Choose timezone
Your profile timezone:
MPP meeting
MPP meeting 16 January 2008
Extension of CRYO_START and CRYO_MAINTAIN logic for 120 A dipole orbit corrector circuits. M. Zerlauth.
Similarly to what was already observed for 60 A corrector circuits, some conditions necessary for safely run the 120 A corrector circuits are missing from the CRYO_Maintain and CRYO_start. An ECR has been circulated that proposes to prepare a loop from the 120 A cryogenic conditions and include it into the CRYO_Start and CRYO_maintain signals (can be an exact copy of the 60A according to A. Ballarino). The disadvantage of this first proposal is that, if the conditions for the 120A circuits are not fulfilled, it prevents the powering of any circuit in the same powering sector. Problems with the cooling of the beam screen can still be expected during hardware commissioning. S. Claudet added that the conditions for CRYO_Start could be simulated but not those for CRYO_Maintain.
A second proposal consist on create independent signals for the CRYO_Start and CRYO_Maintain of the 120 A circuits exclusively. This will keep the flexibility during hardware commissioning and both permits will be merged at the end of commissioning. The implementation however of these new signals has very heavy implications in AB/CO.
An intermediate solution would be to prepare the loops and commissioning them independently. They would be linked to the CRYO_Start and CRYO_Maintain signal circuit by circuit only after PCC tests (10 A) have been done. This will leave powering of 120 A circuits after all other circuits in the powering sector problems are solved.
MPP supports the third option.The powering procedures should be changed accordingly. A signal or a flag that the activation is done could be visible somehow from the CCC. The ECR will be modified in this way and released. (Action: M. Zerlauth)
Problems of level gauges in the stand alone magnets
Loss of the level signal in the matching section magnets happened already in sector 7-8 and is now preventing powering in sector 4-5. Typically, a pressure reduction is seen in the line D (~8 min) together with loss of level in the magnet (~3 min). The nominal heat load in the LT enclosure is estimated to 0.31 W which taking into account the impedance of the capillary, will create a pressure increase and lower the level in 10 mm or 2.5%. The heat in-leak necessary to lower the level by 75% is of 2.2 W. The possible source of such heat in-leak is unexplained. Besides, the behavior of the level gauge during the boil-off of magnet Q5L5 also shows regular steps that are not well understood. This increases the reading error from 2.5% to 10%.
The origin of these problems is not understood yet. It is believed that the level read is not representative of the magnet level as the cold mass cannot empty and refill so fast. However, there is not reading of the actual level on the cold mass during these periods. ElQA and powering cannot be done until the problem is fixed.
The full system has never been tested in SM18 or in the lab. The preparation of such a test will require weeks and some extra resources. It could also be part of the tests of SSS603 if it is re-cryostated.
AT/CRG proposes to add a path to equilibrate the pressure on top of the level gauge enclosure using the PT991 manifold and avoiding the capillary path. A pressure measurement device will be also added. This solution would be applied to Q5L5 and may be extended to the other magnets if successful.
MPP endorses this proposal and recommends to start preparing laboratory tests in the case that the solution proposed is not satisfactory.
The instability of the level gauges has been observer in stand-alone and semi-stand-alone magnets but is less important in the last case. The level is lost in 30 s and it may stay down from 30 s to 10 minutes. The pressure in the cold mass bath increases slightly and drops down suddenly as much as 0.135 bar just before the level loss. The control of the valves is done in a closed loop according to the level readings but the time delay is of several minutes and cannot explain it.
In the D2, the same level reduction has been observed but represents only 3%.
It is agreed that operating the wiggler below 50 A does not represent any risk and that the threshold could be increased to 3 V below this current. Above 50 A, the nominal threshold is 100mV. QPS team need to evaluate if this threshold is sufficient specially for ramping down the magnet from nominal current. (Action: R. Denz, W. Venturini-Delsolaro)
This circuit can not be powered during the present run as the correspondent DFBM has not been released yet due to a lack of understanding of the level gauge inside the D3. A separate procedure almost equal to that of the 600 A circuits will be written. (Action: W. Venturini-Delsolaro)
The worst case scenario consists on an undetected short from the quench heaters to the coil and a simultaneous accidental discharge of a quench heater. In this case, the fuse blows and protects the equipment. The maximum observed voltage is less than 22 V. There is no risk for personal if the circuit is properly earthed with low resistance and inductance. Earthing is done for ElQA purposes. (After the meeting, it was found that the ElQA earthing is not appropriate and proper shorts should be prepared. However for sector 4-5 it has been decided to proceed with the connection of the leads before the QPS individual tests.)
It is proposed to repeat PIC2 in the main bending together with a ramp to 2 kA followed by a fast power abort to check the energy extraction. The program will then follow the procedure from where it was left in 2007.
The capability to follow the demanded ramp rates for the 600 A circuits has been investigated. The current cycles for nominal optics and the feed forward systems does not represent an issue. The real time feedback system will however send functions to the power supplies with too high ramp acceleration. It needs to be further investigated after consultation with AB and the QPS team.
General Information.
In sector 4-5, the temperature of the arc is stable within a few mK. ElQA activities prior to powering can start. MPP congratulates CRG. This week is dedicated to ElQA of both DFBAs. The individual system tests for the QPS system is scheduled for Friday but could be done as soon as tomorrow if ElQA finishes before. The QPS tests should be done in all DFBA circuits including individually powered quadrupoles. Two of them are given priority but it should be done in as many as possible. Next week will start with powering of the MB circuit. ElQA-DOC tests will also be done in as many correctors as possible but only in the available time during main circuits ElQA. There are two tests of power converter stability. MPP requested a description of the tests. These tests could be done in all the already commissioned circuits. Extension of CRYO_START and CRYO_MAINTAIN logic for 120 A dipole orbit corrector circuits. M. Zerlauth.
Similarly to what was already observed for 60 A corrector circuits, some conditions necessary for safely run the 120 A corrector circuits are missing from the CRYO_Maintain and CRYO_start. An ECR has been circulated that proposes to prepare a loop from the 120 A cryogenic conditions and include it into the CRYO_Start and CRYO_maintain signals (can be an exact copy of the 60A according to A. Ballarino). The disadvantage of this first proposal is that, if the conditions for the 120A circuits are not fulfilled, it prevents the powering of any circuit in the same powering sector. Problems with the cooling of the beam screen can still be expected during hardware commissioning. S. Claudet added that the conditions for CRYO_Start could be simulated but not those for CRYO_Maintain.A second proposal consist on create independent signals for the CRYO_Start and CRYO_Maintain of the 120 A circuits exclusively. This will keep the flexibility during hardware commissioning and both permits will be merged at the end of commissioning. The implementation however of these new signals has very heavy implications in AB/CO.
An intermediate solution would be to prepare the loops and commissioning them independently. They would be linked to the CRYO_Start and CRYO_Maintain signal circuit by circuit only after PCC tests (10 A) have been done. This will leave powering of 120 A circuits after all other circuits in the powering sector problems are solved.
MPP supports the third option.The powering procedures should be changed accordingly. A signal or a flag that the activation is done could be visible somehow from the CCC. The ECR will be modified in this way and released. (Action: M. Zerlauth)
Problems of level gauges in the stand alone magnets
Loss of level for liquid helium gauges. J. Casas Cubillos.
Loss of the level signal in the matching section magnets happened already in sector 7-8 and is now preventing powering in sector 4-5. Typically, a pressure reduction is seen in the line D (~8 min) together with loss of level in the magnet (~3 min). The nominal heat load in the LT enclosure is estimated to 0.31 W which taking into account the impedance of the capillary, will create a pressure increase and lower the level in 10 mm or 2.5%. The heat in-leak necessary to lower the level by 75% is of 2.2 W. The possible source of such heat in-leak is unexplained. Besides, the behavior of the level gauge during the boil-off of magnet Q5L5 also shows regular steps that are not well understood. This increases the reading error from 2.5% to 10%. The origin of these problems is not understood yet. It is believed that the level read is not representative of the magnet level as the cold mass cannot empty and refill so fast. However, there is not reading of the actual level on the cold mass during these periods. ElQA and powering cannot be done until the problem is fixed.
The full system has never been tested in SM18 or in the lab. The preparation of such a test will require weeks and some extra resources. It could also be part of the tests of SSS603 if it is re-cryostated.
AT/CRG proposes to add a path to equilibrate the pressure on top of the level gauge enclosure using the PT991 manifold and avoiding the capillary path. A pressure measurement device will be also added. This solution would be applied to Q5L5 and may be extended to the other magnets if successful.
MPP endorses this proposal and recommends to start preparing laboratory tests in the case that the solution proposed is not satisfactory.
On the observed instability of the He level gauge readings in stand-alone magnets. V. Parma
The enclosure housing the level gauge in the matching section magnets is always the same and it is always located at the highest point according to the tunnel slope. The bottom part of the enclosure is connected to the bottom part of the cold mass, and the top is connected to line LD through a capillary 2 mm of diameter. The enclosure is 10 mm diameter at the level of the capillary and the level gauge made of a superconducting wire lodged in an epoxy cylinder is 6 mm diameter. If the level gauge leans against the enclosure, it could restrain the access of gas to the capillary and create an additional heat load.The instability of the level gauges has been observer in stand-alone and semi-stand-alone magnets but is less important in the last case. The level is lost in 30 s and it may stay down from 30 s to 10 minutes. The pressure in the cold mass bath increases slightly and drops down suddenly as much as 0.135 bar just before the level loss. The control of the valves is done in a closed loop according to the level readings but the time delay is of several minutes and cannot explain it.
In the D2, the same level reduction has been observed but represents only 3%.
ElQA
The ElQA of the DFBA left of point 5 is progressing well and it will be finished today. In 4R, a solid short to ground has been found in the MCS circuit. Three more shorts have been found in line N circuits. They could be in the warm part of the instrumentation. The diagnostics is not being done systematically. It should be decided in the next hardware commissioning meeting if it is worth investing time now. The low current module in the DFBAO in sector 7-8 has been checked and it is ok.Protection of the wiggler. W. Venturini Delsolaro.
The protection of the wiggler is done through parallel and coupling resistors and through energy extraction. The magnet is well protected if a quench is detected. However, there is no voltage taps in the magnet so the detection of the quench should be done through measurements of the voltage and inductance tables as for other 600 A circuits. A current reading device has to be added to the circuit. The inductance tables have been deduced from magnetic measurements and will be provided to the QPS team. AB/BI has requested the undulator to be ramped up before injection and ramped down after 2TeV where synchrotron radiation in the dipole will be sufficient for diagnostics. Given the high inductance at low current and a threshold of 100mV, the acceleration has to be very small (5mA/s2) and the total ramp time of the order of 15 minutes. The ramp up could be done in the shadow of the main bending cycling that last 18 minutes.It is agreed that operating the wiggler below 50 A does not represent any risk and that the threshold could be increased to 3 V below this current. Above 50 A, the nominal threshold is 100mV. QPS team need to evaluate if this threshold is sufficient specially for ramping down the magnet from nominal current. (Action: R. Denz, W. Venturini-Delsolaro)
This circuit can not be powered during the present run as the correspondent DFBM has not been released yet due to a lack of understanding of the level gauge inside the D3. A separate procedure almost equal to that of the 600 A circuits will be written. (Action: W. Venturini-Delsolaro)
Electrical safety of the LHC quench heater circuits in case of faulty heaters. R. Denz
The connection of the current leads has been scheduled after the quench protection system individual tests and thus, after the quench heater power supplies are charged. In the case of a DFBM, the number of supplies is small and the QPS team proposes to disconnect them before any work in the live part of the circuits. However in the case of the DFBA, more than 600 supplies would need to be disconnected to work in the main arc circuits. The risks of connecting the leads when the heaters are charged has been evaluated.The worst case scenario consists on an undetected short from the quench heaters to the coil and a simultaneous accidental discharge of a quench heater. In this case, the fuse blows and protects the equipment. The maximum observed voltage is less than 22 V. There is no risk for personal if the circuit is properly earthed with low resistance and inductance. Earthing is done for ElQA purposes. (After the meeting, it was found that the ElQA earthing is not appropriate and proper shorts should be prepared. However for sector 4-5 it has been decided to proceed with the connection of the leads before the QPS individual tests.)
AOB
During the winter shutdown only the QPS quench heater supplies were disconnected. Minor checks in the QPS system are scheduled for the RB. It is not necessary however to make any discharge.It is proposed to repeat PIC2 in the main bending together with a ramp to 2 kA followed by a fast power abort to check the energy extraction. The program will then follow the procedure from where it was left in 2007.
The capability to follow the demanded ramp rates for the 600 A circuits has been investigated. The current cycles for nominal optics and the feed forward systems does not represent an issue. The real time feedback system will however send functions to the power supplies with too high ramp acceleration. It needs to be further investigated after consultation with AB and the QPS team.
There are minutes attached to this event.
Show them.