| channel |
problem |
affected run and LBs |
actions |
channel overall status |
supporting plots |
| EBA51 ch00-23 |
unstable timing for half a module
- happens during the run
- related to LVPS power-cycling of the EBA51 module (typically)
- change in mean time happens simultaneously for each channel
- change in mean time is approximately the same for all digitizers
|
two periods with unstable timing in 2018, up to now
- few from 348251-349693 (frequent reconfigurations)
- some from 355109-357193 (multiple automatic reconfigurations, manual power cycle)
|
- periods with time jump will be corrected for 2018 data reprocessing
- LBs with very unstable timing to be flagged with BAD_TIMING
|
- The reason was found out to be unstable 3V output current on TTC - interface card to be replaced during the maintenance
- discarding >70% events from 18-Oct-2018
|
 Timing instability in run 348354, all channels in digitizer (and generally ch00-23) simultaneously affected.  Timing instability in EBA51 ch00-23 during the run 357193. |
| EBA51 ch00 |
time offset about (+3) ns |
- occurred during 360161, since LB=296, directly following power cycle of EBA51
- up to now
|
time constant to be corrected for reprocessing
- provided good pulse shape
- no change in reconstructed energy
|
no response in LG in laser and CIS calibration runs since 350161 |
 Timing shift in EBA51 ch00 occurred during the run 360161: at LB=298 (power cycle of the module) and LB=331 (LVPS trip) . |
| EBA12 ch 20 |
fraction of events has 2x bunch-space offset (-50) ns
- a fraction of events with (-50) ns offset is typically below 1%
- mean time in physics seem sot be not affected
|
almost each run since 348403 when number of colliding bunches increased from 8 to 62 |
The issue concern all DMU 6 (ch 18 and 19 are absent). Out-of-time events could be separated in laser (provided the signal is known). The cut which will help to distinguish offset events in physics data is under development. Decided not to mask the channel
- provided relatively small fraction of affected events
- no significant impact of physics data
|
OK on DMU level |
 Fraction of events registered by EBA12 ch20 has (-50) ns (run 357193 shown).  Pulse shape in EBA12 ch20 has (-50) ns (run 351223, all LBs, shown). |
| EBA40 ch 39-41 |
fraction of events has 2x bunch-space offset (+25) ns
- a fraction of events with (+25) ns offset is typically below 1%
- mean time in physics seem sot be not affected
|
almost each run since 348403 when number of colliding bunches increased from 8 to 62 |
The issue concern all DMU 13 (ch 18 and 19 are absent). Out-of-time events could be separated in laser (provided the signal is known). The cut which will help to distinguish offset events in physics data is under development. Decided not to mask the channel
- provided relatively small fraction of affected events
- no significant impact of physics data
|
OK on DMU level |
 Fraction of events registered by EBA40 ch39-41 (DMU5) has (+25) ns (run 340697 shown). |
| EBA27 ch01 |
time offset about (-4) ns |
- occurred during 357679, since LB=550, directly following automatic power cycle of EBA27 as over-current
- up to now
|
time constant to be corrected for reprocessing
- provided slight undershoot in pulse shape in HG and not reasonable pulse shape in LG
- change of reconstructed in HG energy by 20-25% (from 1.3 pC to 1.6 pC) although eFit distribution is Gaussian-like
|
no CIS response in both gains since 357679 (supposedly burned 3-in-1 card) |
 Timing offset in EBA27 ch01 (run 357713 shown).  Pulse shape in EBA27 ch01 (run 357713 shown). |
| EBA53 ch01 |
sporadic offset (+3) ns seen in laser-in-gap events
- happens during the run
- concerns only one channel -- ch01
|
- some from 349014-350751
- 354309, 355261, 355848
- some from 357409-359191
- 360293
- mean time in physics data is not seemingly affected
|
zeros in HG followed by switch to LG where signal is mis-reconstructed (bad pulse shape in LG)
- increasing pedestal for this channel to be performed on software level
- to be flagged with BAD_TIMING
|
LG ADC masked (unspecified bad). The channel is known for problems with gain switch. |
 Timing instability in EBA53 ch01.  Pulse shape in EBA53 ch01, during LBs with affected timing. |
| EBA16 ch00 |
sporadic offset (-3) ns
- happens during the run
- concerns only one channel -- ch00
|
from 349534 up to now
- observed from run to run
- mean time in physics data is about (-1) ns
|
zeros in HG followed by switch to LG where signal is mis-reconstructed (bad pulse shape in LG)
- increasing pedestal for this channel to be performed on software level
- to be flagged with BAD_TIMING
|
since (335237,0) HG marked as no CIS calibration. The channel is known for problems with gain switch (supposedly burnt 3in1 card). |
 Timing instability in EBA16 ch00-23. |
| EBA08 ch07 |
time offset about (-4) ns |
- occurred during 359191, since LB=922, directly following automatic power cycle of EBA08 as over-current
- up to now
|
set bad timing from (359191, 922)
- provided distorted pulse shape in HG and reasonable pulse shape in LG
- change of reconstructed in HG energy by 20-25% (from 3.2 pC to 4 pC) and eFit with pultiple-peak structure
|
CIS and laser calibrations
- no CIS response in both gains since 360833 (supposedly 3-in-1 card problem)
- laser constant drift since 360833
|
 Timing in EBA08 ch07 during the run 359286.  Pulse shape in EBA08 ch07 (run 360560 shown).  Time offset in 359279 in physics data. |
| EBC57 ch11 |
time offset of varying value
- observed only on a single channel (ch11 in extended barrel - A13)
- seen on laser-in-gap events and physics data
- time offset happens out of run, first LB already has the offset
|
some runs from 349114-351455
- recovered after power cycle during 351455
- seems to be stable in 351550 and onwards
|
- pulse shape is OK
- to be corrected for reprocessing
|
calibration (observed throughout 2018 data-taking, not only in the runs with affected timing)
- mono-gain CIS: 100% SB5-9 in LG
- x2 larger RMS of amplitude in pedestal runs
|
 Time offset in EBC57 ch11 in run 349335.  Time offset in EBC57 ch11 in run 351062 (physics data).  Pulse shape in EBC57 ch11 in run 351062. |
| EBA19 ch00 |
mean time changes (+- 10 ns roughly)
- seen in laser-in-gap events and physics data
- following HV change and change of data-taking conditions
- since 348836 (+2.0) ns in laser-in-gap-events
- could vary in physics (due to non-gaussian shape of time distribution and/or low statistics runs). Please, report if the offset is above 10 ns in physics, and time in laser changes (about (+2.0) ns now)
|
two periods (up to now):
- 348403-348836 (changes of HV applied)
- 354124-355468 (vdM, beta and low-mu runs)
|
pulse shape in LG is not good. Overflow in HG -> switch to LG |
laser LG calibration run: very large reconstructed response (evolving: 900+-35 pC at the beginning of data-taking, 4500+-180 pC in September 2018) |
 Time offset in EBA19 ch00 in run 348836.  Pulse shape in EBA19 ch00. |
| EBC25 ch15-17 |
time offsets in a few LB seen in laser-in-gap events
- no effect on physics (several LBs with time offset)
|
throughout data-taking
- run 348895, 350803, 351270, 351698, 352030, 352494 , 354494, 355224, 355563, 355599, 357293, 357539, 358175, 358300, 358325, 358333, 358577, 358615, 359471, 359472, 359717
- may happen another time. Please, report about it.
|
pulse shape is not good, even in runs with well reconstructed timing
- candidate for masking , DMU replacement
|
- large noise in mono-gain CIS and laser calibration runs (in HG x30, in LG ~x7)
- saturation 1% in laser HG run (could be indication of wrong switch circuit)
|
 Time in EBC25 ch15-17 (run 362345).  Pulse shape in EBC25 ch15. |
| EBC38 ch01 |
time offset (-4) ns
|
runs 352436-353907 (laser-in-gap)
- since (354021, 0) EBC38 ch01 LG flagged with bad timing
- since (354477, 0) EBC38 ch01 LG no CIS calibration
|
zeros in HG (in last sample) leading to switch to LG
- pulse shape distorted
- energy affected by 10-20%
- flagged with bad timing
|
no CIS calibration |
 Time in EBC38 ch01 (run 352436), after offset appeared.  Pulse shape in EBC38 ch01, after the offset appeared. |
| EBA10 ch08 |
time offset (-3.5) ns
- occurred during the run
- affected after overcurrent and automatic power cycle as over-current (LBs affected due to the trip: 442 - 470)
|
since (363710, 480) |
- distorted pulse shape
- undershoot in pulse shape
- reconstructed energy is overestimated w.r.t. its usual value: ~25%
- decision: correct timing for reprocessing (?)
|
- monoCIS LG: no response in LG, noise-like amplitude in HG
- monoCIS HG: no amplitude reconstructed in HG
- laser LG: 10% zeros, time offset
- laser HG: response in both gains
|
 Time shift in EBA10 ch00 happened during the run 363710.  Pulse shape in EBC10 ch08, in the run 363830, after the offset appeared. |
| EBC50 ch01 |
time offset (-3.5) ns
- occurred during the run
- affected after automatic reconfiguration (due to overcurrent) and trip
|
since (363979,500) |
- distorted pulse shape with undershoot
- reconstructed energy is over-estimated by ~25%
- decision: correct timing for reprocessing, the energy adjusted by the laser calibration factor
|
- monoCIS LG: no response in LG, noise-like amplitude in HG
- monoCIS HG: no amplitude reconstructed in HG
- laser: time offset, indication of zeros (~16%)
|
 Time shift in EBA50 ch08 happened during the run 363979.  Pulse shape in EBC50 ch08, after the offset appeared, in the run 364030. |
| EBC40 ch16 |
time offset (-3.5) ns
- occurred on the fly
- neither LVPS trip nor power cycle at that moment
- similar pattern is seen for channels in odd-numbered DMUs: 1, 3, 5, 7; time offset is present for all channels, but to smaller extent
|
since (364214, 492) |
- undershoot in samples [0..2] in some events
- double peak in reconstructed energy
- decision -- to mask this channel
- the channel is already masked in runs 364214, 364292
- will be observed in heavy ions collisions
|
- ch21,22 are dead in monoCIS (LG and HG) and laser
|
 Time shift in EBC40 ch16 occurred during the run 364214.  Pulse shape in EBC40 ch16, in the run 364292, following one after the offset appeared.  Reconstructed energy-time scattering plot: events with ~(-4) ns have 25% higher energy reco'ed. |
Home 
Plugins
Sandbox for tests
Support
Alice
Atlas
CMS
LHCb
Public Webs
Sandbox Web 
Create New Topic
Index
Search
Changes
Notifications
RSS Feed
Statistics
Account 
Copyright &© 2008-2023 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.