>>
>> If the Moeller measurements are important then we have something to argue
>> about. If you take the position that the measurements don't matter, then
>> what can I say?
>>
>> I'm getting close to normalizing the polarization to the eta 1525 MeV
>> measurement. Perhaps there is nothing to discuss.
>>
>> Take care,
>> Michael
>>
>>
>> On Thu, 10 Mar 2011, Eugene Pasyuk wrote:
>>
>>> I don't think anything is wrong with measurements. Our measurements were
>>> consistent with other halls with given linac energy, Wien angle setting
>>> and Mott measurements in the injector.
>>> If there are no changes in the accelerator configuration thera no
>>> reasons for beam polarization to change. There are a few parameters that
>>> may affect polarization: linac energy, Wien angle and there is some
>>> correlation with quantum efficiency of the photo cathode.
>>> The first two were kept constant for given accelerator energy. The
>>> energy also was constant within certain periods when we intentionally
>>> changed it. We always did Moeller measurements after they did spot move
>>> on the cathode or any other changes that may affect polarization.
>>>
>>>
>>> -Eugene
>>>
>>> Michael Dugger wrote, On 03/10/11 00:28:
>>>>
>>>> Eugene,
>>>>
>>>> We have three Moeller measurements that do not have a +/- pair. The runs
>>>> lacking the pairwise measurements are: 55608, 56194, and 56202. It looks
>>>> like these Moeller measurements need to be thrown out.
>>>>
>>>> I still think that the Moeller measurement for run 55552 is probably
>>>> garbage. I do not see how that measurement can be trusted.
>>>>
>>>> -Michael
>>>>
>>>>
>>>>
>>>> On Thu, 10 Mar 2011, Eugene Pasyuk wrote:
>>>>
>>>>> There was some incorrect handling of beam charge asymmetry correction in
>>>>> Moeller measurements at the beginning. For that reason we always did two
>>>>> measurement with opposite Moeller target polarization.
>>>>> If you take an average of the two without sign, the it cancels that
>>>>> problem.  Uncertainties shown there are statistical only. Sytematics of
>>>>> our Moeller polarimeter is believed to be between 2 and 3%
>>>>> If you do this average, than you will see that all three pairs of
>>>>> measurements are consistent with each other within their uncertainties.
>>>>> For the first pair it is 81.9%, for the second one it is 83.9 and for
>>>>> the third one it is 84.4%
>>>>>
>>>>> -Eugene
>>>>>
>>>>> Michael Dugger wrote, On 03/09/11 18:23:
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> It looks like there are potential problems with some of the Moeller
>>>>>> measurements.
>>>>>>
>>>>>> Run 55544:
>>>>>> We have 4 measurements taken over a 14 minute time period with
>>>>>> P = 85.228 +/- 1.42
>>>>>> P = -78.523 +/- 1.35
>>>>>> P = -79.150 +/- 1.26
>>>>>> P = 88.700 +/- 1.48
>>>>>>
>>>>>> There is a large change between the polarization values (>    10%). The Two
>>>>>> negative measurements are much lower than is typical and I want to take
>>>>>> these out of the average. Is there a reason not to remove these negative
>>>>>> measurements?
>>>>>>
>>>>>> Run 55552:
>>>>>> We have 2 measurements:
>>>>>> P = 84.167 +/- 1.33
>>>>>> P = -84.725 +/- 1.53
>>>>>>
>>>>>> The first measurement had True Rates = 0.00 and Accidental Rates = 0.00,
>>>>>> also the SLM Current = 0.125 (whereas most all other runs had SLM
>>>>>> Current ~ 6). The second measurement had True Rates = 1002.00 and
>>>>>> Accidental Rates = 3.50, with SLM Current = 1.192. (Typically the True
>>>>>> Rate is about 6000). Moreover, 2C21 X = 0.0 (-4.455) and 2C21 Y = 0.0
>>>>>> (1.560) for the first (second) measurement.  I think that the Moeller
>>>>>> measurements for run 55552 are highly suspect and should be thrown out of
>>>>>> the determination of the polarization average.
>>>>>>
>>>>>> Additional question: How important are the 2C21 X and Y values for the
>>>>>> Moeller measurement?
>>>>>>
>>>>>> Take care,
>>>>>> Michael
>>>>>>
>>>>>> _______________________________________________
>>>>>> Frost mailing list
>>>>>> Frost at jlab.org
>>>>>> https://mailman.jlab.org/mailman/listinfo/frost
>>>>> _______________________________________________
>>>>> Frost mailing list
>>>>> Frost at jlab.org
>>>>> https://mailman.jlab.org/mailman/listinfo/frost
>>>>>
>>>> _______________________________________________
>>>> Frost mailing list
>>>> Frost at jlab.org
>>>> https://mailman.jlab.org/mailman/listinfo/frost
>>> _______________________________________________
>>> Frost mailing list
>>> Frost at jlab.org
>>> https://mailman.jlab.org/mailman/listinfo/frost
>>>
>> _______________________________________________
>> Frost mailing list
>> Frost at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/frost
> _______________________________________________
> Frost mailing list
> Frost at jlab.org
> https://mailman.jlab.org/mailman/listinfo/frost
>