Summary notes from the injection test meeting 12/8/08

Reported by Brennan & Mike

RP - Helmut Vincke

Survey at TDI.L2 showed 23 uSv/h, collimators in P3 were 80 uSv/h on W collimators. Looks OK.

The final radiological classification of the LHC sector 2 and 3 following the LHC-SPS synchronization can be found here. (c/o Helmut and Isabel.)

Totals:

• 3.4 10^12 extracted
• 7-8 10^11 on the TDI
• 2.7 10^12 to point 3

Helmut won't be here for the next injection test, Heinz will take over.

Timing/RF synchronization - Andy Butterworth

• Fast timing/synch: new prepulse system, stable
• Some GMT issues (ring number/bucket number events arriving too late – CBCM issue) which caused random “synch” of kicker pulse.
• Solved manually for the test
• temporary fix will be put in place this week for tests next week...

From Julian Lewis the day after the meeting:

concerning the bug where the Next Ring and RF buckets were sent out after the SPS LHC cycle had started ... I have corrected the RT task on the LHC gateway. The logic is as follows ....

• on receiving SPS BeamOut and if the next SPS cycle is LHC, send the events. This logic will work correctly even if the SPS super-cycle contains only one LHC cycle. No modifications to the CBCM are required.
• The beamout event will obviously occur before PS extraction, and the ramp down time must be at least 450ms so that the events arrive 450ms before PS extraction.
• The new RT task is therefore a definitive way to handle sending the synchronization events, Next-Ring and RF bucket.
• I have a temporary installation in place so that we can go back to what was running during the last test. We can go back in seconds if need be.
• As I have made a major revision of the FESA RT task, a test of the logic is required. This means a SPS beam with TO_LHC sequencer control and an injection test. If all these conditions are met, the events should be sent out. The sending of the NextRing event by the sequencer should be disabled.

• RF frequency control – phasing from LHC back to SPS – limited frequency range to +/-800 Hz around LHC nominal frequency – can double this for dispersion measurements next time(will give 2 per mille momentum shift)
• Used rephasing error signal – worked nicely

Injection systems - Brennan Goddrad, Volker Mertens

Injection systems worked fine.

MBIBH293 Malika reports after meeting: following the change of current we had to apply on the MBIBH293, I've checked the nominal current requested for this magnet. The MADX data is K= 0.002999255 which corresponds to 722.519A. In the file I checked (in Ks) the K data was correct. However the corresponding "I" was wrong from what I retrieved from the logging data base: Logging data base: 1- Before the field correction: 735.832 A <=> K= 0.0030373 2- After the field change: 723.526 A <=> 0.00299926 So it looks like we made the change in the current of the B280 exactly to come back to nominal requested value. Need to check calibration curve.

Aperture scans showed aperture limit of 5-6 mm between MSI and Q5

• Confirmed by RP survey to be at valve/pump at 6L2, 7 m upstream of Q5.
• News from Jan the day after: The cause of the misalignment seems to be understood. To align this region the Q5 and the MSI have been taken as reference, assuming that the MSI is aligned with the beam. However, the MSI has a vertical off-set of +10.4 mm. So this would explain at least a 5 mm error at the level of the vacuum valve, which is in between these elements. Also, this assumption made the alignment very difficult, as the equipment was not designed to be installed like this. Vincent is going to check this in the machine asap (this afternoon if possible); first in Point 2 and than in Point 8 and correct at both positions.
• MKI kickers flashover after ~5e10 p+ lost during aperture scans – care to be taken with these systems – method of loss interlocking being discussed
• Some updates/improvements being implemented to MKI controls and SW 5e9 lost several times at aperture limit upstream Q5
• No quench of Q5
• Kicker application urgently needs to be made more ergonomic.

Optics - Kajetan Fuchsberger

Beam sizes were 2 nm in horizontal, 1 nm in the vertical.

Response measurements

• used about 50% of correctors and all BPMs
• found a few issues (1 dead BPM, 1 inverted BPM, BPMWs systematically with inverted non-linear calibration factors).
• Found main focusing quadrupole strength about 1% too high.
• Monitor and corrector gains all look good, correctors within 15%, with one exception. See clear family dependence. Need error bars.
• Saw problems with fit after ~11L3.Candidate was polarity of MQTL1.11L3 – tried to verify this experimentally – results inconclusive

Dispersion measurement & screen matching- Verena Kain

• Energy matching – trimmed SPS by 0.2% - had a residual of about 0.05%. Some effect seen after recycling – 2 units of B1.
• SPS frequency trim limited to 800 Hz (+/-0.1%).
• Looks very good in the horizontal plane – starts diverging after ~11L3.
• In vertical plane might be a bit of residual dispersion, but maybe just due to the noise on the BPMs at these low intensities.
• Optics at injection point with screens – parasitic measurements – find of the order of 10% H mismatch and 5% vertically.

Higher order corrector polarities - Frank Zimmermann

• Need larger trims next time.
• Suspicion that the QT11.R2 is inverted (improved agreement with measurements in arc).
• Checked polarity of QL12.R2, and of QL13.R2 – look fine.
• Skew quadrupole QS polarity checked, results inconclusive.

Checked optics in warm region with TCP collimators out and beam onto TCLA.

BLMs - Bernd Dehning

• System worked from first shot! Some channels with wrong references settings – some fixed during test – some still being fixed.
• Monitors proved to be very sensitive – for 40 us quench at 2.1G/s, saturation at 23 G/s at least a factor 100. Can also see 1% of a pilot bunch (5e7).
• Some wrong SEM cabling in P3 – inverted channels – need modified test procedure.
• During MBB quench – measurements consistent with quench estimates and energy deposition calcs. Rates were 0.01 to 2.1 G/s.
• During aperture scans - for MQ the measurements of losses a factor of ~4 above estimates.  Very good case study – almost all required information recorded. In MBB rates were below 0.2 G/s.

Preparations and modifications:

• –all threshold settings included,
• connect BLM system to BIC,
• work on SEMs calibrations and channel allocation,
• application SW, - ergonomics - freeze on acquire
• capture data in concentrator, …

Response of BLMs to losses on collimators calibrated 

Aperture measurements in arc - Stefano Redaelli

• “TL” method of free oscillations to check overall  clear aperture.  About 4 hours available.
• H +/-18-20 mm in arc, losing in Q6.L3.
• V at least +/-10 mm, again found bottleneck in Q8 and Q7 to about 10 sigma. Did not go to aperture limit in the arc.

Collimators - Stefano

No problems with collimators

• made calibration scans of beam losses on the different devices
• found one small issue with TDI jaw losing steps
• mechanical problem found and now fixed...

BPMs - Lars Jensen

BPMs worked first shot - used asynch mode (autotriggered – no fine timing needed)

• Trajectory acquisition looked fine.
• Publication to LSA concentrator fine
• Also tested capture data – fairly complex setup for injection using prepulse over BST. Worked fine
• – jitter below 0.25 ns
• - some issues with the timing of the injected event to look at and maybe some issue with concentrator.
• Working on algorithm for setting up of expert settings for capture.
• Tested XPOC (PM) – worked fine – small issue with alignment of turns BPM calibration factors good – to 15-20%
• Intensity range fine – should work down to 1.5e9 with 0.2 – 0.3 mm noise. Will be sensitivity change at 4e10 charges
• A few probs – 1 BPM not working, 1 polarity error, 1 cable inversion (?) BPM intensity modules less sensitive than expected – non-linearity to be corrected with polynomial – should be OK to 5% when fixed.

BCTs End TI2 - worked fine once threshold for triggering was lowered. Difficulty with seeing analog signal for kicker timing in – will be improved

Screens Worked fine – big degradation of signals in injection region when losing on Q5 – but no permanent effects seen yet.

SW issues for BI - JJ Gras

• Stability of servers used was an issue – SR2 and SR3 BPM crates crashing. Still under investigation
• BLM threshold handling not fully operational – being worked on.
• Some concerns about new concentrators using XTIM and GET methods rather than subscription and time response of published data – being addressed with clients.
• Need some RBAC deployment to limit access to some data at low level.
• DIAMON monitoring didn’t see the server problems. Still missing or to be tested
• Full and reliable implementation of BLM logging (OK when limited to few crates for the concentrator).
• RBAC deployment
• BLM thresholds via MCS
• full implementation of BLM/BPM capture data
• PM data. To do next time -
• Threading - Worked pretty well!!!

Could we try 2 bunches (separated by 525 ns.) Would involve a different SPS cycle - to be discussed.

Control system - Eugenia Hatziangeli

• Slow timing (affecting prepulses and injection/RF synch). Quick fix available for b2 injection test – 2 weeks for full fix. Test before 10/09
• FESA server crashes – being worked on
• XTIM – unexpected crashes after major fix a few weeks ago – auto reboot a possibility for critical periods. (From Eugenia after the meeting Since Thursday evening the XTIM is automatically restarted in case of a crash to make operations easier. The time of each restart is being logged. In the meanwhile, both Julian and Maciej have collected a large amount of debug XTIM information and they are trying to identify the cause of the crashes.
• SDDS logging disks getting full – added ~50 Gb over the w/e – need more space and cleanup urgently.
• JMS traffic much higher than expected – try restricting access to key applications with RBAC for 22/08.
• Need to test RBAC - phase in gradually - do not disturb the injection test please.