MP3 meeting

Wednesday 8 Jul 2009, 09:00 → 10:30 Europe/Zurich

30/7-012 (CERN)

Jean Philipppe Tock (CERN)

Slides

• S23_recommissioning.pdf
• S23_recommissioning.pptx
• The_Pulse_of_the_Silent_Thriller.pdf
• The_Pulse_of_the_Silent_Thriller.ppt

MP3 meeting 8 July 2009

Matters arising

A number of MP3 steps signatures are still pending from the recent campaign of commissioning in sector 23. One of the issues concerns the PCS tool, which is now able to use U_DIFF as well as U_RES to measure the splices.
The default option in the tool is still U_RES, the other signal is more noisy and the resulting statistical error on the resistance is higher. Moreover the number of points recorded is in some cases much lower, and insufficient, for U_DIFF.
If the U_DIFF signal is to be used, the tolerances on the voltage standard deviation will have to be relaxed

A proposal to pulse the magnet buses to validate the splice quality (Howard Pfeffer)

Howard presented the idea of using high current pulses through the buses to test the connections. The proposal is spelled out very clearly in the annexed slides
The presentation triggered several comments and questions
Rob Van Weelderen noticed that the method is very sensitive to the time constant for heat diffusion through the copper: this being very fast (as confirmed by Arjan Verwej), an overheated connection might have recovered by the time of the measurement after the pulse.
Arjan stressed that the resistance measurements at 100 A before and after the pulse must be accurate enough to allow deciding whether a thermal runaway is close to happen
Karl Hubert Mess noticed that the diodes will not all behave the same way, which might be a problem for the splice protection
Nuria asked what was the uncertainty on the stabilizer resistances in the quadrupole circuits measured at warm
Howard said at least 60 micro Ohm.
Ezio asked how long the proposed measurements would take
Ruediger commented that this technique is not envisaged for the first run with beam, but rather to assess the energy reach of the machine, relative to splice stability, after the first physics run. He said the measurements would take several weeks.
Karl Hubert underlined that this technique only detects the worst connection, once this is detected the sector must be warmed up to be repaired.
Arjan said that the measurement should be seen as a last commissioning step to qualify the machine for a given energy, once the value is decided by the management.
Walter said that the current pulse would put a lot of power over a wide frequency band, and inquired on the sensitivity of the exposed electronics, for example the QPS.
Knud answered that the real pulses will be tested in the lab to avoid breaking anything
Nuria asked whether the current leads could take the load
Bob answered that this had been checked with Amalia and should pose no problem as the DFB can be cooled separately.
Ruediger said that the pulses would put a noticeable heat load on the cryogenics system
Karl Hubert asked whether an ECR was needed before putting two RB converters in series
Ruediger answered certainly yes, but we are not yet at this stage

Status of Non Conformities (Michele Modena)

Michele informed that actions are ongoing to cleat MTF from not closed steps with the aim to leave only really important NC. Subsequently the plan is to classify the non conformities to ease the queries

ELQA and commissioning (Matteo Solfaroli Camillocci)

Matteo submitted the following problem: if in sector 23, which is brought at 80 K, the ELQA needs to be repeated, the cables must be disconnected. This would mean that the circuits cannot be considered commissioned any more.
Matteo proposed to skip some steps on different types of circuits in case of recommissionig. (see slides)
Ruediger asked what is the critical temperature above which ELQA must be re done.
A discussion followed, the main points are the following:
there is nothing special about 80K, it is just the value at which the temperature can be stabilized using the helium flow coming from the liquid nitrogen heat exchagers. Most of the thermal expansions happen above 80 K but this is not a sharp borderline. Probably 90 K is still OK, this would need be assessed based on the mechanical tolerances.
Giorgio stressed that when the arc is at 80 K the DS might be above 100 K. He favored repeating the ELQA.
Nuria concluded that the decision will be made per circuit type based on the risk of having to warm up if a problem is detected at cold.
Markus Zerlauth said that if the test plan has to be redone in LSA, he will need the list of steps for each circuit
Matteo will provide that

AOB

Christine Darve reported on the proposed tests on the MCBX magnets powered from the DFBX:
Following the completion of the DFBX leads HC powering phases, they will power several RCBX circuits in order to determine the hard limits of their hydraulic and QPS systems (up to two days of powering test needed).
The main driver of this test is to reduce the beam collapsing time, presently specified @ 40 sec. The motivation of this test was presented by Christine and agreed during the LMC meeting hold on March 25 2009.
Last year DFBX lead stabilization measurements validated possible 600 A lead operating ranges. Still, these test results don't comply with new Physics request and faster ramp and acceleration rates must be investigated.
Mirko pointed out that the maximum current during the test should be raised to 550 A, instead of the 350 A mentioned.
Walter proposed to revisit the option of using the lead voltage (and current) as a process variables, reminding that this would be technically better and had been put in standby last year for lack of resources. He remarked that this should be balanced with the resources being now used to tune up the currently implemented solution.
Karl Hubert said that the solution with the TT was considered the "quick and dirty" fix.
Action --> Nuria will bring up the matter at the TEMB

Knud informed on two proposals:
1) Installing a snubber capacitor across the dump resistor to prevent triggering the symmetric quench detector with the travelling waves that propagate in the line upon switch opening
2) Further reduce the RQF/D time constants, the idea of firing the heaters to react to a splice runaway being judged perilous.

A. Rijllaart did not agree with the last point. After the meeting he explained his position as follows:
<< I am in favour of a study of the possibility of firing all heaters in case of a splice runaway. Such a system, already partly in place during 2008 (firing 1/3 of the heaters in a sector) is interesting, not only for splice runaway, but also for no switch opening in the event of a magnet quench (for this 1/3 is enough, all heaters are better for splice runaway).
Although the system accidentally fired once, it was still in a very early state, and I think it can be made more secure. We have a better understanding of the noise and disturbances now (switch opening, cryo valves, ...). We believe we can reliably operate splice detectors at a 300 microvolt threshold with a 100 ms window. Then we should also be able to reliably operate the sum of them. The threshold and time window can be set higher, we don't need the same reaction speed.>>