Dear Colleagues, As you may have seen the last ~24 hours have been not good for the experiment. Following the access yesterday the beam did not come back until late nite, and even at reduced intensity. Two problems occured that has occupied several of the BNL people today. a) Pii3 just quit. b) The turning on of D3 caused a second QLI (quenc link interlock) and dump of the RHIC beam. -Pii3 The first diagnosis was this is due to a bad disk. The system disk was replaced but erro message still occur and the machine may still go down. Even though data could be taken, the runno was reset to 1 for an ongoing run this morning. Hiro and I have been working on and discussing ways to be able to run w/o pii3. I thiought this was already some, but a few quirks in scripts prevented this. As of this point one should be able to run DAQ from pii12 w/o pii3 present. At least data will be taken, but the RUNDB will not be available. This second point is having Hiro looking into both replication of RUNDB on the rcas etc.. -Magnets The two incidents were caused by a) having the pet pages set at a high field, and in first case waweform generator was not on. This was clearly an operator error from our point, in the second case the 'ramp speed' had changed on the pet page from the value 120 sec that both Dan and I had seen earlier in the day to 1 sec. This caused a large spike that induced the QLI. To ensure this would not happen unless in case of real trips of D3 that we believe are very infrequent the following remidies are bveing implemented. a) hardware protection have been added to D3 powersuplly restricting/mdofiing ramp downs.. A problem was also foun with some of the circuitry. This will also be implemented on other magnets. b)The control of magnets will probably from tomorrow be done though another set of interface pages, that has build in protection of rampspee, will prevent turning on unless waweform is at 0 etc... c)The only issue left is true trips of D3 , and statistics will be pulled out of the database to see how infrequently this happens. The above problems is the rationale for having the spec at a larger angle (8 deg) where the interference is expected to be small, until the issues are settled an we can go back to our planne physics program thanks to Dana, Hiro and Bob for spending considereable time to resolve these issues. Flemming ---------------------------------------------------------------- Flemming Videbaek Physics Department Brookhaven National Laboratory e-mail: videbaek@bnl.gov phone: 631-344-4106 _______________________________________________ Brahms-l mailing list Brahms-l@lists.bnl.gov http://lists.bnl.gov/mailman/listinfo/brahms-lReceived on Thu Feb 5 20:00:56 2004
This archive was generated by hypermail 2.1.8 : Thu Feb 05 2004 - 20:01:17 EST