-- YurySmirnov - 27-Sep-2011

General info

This page is a manual on how to analyze files volt.runNumber.txt that are part of TRT Calibration. The main goal of this study is to find incorrectly mapped TRT HV lines in DB by comparing their voltages and straws efficiencies.

There can be two line types in this file:

1. Line    HV_line_name    has low voltage but regular efficiency

2. All straws belonging to    HV_line_name    are low efficient but its voltage is regular

First of all, we are looking for files having both of these lines. E.g. if you see only 1 line there (of any type) or N lines but all of the same type you can skip this file: the main idea here is that a correctly mapped HV line should have both low voltage and lowered straws efficiency (in case we have a pretty long trip and pretty large statistics); an incorrectly mapped one has low voltage OR lowered efficiency and the one it is mixed up with has lowered efficiency OR low voltage respectively.

"HV line has low voltage" means that there was at least one LumiBlock (more specifically - first event of a given LB) when HV of this line was < 1490 V. "HV line has regular voltage" means that every HV line in every LB's first event had voltage > 1490 V.

"Low efficient HV line": if the efficiency of every single straw connected to this HV line (after accumulating all hits and holes on all tracks for all events the calibration runs over) is < 0.8. If there is at least 1 straw related to a particular HV line with efficiency > 0.8 such HV line will have "regular efficiency".

In most cases you'll see an empty volt.runNumber.txt . In case there are some lines (even only one) my program will notify me via email if calibration was performed using TRT_CalibAlgs-00-00-73 or any newer tag. In case you see 1 line you can skip this file as well: it can not be due to incorrect HV line mapping. Probably it happened because of low statistics or very short trips that did not affect overall efficiency. Please note that in cases of low statistics in pp or HI runs or high occupancy in HI runs you can see hits on HV lines that are known to be switched off like HVA/S19S20/WA4/1T .

What to do if there are 2+ lines of different types

Well, if you've spotted such kind of situation, all you can do is to make sure it is not some kind of bug. This way you have to manually check HV line voltage for a given period of time and to make sure there was a trip on some HV line but efficiency is OK or vice versa - efficiency is low but there was no trip.

1. Voltage

To make sure there was (or there was not any) trip you have to open DbExplorer (or set it up if you don't have it - as described here). In DbExplorer Console choose Plots -> Histogram Trends & 3D. Now DB Data Viewer window is open. In the right upper corner choose ATLAS_PVSSTRT, then in a section below choose End-Cap_A_Hight_Voltage or End-Cap_C_Hight_Voltage (that depends on whether you have HVA_* or HVC_* HV line with lowered voltage/efficiency in volt.runNumber.txt) right under TRT_HV red line. In the section below click on "Aliases" and choose the one you're intersted in from the list. E.g. if you see HVC_S13S14_WB7_T in volt.runNumber.txt you have to choose HVC/S13S14/WB7/T/OutputVoltage ##tbl ATLAS_PVSSTRT ##sys_id 5 in DbExplorer. Then below you have to specify run start time and run end time. You can get run number from the volt.runNumber.txt filename itself and then go to ATLAS Run Query page. E.g if you're looking for run # 189027 just insert f r 189027, press "Show runs". Please note that run start and end time and date are in CEST! So now we have to insert these dates and time into DbExplorer in GMT, not CEST. Then click "reDraw". If voltage of this line was measured at least once in a given period of time you'll see a graph in left upper corner: voltage vs time. Now we know whether a trip happened or it did not according to PVSS (Prozessvisualisierungs-und Steuerungs-System).

As an option (not as reliable though) you can check relevant entry on TRTDetectorHVTripRecord instead.

2. Efficiency

One can calculate efficiency for a given HV line by hand using straws.runNumber.txt. This file has 350848 (# of straws in TRT) + 1 (first line with # of events calibration runs over) lines. Taking lines that correspond to particular HV line one can calculate the efficiency. Unfortunately this file does not provide HV lines names - they can be taken from HVLinesNames_slimmed.txt. So you have to download and untar this tarball

, open it in any text editor with lines numbering and search for an HV line name you're interested in. Once you've found one note the line number and take a look into exactly the same line number in straws.runNumber.txt. Straw efficiency = (fifth digit)/(fifth digit+eight digit). If it is < 0.8 it is low efficient straw. Look for all straws (generally 192 straws per HV line in Endcap, though there are 2 exceptions). If there is at least one straw with efficiency > 0.8 this HV line should not be called "low efficient one". Otherwise it should. Please note that in case of low statistics (number of events calibration jobs runs over can be found as last number on the first line of corresponding straws.runNumber.txt) there can be many false-positive "low efficient HV lines". Please consider runs with N_events >~ 10000 only.

Yeah, I know all that sounds time-consuming but first of all I'll check it too so maybe you don't need to do that. Second of all, I do not expect it will happen more often than once per 1.5-2 months.

So if everything seems to be correct we will keep looking more closely on this HV line in the future. There are pretty good chances this one is an incorrectly mapped one!

Topic attachments
I Attachment History Action Size Date Who Comment
Compressed Zip archivetgz HVLinesNames_slimmed.tgz r1 manage 60.2 K 2011-09-29 - 19:54 YurySmirnov Provides straw-HVline matching info
Edit | Attach | Watch | Print version | History: r9 < r8 < r7 < r6 < r5 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r9 - 2012-04-22 - YurySmirnov
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Sandbox All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback