Tuesday June 19th,
14:00
CCC conference room
874/1-011
There were no comments on the minutes of the 27th meeting.
Roger described the updated LHC hardware commissioning schedule which had been presented at the 21st meeting of the LHC MAC. There is no longer any sector test foreseen. Frank remarked that the MAC might recommend that we be nevertheless prepared for a sector test in case of further delays in the beam commissioning. Oliver commented that any sector test appeared increasingly unlikely, and that the AB department had already decided to abandon any further preparation.
Jean-Jacques’ comprehensive overview of BI readiness first surveyed the overall state and then elaborated some details.
The phase 1 of the LHC beam instrumentation consists of 1000 BPMs, 4000 BLMs, beam intensity monitors, transverse profile monitors (synchrotron radiation, wire scanners, ionization gas monitors), tune and chromaticity diagnostics (BPLs for BBQ tune measurements, BQK tune kickers, Schottky monitors), and BRA luminosity monitors. Jean-Jacques displayed a list of the respective project leaders, the persons in charge of software and electronics, and the commissioners for each instrument. Discussions on software requirements are ongoing between BI and the commissioners.
Schematic views of each LHC sections, prepared by K. Foraz, describe the status of each instrument installed in the machine. The data are continually updated, at the link file:\\cern.ch\dfs\Divisions\EST\Groups\IC\LC\Public
(file
LSS_Schematic_View_Vxxx.xls).
So far the BI installation proceeds according to schedule, except for two ionization-profile monitors (BGIs) installed in LSSR4 which have been accidentally burnt during bake out. One of these detectors may be repaired for the LHC start-up. The plane in which it measures can be chosen.
Roger asked whether the BGIs are intended for ions or protons. Jean-Jacques and Oliver responded that these monitors are employed also for protons, and that they may be useful in particular for the early commissioning. Ralph S. pointed out that the same bake-out mistake had been made on some BPMs used for the tune measurements. Luckily, these BPMs require only cleaning to be again operational. Ralph S. added after the meeting that the BPMs were cleaned and they seem to be OK.
Searching for an alternative to the missing BGIs, Frank asked whether the synchrotron light monitors will not be available for the start up. Oliver replied that the synchrotron light monitor can be used only above about 2 TeV, without the undulator. At an intermediate beam energy we may observe the transition between undulator and dipole radiation. Oliver inquired the energy up to which wire scanners can be used. Jean-Jacques responded that wire scans may be performed at any beam energy, but that the beam intensity is much more limited at 7 TeV.
Jean-Jacques summarized the overall state of the BI hardware commissioning. He presented some details for how the properties for each individual BPM are documented and identified via the slot position. BI is in the process of assembling this information for the start up, with the aim to keep traces of all values and parameters that will be needed for operation.
Oliver asked whether the BBQ system requires time for hardware commissioning. Jean-Jacques answered that this would be discussed later in his presentation among the details. Jean-Jacques then answered affirmatively to Oliver’s next question whether radioactive probes are used for the calibration of the BLMs. Indeed the radioactive probes are scheduled as the last step of the BLM calibration.
Now Jean-Jacques turned to the software side, where the coordination is done in collaboration with the instrument commissioners and with the LHCCWG scenario builders. Operational interfaces have already been defined for the beam position, the beam loss, and beam profiles. Work is in progress for intensity, and Q/Q’. Oliver asked whether coupling is also included. Ralph S. responded that, yes, the latter was automatically taken into account.
The BI documentation is well advanced. Web pages describe the properties, features, people, standard usage, device details, software interface, etc. AB/BI/SW deploys servers on front end computers (FECs) emulating these interfaces. Already some BPM and BLM FECs are running. From the beginning of July, FECs will also run the software interfaces for BWS and BSRT. The interfaces for the remaining instruments will follow during the summer. Timing and logging requests will be transmitted to CO.
Now addressing the commissioning with beam, Jean-Jacques reported that the procedures for the BLMs are developed by the MP Commissioning WG. The beam commissioning for all other instruments is defined by BI. The work will start this summer and could be presented to the LHCCWG during this autumn.
In the second part of his talk, Jean-Jacques discussed various details.
First, for the beam position monitors, the monitor hardware, electronics, front-end software will all be ready for the start up.
Oliver asked whether the sum signal mode was also included. Jean-Jacques replied that for intensity measurements only expert software would be available, but no application for users. This expert software is under development.
=> ACTION: BPM sum signal mode ready for start up? (Jean-Jacques)
Five types of BPM acquisition can be distinguished: (1) threading – asynchronous acquisition, which is OK for a few turns; we can at least see where the beam is; once the beam is captured we have access to the other modes; (2) orbit [initially: 1 Hz, later: 10 (25) Hz for orbit/energy feedback]; (3) capture, (4) XPOC, and (5) Post Mortem Buffers. Jean-Jacques explained that BI will try to get information on the revolution frequency in threading mode to be sure about the number of bunches. Operation with a single bunch will avoid any confusion. The capture mode samples the beam position for a selected number of bunches over a selected number of turns. XPOC will obtain the bunch positions of the last turns prior to each beam dump event. The Post Mortem (PM) Buffers will be sent on PM events via the CO PM Channel.
Gianluigi asked whether fine delays needed to be tuned for identifying the individual bunches. Jean-Jacques replied that, yes, the abort gap would be used for identifying the first bunch. The HW fine delays tuning is done as soon as the beam is captured.
Gianluigi next inquired if the threading mode will also provide the average orbit over a number of turns. Jean-Jacques explained that it is not foreseen to implement the full DAB acquisition functionality for initial operation.
=> ACTION: Average orbit available in threading mode? (Jean-Jacques)
Remaining issues for the BPM FE software include the handling of calibration settings, the swapping of trigger sensitivity (200 ms needed for the swap), impact of capture data request on the real-time (RT) data transmission, and the development of XPOC position interlock in point 6. Work on all these items is in progress.
Roger commented that sensitivity may be an issue when we inject on top of a circulating pilot. Jean-Jacques remarked that the triggers are changd via the worldfip.
Operational aspects comprise a full scale test for the concentrators, logging, etc. For the RT feedback the BI-OP proposal is to start with 1 Hz and then to later increase the rate to 10 Hz. RT loop control includes, e.g. updating the feedback matrix. Other items are fixed display definition and ad-hoc developments.
After the meeting Massimo recommended addressing the issue of communication of the measured positioning (geometrical measurements of the magnets) and the required misalignment (to optimize the aperture) from the AT/MEB activities to the BI experts.
=> Action: Communicate AT/MEB results on magnet geometries and voluntary misalignments to BI (Luca?)
Jean-Jacques next commented on the loss measurements. The BLM monitor hardware, electronics and software are all ready for the start up. The BLMs have 4+1 acquisition modes: (1) maximum losses [12 time periods, maximum since the last acquisition, at 1 Hz]; (2) study data [on request, data points are acquired every 40 microseconds covering a span 1.3 s; reading out the buffer]; (3) XPOC [40 microsecond granularity, from 100 ms before dump, until 40 ms after dump]; (4) Post Mortem Buffers sent on PM requests via the CO PM channel. The 5th BLM mode refers to collimation data. Loss data are fed into the collimation control system to ensure that the losses do not exceed preset limits during collimator jaw motion. This mode uses a dedicated interface, triggered by the collimation system, and with data being displayed via the collimator controls applications.
Ralph asked about the availability of beam-loss data integrated over one second. Jean-Jacques replied that the loss rate integrated over 1.3 s is provided at 1 Hz. The data comes either in the form of a sliding window or as a maximum value depending on the selected time period.
Remaining issues for the BLMs include threshold handling (database, tools, procedures; all under discussion), the impact of MD data requests on the RT data transmission, the HW threshold evaluation for which work is in progress by BI, full scale tests, fixed display definition and ad-hoc developments.
The transverse profile monitors, their electronics and their software will be ready for the startup with the agreed interfaces. It will be possible to scan two beams at the same time, but not the two planes of the same beam. Two wire scanners exist per plane and per beam, one operational and one spare
Oliver repeated his earlier question up to which energy the WS can be used. Jean-Jacques replied that the intensity limitation was being studied in Bernd’s section. Tight limits are set by the possible quench of Q5 while scanning. The limit is estimated to lie between 1 and 6e12 protons. The wire damage limit seems to be close to the quench limit. Therefore, at 7 TeV the wire scanner usage should be limited to 1 nominal bunch. In any case the wire scanners (BWS) are meant to be calibration devices for BSRT and BGI. If we are lucky Q5 quenches will occur before damaging the wires. The limit for 450 GeV is under study. According to the original specification we should be able to scan 1 PS batch (72 bunches) at 450 GeV.
Gianluigi asked about the speed of the wire scanners. Jean-Jacques replied that for the moment the speed was 1 m / second. Gianluigi wanted to know whether there is a limit on the BGI from quenches, similar to the wire-scan limit. Jean-Jacques replied that no such limit had been mentioned so far. He will check this question with the experts.
=> ACTION: Intensity limit on BGI from quenches? (Jean-Jacques)
Ralph suggested relocating the wire scanner to the cleaning insertion to avoid quenches. Gianluigi commented that the cleaning insertion was going to be a highly radioactive area. However, Ralph clarified that the entrance of the insertion was not expected to become terribly radioactive.
The BGI monitors for beam 2 were destroyed. Those for beam 1 are still OK. Oliver asked whether the beam-1 monitors had already been baked out. The answer was not known. Possibly one beam-1 monitor could be used for beam 2. The BGIs have 1 continuous acquisition mode, at 1 Hz, providing the beam profile, beam size and position. One image will be displayed derived from an integration over 20 ms.
Oliver inquired why a gas injection was necessary, recalling that the ionization monitors at HERA worked well without additional gas and delivered superb signals. Jean-Jacques responded that in LHC the vacuum was expected to be much better and that also the revolution time was lower. If gas injection turned out not to be needed, it would be a great relief.
Another issue is the effect of the magnetic field of te BGIs. Some encouraging feedback on this point had been received from Jorg.
Replying
to a question by Oliver whether there was a plan to change the B field,
Jean-Jacques explained that the BGI will operate at constant maximum strength.
A correction of orbit effects during the ramp must then be considered. Gianluigi proposed using several dipoles as in the SPS
e-cloud tests. Jean-Jacques responded that two dipoles, not three, are used,
fed by a standard power supply in parallel, not in series.
=> ACTION: Check residual effect of BGI magnetic field with experts (Jean-Jacques)
The synchrotron light monitors feature 3 acquisition modes. (1) A non-gated full image is recorded at 1 Hz. (2) The gated physics mode provides signals for all bunches integrated over 1 turn, updated every minute for 3564 buckets. The modes are mutually exclusive. (3) The MD mode acquires data for the gated part of the beam, down to a single bunch, over several and up to 3000 consecutive turns.
Issues for the synchrotron-light monitors include the ethernet link towards the fast camera, where a dedicated CPU is used in order not to “pollute” other networks, the implementation of automatic gain setting for different beam energies, the HW commissioning of the light sources, the signal calibration with beam at different energy and intensity, and the reliability of the fast camera in the tunnel conditions.
Oliver asked how much time would be needed for the calibration with beam. Jean-Jacques replied that this question will be discussed over the summer. Oliver then wanted to know the sensitivity to the orbit. He stressed that a detailed discussion was needed on what needed to be done with beam. Another example he mentioned was the ramp of the undulator. Jean-Jacques agreed that the switch between the two sources of synchrotron light needed to be checked. Gianluigi inquired the methodology for profile measurements, involving wire scanners, BGI, and SR monitors. Jean-Jacques commented that the most accurate calibration will be achieved with a single bunch. For the synchrotron light monitors, the energy dependence will be a concern. Gianluigi recalled heat deformations of the optical mirrors observed in LEP. Jean-Jacques assured the team that no such deformations were expected at the LHC.
After the meeting Laurette commented that it is planned to operate the undulator always at its maximum field, even at injection. The ramp concerns only the D3 magnet. If the undulator field were lowered at injection, the emitted light intensity would probably not suffice for a satisfactory calibration with single bunches.
As for the intensity measurement, two BCTDC monitors will be installed per beam, one operational, and one spare. The BCTDCs have 3+1 acquisitions modes: (1) injection, (2) standard [1 Hz], (3) safe beam flag [10 Hz, link to MP], and (4) post mortem [on PM requests]. A first measurement is performed before and after every injection. The remaining BCTDC issues include the absence of automatic arbitration between the two redundant monitors for the start up.
Oliver asked for the performance specification. Also for this instrument he pointed out that a list of detailed commissioning steps was needed. Mike commented that 1% accuracy had been asked for. Jean-Jacques replied that the detailed steps will be documented together with commissioners during the coming months. Replying to a question by Gianluigi on the repair and access needs, Jean-Jacques explained that the BCTDCs were easy to exchange. He recalled a discussion with MP, where no clear result was obtained, except that MP did not want to receive two signals. Jean-Jacques remarked that the BCTs are solid devices in principle, except for calibration issues.
=> ACTION: Performance specifications for BCTDC. Detailed list of what needs to be done for the instrument commissioning, both for the BCTDC and for other devices (Jean-Jacques)
Verena asked if only one BCT was used for creating the safe beam flag. Jean-Jacques confirmed that this was the case, saying that the safe beam flag had not been considered critical. Jan pointed out that now it had become critical. Ralph S. and Jean-Jacques both commented that without signal from the BCT the LHC was automatically safe. Jan cautioned that different failure modes needed to be considered.
Jean-Jacques remarked that information could be sent from both devices if there were such request, and that it would be no problem for BI to implement any proposal.
Ralph S. commented that unless two completely different devices were available, the redundancy of two structurally identical intensity monitor was not useful due to the intrinsic common-mode failures and ambiguity in case of a single device failure. Jan commented that this was to be checked with Bruno. Jean-Jacques said if the second device was not needed, he would prefer using it for BI instrument development. There is one crate with signals from the two beams, and one duplicate.
After the meeting Rudiger commented he did not recall any discussion that the interlock system did not want to receive two signals. On the contrary, if two signals were available, they would certainly be used, since the use of only one BCT might limit the safety level of the Safe Beam Flag.
Reyes asked if the lifetime was published at 1 Hz. Jean-Jacques responded that, yes, but probably it would be computed over a longer time, since otherwise one could not resolve long beam lifetimes. The BI group has experience with algorithms developed for LEP. In any case, all data are provided to the users, who could calculate the lifetime according their own favorite algorithm if they so desired.
Jean-Jacques now provided information on the fast BCTs (BCTF), their status and acquisition modes.
Ralph commented that the beam lifetime was very important, as the collimators were designed to withstand a 0.2 h lifetime over 10 seconds per specification. Therefore, the beam must be dumped after 10 s, if the lifetime drops below 0.2 h. Oliver inquired whether this lifetime drop was not also seen on the BLMs. Ralph replied, that, yes, in principle, but that the collimator design was not prepared for this modus of operation. He added that the beam dump was triggered by a software interlock, and that there were no temperature sensors.
Jean-Jacques asked Ralph to send him the specifications. Roland commented that this question may not be relevant for phase 1, only for higher intensity. Ralph remarked that it would not hurt to have this interlock in place right from the start. Jean-Jacques agreed that he would explore if the BCTF resolution was sufficient to do this calculation in real time.
=> ACTION: Send specification of beam-lifetime signal for collimation to Jean-Jacques (Ralph). Check if BCTF resolution is sufficient to do the lifetime calculation for collimation in real time (Jean-Jacques)
Another derived signal is the dI/dt monitoring for MP purposes. The BI objective is to have a prototype by summer 2008. The dI/dt is obtained after splitting signals from fast BCTs.
Frank asked whether this dI/dt signal could not be used for collimator protection. Jean-Jacques answered that the dI./dt measured only really fast losses, which were detected over a few turns, not slow losses.
Another device measuring intensity is the BSRA abort gap monitor. Present hardware is used just for monitoring, and at the moment no beam dump or other action is planned to be triggered from the corresponding FEC. The development of the operational application is to be started.
After the meeting, Brennan pointed out that the abort gap monitor must work at 450 GeV and that it uses light from the BRSA, which will need the wigglers below 2 TeV. So the latter must be commissioned before the abort gap monitor can work at injection.
Ralph commented, however, that the beam present in the abort gap is relevant only for avoiding quenches. There could never be any damage. Mike and Verena remarked that a hardware signal was foreseen originally. Verena elaborated that according to the specification the signal should be hardwired and generated below the quench limit of Q4.
=> ACTION: Hardware signal triggering beam dump below Q4 quench limit? (Jean-Jacques)
Post-meeting, Brennan added a qualification to Ralph’s comment that the population in the abort gap is of importance solely for quench protection. Namely this statement is true only if the TCDQ is correctly positioned with respect to the beam - if not, a substantial abort gap population may damage the TCT collimators, which have a very low damage threshold. So in a sense we would obtain some redundancy if the BSRA abort gap monitor would be part of the interlock chain. This issue (i.e. the action listed for Jean-Jacques) will be discussed at a forthcoming MPWG meeting.
The discussion now turned once more to intensity measurements via the BPMs. Jean-Jacques explained that the development of monitor electronics and software are in progress and will be ready for the LHC start up. He recalled that we will use the ac channel of beam 2 to read the beam 1 intensity and vice versa, and therefore we will lose the position information for the other beam when this mode of operation is chosen. This mode will help to identify and localize unexpected obstructions. The BPM intensity monitoring will remain an expert intervention. No operational application is foreseen.
Roger pointed out that the BPM intensity mode is not a back up, but that we would like to start the LHC in this mode. Jean-Jacques responded that this BPM intensity would not be available for a sector test in December. Oliver commented that the sector test would not be available either.
Verena asked about the beam presence flag, which needs to
be set 1 ms before the next injection. Jean-Jacques replied that the beam
presence flag is based on the fast BCT signal, with good resolution for pilot
bunches. A hardware comparator is employed. The signal is transmitted to
Now Jean-Jacques addressed the tune and chromaticity measurements. The necessary monitors, electronics, and software will all be ready for the start up. It is planned to start with a single-kick method and with single acquisitions, returning raw data, FFT and peak detection. Next the PLL continuous measurement will be commissioned. Oliver asked whether the PLL commissioning can be done in parallel, which Ralph S answered in the affirmative. The latter pointed out that if the transverse damper is unavailable during initial LHC commissioning, there will be no tune, coupling or Q' tracking unless the tune-tickler which was initially intended mainly for nominal operation is boot-strapped. Part of the resources may need to be allocated to prepare this 'tickler'.
=> ACTION: Tickler for tune and chromaticity measurements for initial operation? (LHC-CWG)
Jean-Jacques pointed out MDs will be needed to set up tune, coupling and Q’ tracking with damper excitation. Oliver stressed that we must justify the dedicated beam time for LHC MDs in view of the tight schedule. It should be specified how many MDs are required and the detailed commissioning steps must be provided.
=> ACTION: Tickler for tune and chromaticity measurements in nominal operation? (Jean-Jacques)
=> ACTION: Justification of dedicated beam time in view of tight schedule; specify number of MDs and details of commissioning steps (Jean-Jacques)
Jean-Jacques described that the tune can be measured via the BBQ or via the Schottky monitors [and during injection or in combination with large kicks also using the head-tail monitor or LHC BPMs]. The Schottky monitors were meant to come later, in phase II, but they are already installed and will deploy a similar (the same?) FESA interface as for the BBQ. Prototype applications exist and are already in operation at the SPS. The present LHC base-line Q' measurement relies on radial momentum modulation plus continuous tracking and on-demand measurements of the induced tune changes.
Finally, Jean-Jacques turned to the luminosity monitors BRANA (at IP1 and 5, built by LBNL) and BRANB (at IP2 and 8, built by CERN). The total luminosity will be recorded at 1 Hz, the bunch-to-bunch luminosity at 0.1 Hz.
Oliver asked whether these monitors will detect collisions even without prior beam commissioning and calibration. Jean-Jacques answered yes. Ralph. S. commented that this question had been discussed in a previous meeting of the working group, where the beam rest-gas interaction rate had been estimated to be about the same as the rate from collisions. Mike added that especially at 450 GeV observing luminosity signals would be extremely difficult. Alick commented that the experiments could also give some collision feedback already on day 1. Jean-Jacques asked if the conditions for useful signals could be specified. Once these are known, he would contact Enrico.
=> ACTION: Specify conditions for useful signals from luminosity monitors (Helmut?)
Concerning a hypothetical sector test, new enlarged BPMs (BPMWI) will be available soon. Jan and Roger mentioned that beam tests with these BPMs could be done before the end of October and that they do not require the sector test.
=> ACTION: Prepare beam tests with new enlarged BPMWI (Jean-Jacques, Jan)
Reyes, Oliver and Mike clarified and summarized that beams will be collided at both energies. The monitoring situation appears OK at 7 TeV, but indeed it looks critical at 450 GeV. Discussions with the experiments for additional signals are underway. It is planned to obtain an ad-hoc luminosity signal, which may differ from the final one.
=> ACTION: Follow up ad-hoc signal from experiments for collision optimization (Alick, Reyes, Mike, Jean-Jacques?)
Oliver asked about the wall current monitor from the rf group and whether its signal would also be available in the control room.
=> ACTION: Clarify status of wall-current monitor (Jean-Jacques, rf group)
Jean-Jacques concluded his presentation by stating that the targeted performances of all instruments are those from the functional specifications. Commissioning procedures and MDs with stable beam are needed to reach this performance. He will discuss and document the needed procedures during the coming months and come back to discuss how the result can be fitted into the two months allocated to beam commissioning.
Jan’s presentation was cancelled due to lack of time. His slides presented to the 21st meeting of the LHC MAC are available on the web.
Roger announced a summer recess for the LHCCWG. The LHCCWG meetings will resume in the middle of August. Gianluigi asked about the sector test preparation. Jan replied that this had been put on hold for the moment, but that the preparation had been fairly well advanced.
The date of the next meeting will be announced in due time.
Reported by Frank