[Sbs_gems] [EXTERNAL] RE: Efficiency plots from SBS-14 GEM studies

Tue Feb 8 14:51:28 EST 2022

Hi Andrew,
Thanks for getting back. Yep I agree with the approach you are advancing.
Also I quickly looked at the replay for the high beam current / high rate studies yesterday and I agree that the initial results looks indeed very interesting. The discussion of the results at tomorrow's meeting will be very exciting indeed

Best regards
Kondo

From: Andrew Puckett <puckett at jlab.org>
Sent: Tuesday, February 8, 2022 1:18 PM
To: Gnanvo, Kondo (kg6cq) <kg6cq at virginia.edu>; Liyanage, Nilanga K (nl8n) <nl8n at virginia.edu>; Rathnayake, Anuruddha (adr4zs) <adr4zs at virginia.edu>
Cc: sbs_gems at jlab.org
Subject: Re: Efficiency plots from SBS-14 GEM studies

Hi Kondo,

Thanks for putting these plots together, they are very informative. I agree with your interpretation of the efficiency drop in the back layer, it must be caused by an increase of "fake" tracks from the front four layers. One option I am considering to improve the tracking performance at high rate (for BigBite anyway) is to basically REQUIRE the back GEM layer to be on the track. This of course would lead to a slight drop in overall reconstruction efficiency but would presumably allow us to reduce the "fake track" fraction quite dramatically. My original presentation about the software to the GEN-II ERR committee basically assumed we would do this. It would allow us to shrink the track search region area at the front layers by at least a factor of 4 compared to what we are doing now, which relies only on calorimeter information. I started to look at the plots in the logbook from the high current test. They are very interesting. Even though I cannot attend the full GEM meeting I will try to join for the second half of tomorrow's meeting after my class is done.

Andrew

From: Sbs_gems <sbs_gems-bounces at jlab.org<mailto:sbs_gems-bounces at jlab.org>> on behalf of Gnanvo, Kondo (kg6cq) <kg6cq at virginia.edu<mailto:kg6cq at virginia.edu>>
Date: Monday, February 7, 2022 at 2:31 PM
To: Liyanage, Nilanga K (nl8n) <nl8n at virginia.edu<mailto:nl8n at virginia.edu>>, Rathnayake, Anuruddha (adr4zs) <adr4zs at virginia.edu<mailto:adr4zs at virginia.edu>>
Cc: sbs_gems at jlab.org<mailto:sbs_gems at jlab.org> <Sbs_gems at jlab.org<mailto:Sbs_gems at jlab.org>>
Subject: [Sbs_gems] [EXTERNAL] Efficiency plots from SBS-14 GEM studies
Dear all,
I produced a couple of efficiency plots from the GEM scan studies performed during SBS-9 with LD2 with beam current from 1 to 10 uA.

Please check out the attached slides:
1-) Slide 1:

  *   on the left: is the GEM current vs. beam current for each of the 8 GEM chambers so basically a different way to look at Anu linearity plots. So of course the current increases (negative value ==> drop on the plots) as the beam current increases  and the drop is more significant for the UV layers than the 4 X-Y modules of the back layer

  *   On the right: is the efficiency for the same layers vs. beam current. The trend is the similar although not necessarily correlated with the current drop (or excess current). For example UV layer 3 efficiency drops more faster than the other 3 layers even though the GEM current drop (or excess current) is less pronounced.

2-) Silde 2:

  *   This is more interesting, Here I am plotting the efficiency as a function of the excess current (relative to the measured current for 1 uA) for each chamber
  *   I keep both x and y axis to the same scale so that we can concentrate on the amplitude of efficiency drop for same range of excess current

  *   Here, it more interesting to look at the 4 bottom plots for the XY modules where the GEM current drop is less than 10 uA for all 4 modules which represent less than 1.5% of the nominal GEM so this should not cause any efficiency drop, but we are still looking at 20 percentage point drop at the minimum for each of these modules

  *   This 20 point drops in efficiency for such small current increase can only be explained by fake tracks from the 4 U-V layer. My hypothesis is that these fake tracks are partially caused by the negative pulses impact on the common mode calculation

  *   Efficiency drop on the 4 UV-module for a excess current range of 10 uA is limited to a couple of percentage point typically less than 5% which is 4 time smaller and I suspect that the XY layer which is less prone to these negative pulses reduce the probability of fake tracks and therefore the efficiency drop.

I think this is another clear indication that solving the divider hardware issue is only going to get us that far unless we address the negative pulses issues.

It will also be good that the similar plots are produced for the other GEM studies and correlated with the common mode calculation. I would not have time to do so though

Best regards
Kondo

_______________________________________________
Sbs_gems mailing list
Sbs_gems at jlab.org<mailto:Sbs_gems at jlab.org>
https://mailman.jlab.org/mailman/listinfo/sbs_gems
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/sbs_gems/attachments/20220208/dc1d56b2/attachment-0001.html>