Choose timezone
Your profile timezone:
MP3 Meeting
→
Europe/Zurich
874/1-011 (CERN)
Description
Agenda:
- BLM threshold wrt quench limit (Bernd)
- Measure signals on switches, PC and 3 magnets in the RB circuit of sector 81 during next TS (Emanuele, Arjan)
- Feedback from CCC (Michele,Arjan,Markus)
- AOB
PARTICIPANTS:
Matteo S, Fabio F, Hugues T, Knud DP, Zinur C, Arjan V, Emanuele R, Rudiger S, Bernd D, Michele M, Nuria CL, JPhilippe T, Mike K, Richard M, Sandrine L.
- MINUTES OF LAST MEETING
There was no specific minutes. Everything is included in the MP3 presentation done at LMC on the 18th of August and that is on the MP3 indico site. http://indico.cern.ch/conferenceDisplay.py?confId=105294
- BLM THRESHOLD WRT QUENCH LIMIT (by Bernd Dehning, see slides):
After some discussion, it was clarified again the scope of the measurement that is: to match the BLM thresholds to the trigger of the QPS system in such a way to avoid a magnet quench without excessive margins. The BLM threshold may then be raised to reduce the number of unnecessary beam dumps.
The test configuration will require modified nQPS cards which start a buffer acquisition above 50 mV threshold without firing the quench heaters. These cards have been prepared by Jens and are already installed in a cell. The threshold has not yet been modified but it can be set remotely at any value. The threshold at which the heaters are fired in the event of a symmetric quench (now at 600-800mV) are not modified. The protection of these magnets is thus not compromised and always active.
The test will consist on creating transient and continuous losses while recording at the same time the voltage seen in the coils. This voltage is not a direct input to the threshold simulation but it is thought that a simple scaling law will be sufficient for estimated the margin to quench. The test should not trigger any heater discharge as the threshold on the BLMs will stay nominal. The total beam intensity should be lower than the quench limit in order to avoid saturation on the QPS signals which will invalidate the measurements.
Answering a question of Arjan, Bernd assured that the test will work successfully if particle showers effects on coils (thermal) are “reproducible”. This is the case during the test when the location of the loss is controlled.
The committee reminded that at such a low value, the acquisition buffers will be filled during any rap. A reset needs to be dump on all nQPS buffers after any ramp.
Following a suggestion of Rudiger, a future test “BLM signal versus current level in the magnet” could be done on an IPQ since there we can individually change the current to simulate energy dependent quench levels. Nuria reminded that there is no nQPS for IPQs yet but that the PXI system currently used for slice resistance measurements could be used instead. To do this, the configuration of the cabling has to be changed.
MP3 don’t see objection to the test as the protection of the QPS remains unchanged.
- PROPOSAL FOR MEASUREMENTS ON RB.A81 ON 1st SEPTEMBER (by Emanuele Ravaioli, see slides):
Looking at previous tests, two different reactions have been seen to the power supply ringing from different magnets. About 60 % of the magnets seems affected and show a peak to peak perturbation up to 5 times larger than the simulation. Emanuele has checked for possible correlation with: location, magnet manufacturing, magnet installation, hardware differences but nothing was found. It was anyway remarked how QPS cards are coming from 2 manufacturers and this could be a track.
A test is planned for 1-2 September night (from 18:00) where scopes will be used to monitor the response at higher frequency. The voltage signals coming from the power converter, both switches and three magnets will be monitored. Patches are in preparation and must be installed in between the QPS protection circuits and removed after the test. Similar patches are being used by the EI team for IPQs measurements. Some kind of re-commissioning will be needed after the test to verify the circuit and protection integrity. A team of expert will be present (Arijan and Mike from MP3 Team) for the test run.
The test execution seems to be compatible with the technical stop apart for unforeseen events. Everything must be completed for 7 o’clock in the morning. Since several actions must be put in place (hardware preparation, hardware installation, RESTORING nominal hardware conditions WITH CHECKS after the test completion) Rudiger suggest that Emanuele will prepare a full list of steps and actions that will be presented to the MP3 and LHC community before the technical stop.
- FEEDBACK FROM CCC (by Arjan,Markus,Michele):
Nothing specific to be reported.
- AOB
Nuria reminded how we have to go on with IPQ splices measurements:
For the actual location of the hardware only ONE test remains (RQ10.R8). It will take about 1 hour. Any person in shift from now to the end of the week that will know about a run stop of minimum 1.5-2 hours should try to do the test (procedure is available in MP3 repository).
During the technical stop the hardware will be moved to Points R2 and R4. A total of 8 tests should be done before the next technical stop expected in November (about 8 hours of tests).
There are minutes attached to this event.
Show them.
The agenda of this meeting is empty
