MP3 meeting

Wednesday 30 Sept 2009, 09:00 → 10:30 Europe/Zurich

30/7-012 (CERN)

Nuria Catalan Lasheras

Slides

• 20090928_CalorimetryRunPlots.pdf
• ELQA_MP3_30-09-2009.pdf
• ELQA_MP3_30-09-2009.pptx
• mp3_090930_zch_01.pdf
• mp3_090930_zch_01.pptx

MP3 meeting 30 September 2009

Splice mapping of the main circuits in sector 12 (Zinur Charifoulline)

Zinur presented the results of the PLI2.s1 step performed last Friday on RB and RQF/D circuits in sector 12 (see slides)
Once normalized to the number of splices, the measurements feature a nice distribution with about 300 pOhm average and about 100 pOhm sigma for both circuits. The presentation triggered a few questions and comments.
Sandor commented that the distribution should be skewed towards the upper end, as there is a lower bound to the possible resistance values
Walter asked whether the figure of 200 pOhm for the lower bound was correct
Francesco confirmed that this corresponds to the best splices that were done
Christian Scheuerlein said that in FRESCA they had found a somewhat higher value between 200 and 300 pOhm
Jean Philippe Tock asked whether the "reduced pyramid" with less current plateaux was enough
Zinur said the resolution was sufficient
Ruediger agreed the reduced test was adequate and proposed to change the specification
Action-> Authors of the 13 kA circuits specs
Francesco asked what are the error bars on the splice mapping results
Zinur clarified that these indicate the regression of the linar fits of the current voltage curves
Walter remarked that the results of sector 12 indicate that any splice with a few nOhm is (in some unknown way) different from all the others, and asked whether we should revisit the tolerance of 10 nOhm.
Francesco cautioned that in sector 12 many splices had been redone, and suggested to wait for the results of sector 78 for a realistic picture of the machine as built
Knud pointed out that in the coming QPS IST campaign the heaters will be fired twice, to check the two independent triggering channels (old and new QPS)
After some discussion the present agreed that this is a necessary ill.
Jean Philippe, in view of the results, proposed to give green light to the next step at higher current
Walter said that the interlock was not yet connected and asked if the reaction of the nQPS to the switch opening had been checked
Knud said the test was done but the PM data were still being collected

Status of IPQ splice mapping (R. Mompo)

Richard summarized the situation with the system setup by the ELQA team to measure the splices (see slides). Most software problems have been solved.
The most outstanding issue is now the level of noise ranging from 1 mv to 20 mV.
This must be greatly reduced in order to reach nOhm resolutions.
Hugues asked about the time structure of the noise.
He recalled that the converters generate 1 MHz noise which can be easily filtered out
Adriaan agreed, saying the the system has too high a bandwidth
Knud asked whether the noise floor with the converter off had been measured
Richard answered yes, and the level is the one presented
Knud then commented that the noise is likely to be low frequency and coming from the environment, which is more difficult to filter
Walter commented that even if the noise was purely white it would take hours of averaging to reach the required resolution
Zinur asked how far this system was from the main busbars, and proposed to correlate with the noise on the nQPS
Ruediger recalled that last year the results (at higher currents) were better
Andrzej noted that the same people were involved at the time
Jean Philippe summarized the situation noting that there has been progress, however it seems that a lot of work is still needed to achive results. Therefore a decision is needed on how to go ahead. He asked wheter there was any other way at hand to measure the splices
Ruediger said that one option would be not to commission these circuits at 5 TeV, to reduce the risk. The new I_PNO would be 3.5/5*(old I_PNO) rounded up to the nearest 100 A
Andrzej commented that the study should be finished, but he agreed to go for a 3.5 TeV commissioning on the other circuits,
thus skipping the systematic splice mapping
Outside the meeting: the chosen way to trace this is changing the specificities documents in edms, reducing I_PNO in a new version
Ruediger said that in late spring the splices may be measured before rising the energy of the machine, possibly with different techniques
Knud confirmed that a QPS based system should be available by then
Gianluigi asked if the circuit will need to be recommissioned after the disconnection of the measuring system
the present agreed that at least some checks had to be done

Concerning the similar issue of the IPDs: D2 D3 D4 were covered by QPS logging, while it was not clear what to do with D1.
Jim and Sandor said that the busbar design is such that any splice quenching would propagate to the magnet fast enough and therefore the system is safe.
Sandor will double check the situation for D1 and present it to MP3 next week
It was already that D1 and MQX will be commissioned to 5 TeV
Andrzej commented that the busbar layout for the IPQ is more complex, so we are not sure that the situation is similar to the triplet

Calorimetry in the inner triplet L2 (Jim Strait)

Jim presented the results of a calorimetry test performed in IT.L2 (see slides). The temperature had a negative slope before the test. The slope was progressively curbed upwards in correspondance with the increasing current plateaux
THis is qualitatively in agreement with expectations
However when trying, with different approaches, to work out the effective splice resistances, values of one order of magnitude higher than expected are found
The issue seem to be the cryogenic conditions, which are not enough stable
Jim found a couple of valves (thermal screen and LHe filling) which moved during the test. He will further discuss with the colleagues of the cryogenics to see if the situation can be improved.
However he recalled that similar results had been found last year in the DS an MS, and although not fully understood, were attributed to end effects
In this respect, the results of the IT, which is "all ends" should not be a surprise
The present agreed with Jim's remark that, while it is interesting to spend some more time to understand the calorimetry of the IT, it is not worth continuing if the cryogenic conditions cannot be made stable enough.
The authorisation to skip the splice mapping is supported by:
- the fact that the triplets magnets have been commissioned to 7 TeV last year:
- the fact that in case of a splice quench, it will propagate, will be detected and the energy quickly extracted so no risk is taken.:

AOB

Sandor showed the IT connector on which he proposes to intervene. The time estimate for the intervention is 1 day for the work and 1 day and 1/2 for subsequent requalifications.
The question whether it is also necessary to empty the triplet was also asked
This will have to be clarified by the safety commission
Ruediger said that if this needs to be done, there will have to be a responsible person and a procedure Sandor will write the procedure, while Andrzej, Giorgio and Richard will organize the intervention
MP3 recommended to go ahead, also to comply with the ALARA principle

Jean Philippe Tock illustrated the excel sheet which will be used to track and follow up problematic circuits. ,
the sheet is on the HCC pages (http://hcc.web.cern.ch/HCC/meetings_schedules/?version=2009), and will be maintained by Nuria, Jean Philippe and Walter.