Minutes of the REX integration meeting on WednesdayFriday,13.05.2005, 25.13.05.2005

 

 

Participants: Luis FrailePierre Delahaye, Thomas Sieber, Mats Lindroos (chair), Fredrik Wenander, Nikolai Trofimov, Ivan Podadera, Nicolas David.

 

 

1. General Issues

-          A summary of the 138Xe run has been written. It can be found attached to these minutes. Starting on Tuesday 17.05., we will have a daily short REX meeting at 16.00 in the ISOLDE kitchen to discuss the startup and the status of the running Xe experiment.

-         

 

2. TRAP

-        PierreDue to the problems with the trap controls during the last run, Romain is at the moment reinstalling all programs of the trap CS.

-        For the rest of the week and beginning of next week, Romain and Ivan will continue with the emittance measurements at the trap during daytime. In the evenings the beam is given to WITCH.

-        On Thursday/Friday the setup for the Se beamtime will start.

-        The trap ion source has been changed. It shows a better stability but, a high power consumption.

-        By the end of the week Fredrik will check beam position at the entrance of the EBIS, using beam from the offline source (injection tests).

-           and Fredrik had the problem of discharges in the trap, means there was more current measured at the exit than injected from the test source. Nicolas David found a broken cable at a transfer element. The bad contact was responsible for the discharges.

-The test ion source runs stable now after the regulation has been improved.

-The application for the frequency scan of the trap has to be improved.

 

3. EBIS

-        The noise on the EBIS drifttubes, which made dry injection tests difficult, will be checked by Fredrik on Friday.

-        The work on the time of flight measurements at the EBIS has started. Cabling of the mass scan diagnostic box has been changed to allow for easy switching between beam monitoring and TOF.

-        The bakeout, which was planned for the time between the first and second REX run has been postponed.

-        There is still sparking observed in the extractor region. A new flange will be produced to have better insulation there.

-The bad transmission through the mass separator, which has been observed on Thursday, was partially caused by two switched cables in one of the steerers.

-There was again a problem with the cabling of the Wago module for the high voltage cassettes, which could be repaired by Nicolas.

-The transmission through the separator is at the moment 70- 80%, due to the high residual gas pressure, the current is about a factor 10 higher than usual.

-Stable beam for Linac optimization will be available on Monday.

-The baking after the run will be done now with the old NEG strips.

-One day is needed to optimize beam transport through the BTS.

-Dry injection (without electron beam and magnetic field) into the EBIS is still problematic because of the high noise level on the electrodes.

 

4. Linac

-        The experts from AB-RF, supported by BERTRONIX, managed to do a fast repair of the two 100 MHz amplifiers right before the run. The two broken high voltage transformers were replaced by our own spare transformer and a spare part from Linac3. Therefore (besides the constant problems with the 200 MHz system) the duty cycle of the machine is at the moment limited to 1%. The broken transformers have been shipped to Munich and will return to CERN with improved insulation within 2-3 weeks.

-        There is still one HV transformer of the weak type left in the RFQ amplifier. It will be replaced as soon as the repaired transformers arrive from Munich.

-        Due to the (almost) completed improvement of the low level control modules, the 100 MHz part showed during the Xe run for the first time a reliability which is comparable to the other CERN Linacs.

-        The 200 MHz system still requires a modification of its rf cavity. The problem seems to be, that at duty cycles above ~1% the ferrites in the cavity loose their properties (due to the high temperature). A parasitic mode at low duty cycles damped by these ferrites causes unstable operation and breakdowns at the anode. If better thermal contact of the ferrites and better cooling will solve the problem will be checked in the break we have in June.

-        There is also a relevant certain number of switch offs due to power reflected from the 9-gap cavity. This is not in connection with the amplifier, but a problem of sparking in the cavity. Since the cavity went back to operation only one week before the run, this is considered to be a conditioning effect, which should vanish with time. We plan to install a circulator to make operation more stable concerning at least this problem.

-          BERTRONIX could not solve the principle problem of the limited power level of the 200 MHz amplifier. Also the unstable operation at 10% duty cycle is still a problem. The amplifier could be tuned to ~100kW output power on a dummy load. It is running quite stable now at 30Hz, 600us. So we can expect closed loop operation on the cavity at about 85 kW.

 

4. Control System and Power Supplies

-        There has been a meeting with CO to discuss the work on the migration of the REX Control System. The REX team will publish the results together with the minutes of next week, after re-discussing some details.

-        The tests for control of the Danfysik supplies with the new front end developed by Nikolai have been finished successfully. The response time turns out to be similar than what we have at the moment.

-        Nicolas has reconnected the power supplies to the old CS to allow for beam optimization on Thursday

-        There were problems with a jitter on the GFAS timing pulse (EBIS) which is used as a trigger pulse for the Linac (caused beam losses). Fredrik managed to find a temporarily solution for this, but claims that the REX timing scheme should - in connection with the issue of a “slow extraction” - be investigated in general by an expert from CO. He will also contact Joakim Cederkall (who starts working at ISOLDE 1. July) to help with this.

-        Sylvestre Catin is currently looking into the problems of FECs switching off without obvious reason (BD, EBIS FECs).

-The BD FEC is still stopping from time to time, Sylvestre Catin has been informed about this.

-Nikolai found that there are not enough Ethernet connections in the network rack, where he wants to connect his Front end computer.

 

 

Minutes by Thomas