[Halld-cpp] [EXTERNAL] How much amorphous data should we take?

[Elton Smith]

Hi Naomi,

If we take data continuously for 2 h at 60 kHz, we accumulate about 430 M events. Of course it is unusual to have 100% beam. Looking at the recent history of runs and number of triggers accumulated, it looks like a “good” run has about 300 M triggers. If we take the canonical 50% efficiency (accelerator/detector), we might estimate 215M events.  Here is a suggestion: Aim for 240 M events per setting for polarized beam (may be less than 2 h) and 60 M for amorphous radiator. So the sequence would be the following:


Diamond 45(240 M events)/ Diamond 45(240 M)/Diamond 135( 240 M)/Diamond 135(240 M)/AMO(60 M)

As Rory mentioned this morning, we would
1. take about 6 days of full-target data;
2. switch to empty target for about a day.
3. Take about 10% of the triggers accumulated during full target.
4. Switch back to full target.

Obviously, this rough schedule can be tweaked as we see what the data look like.

Thanks, Elton.

------------------------------------
Elton Smith
Jefferson Lab MS 12H3
12000 Jefferson Ave STE 4
Newport News, VA 23606
(757)269-7625

On Jun 23, 2022, at 8:43 AM, Naomi Jarvis <nsj at cmu.edu<mailto:nsj at cmu.edu>> wrote:

That looks good to me, except that it would be easier for the shifters if you express it in triggers rather than time.
The runs shouldn't go over 2h anyway.  Maybe it would be easier to aim for a little less than 2h per run to allow for some down time for beam trips etc.

Last night I had them aiming for 150M triggers in the hope that they would get through  4 diamonds + maybe a little amo time before beam-down but there were so many daq problems that this didn't happen.

Naomi.

On Thu, Jun 23, 2022 at 8:29 AM Elton Smith <elton at jlab.org<mailto:elton at jlab.org>> wrote:
I received some feedback from Sean, Curtis, Alexandre and Alexander regarding the amount of amorphous radiator data needed.

Another reason for having empty target data that I had overlooked is that it proves very useful for MONITORING. For example, see Alexandre’s log entry  https://logbooks.jlab.org/entry/4006041. Alexander points out that we need regular amorphous data in order to normalize the coherent peak spectrum, which is used for monitoring during data taking. The consensus is that we may not need as much data as was taken for GlueX. Let me make the following proposal for the taking full-target data:

Sequence for Full-Target Data
Diamond 45(2h)/ Diamond 45(2h)/Diamond 135(2h)/Diamond 135(2h)/AMO(0.5 h). That results in 0.5/8.5h = 6% of time allocated to amorphous running. It also makes sure that we take the data uniformly through the running period.

I am inclined to continue the SAME sequence of data taking for empty target, including the amorphous running, so that we have consistent sets between full and empty.

Thoughts/comments?

Cheers, Elton



------------------------------------
Elton Smith
Jefferson Lab MS 12H3
12000 Jefferson Ave STE 4
Newport News, VA 23606
(757)269-7625

On Jun 22, 2022, at 5:31 PM, Sean Dobbs <sdobbs at jlab.org<mailto:sdobbs at jlab.org>> wrote:

Hi Elton,

Another thing we use the amorphous data for is to evaluate instrumental asymmetries.  This may be something to be concerned about since you're looking at tracks at small angles, but I also agree that for this you probably need less data than we have been taking.

Cheers,
Sean

On Wed, Jun 22, 2022 at 5:05 PM Curtis A. Meyer <cmeyer at cmu.edu<mailto:cmeyer at cmu.edu>> wrote:
Hi Elton,

    I think we have used the amorphous data to also check systematics when we analyze data away from the coherent peak. My guess is you probably need less than GlueX.

Curtis


Curtis A. Meyer | MCS Associate Dean for Research
The Otto Stern Professor of Physics
cmeyer at cmu.edu<mailto:cmeyer at cmu.edu> | 412 260 6290
[https://s3.amazonaws.com/htmlsig-assets/spacer.gif]
Carnegie Mellon University
Department of Physics, Wean Hall 8410
5000 Forbes Ave., Pittsburgh PA, 15213
www.curtismeyer.com<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.curtismeyer.com&d=DwMFAg&c=CJqEzB1piLOyyvZjb8YUQw&r=_G1S4SWp4fMj0LNNDBuhBg&m=HRiyoxfOJd_BOwdN44qIWbkRnWNMfWjvMC09D_3AGF0VpHlkWj47Ep4xYqgVnHY8&s=wLOE_B9_Rhjb0vYKlpn_bjPw7n7q2BwbT3ShEX0zo9I&e=>
[https://s3.amazonaws.com/htmlsig-assets/spacer.gif]
[https://s3.amazonaws.com/htmlsig-assets/spacer.gif]





On Jun 22, 2022, at 16:57, Elton Smith <elton at jlab.org<mailto:elton at jlab.org>> wrote:

Dear Hovanes, Mike, and others,

The CPP experiment has been considering how much data should be taken with the amorphous radiator and wanted to get your input. As you know, for GlueX we collected about 10% of the triggers on the amorphous radiator.

As, I understand it, the main arguments for collecting data with an amorphous radiator  are a) to be able to normalize the coherent spectrum and b) check systematics of the polarization. We need very little data for a), but  how much data we need to address b)? Also, if there are other used for the amorphous radiator data, please send us feedback.

We have already started taking data that can be considered production and all of it was taken using polarized beam. We will want to make a plan regarding how much amorphous radiator data we take so that it is interspersed with diamond radiator data. At the moment, we do not anticipate using the amorphous radiator data in the physics analyzes for CPP nor NPP.

Thanks, Elton.


------------------------------------
Elton Smith
Jefferson Lab MS 12H3
12000 Jefferson Ave STE 4
Newport News, VA 23606
(757)269-7625




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-cpp/attachments/20220623/d8e54894/attachment-0001.html>