Monitoring Controls System meeting on November 25, 2003


Attendance:
Mary Convery, Steve Hahn, Jonathan Lewis, Dan Ryan, JC Yun

NEWS:
The long shutdown is over. We are back to the data taking mode.

VNODE:
The COT1 mystery mentioned during the last MCS meeting is solved. An alert ace noticed that there was an error message concerning with tags in COT1. There is no logical link to COT1. The correct logical name for COT system is COT2. One of the high voltage bar in the HV summary page had typo - COT2. Now the bugs are fixed.

Trigger Inhibit:
Jonathan thinks making I/O card working is not so difficult. Aces should mask off system when partition is set in error handler - not from the iFix trigger inhibit page. SVX subsystem matching is not done yet. There are 3 systems NI I/O cards installed but we are not using them yet. These are MUON3, CCU and CLC system. 
Addendum: plug does not inhibit when in stand-by, Steve will ask Luciana to fix this.

TOF:
Steve said TOF still cause some problems from time to time.

ICICLE:
There were frequent ICICLE problems during the last month. Most of the problems disappeared recently. Most problems were caused by the fact that CDF database was on and off all the time.

CLC:
For most part, CLC is working. However there are a few problems: 1) Alarms go off without reason. 2) Can not raise nor lower high voltages due to connection time out. Tests show there are noise problems. Mary and Steve will move the CLC PC to where the CAEN is to cut down noise. Hope this works. 
Addendum: this was done later that same day with the help of J.J. and Mark Knapp.  Doing this made a HUGE difference in the performance of the CLC1 node.  No errors, no timeouts.  Furthermore, the readout seemed to speed up by a large factor (10X?).  I (Steve) suspect that there are error detection/correction routines at the driver level which were slowing readout down considerably even when it did work correctly.  However, I did find one problem which I did not know about until yesterday: according to this PDF file, you have to terminate BOTH ends of the Lemo cable with 50 ohm terminators.  Sure enough, I look at the plug A1303 and its terminated; the CLC A1303 was not until just now.  There is one remaining problem: the inhibit line out of the CAEN 527 does not appear to be working correctly.  I will check into this, also.

PSM:
Roberto is working on the BMV boxes in the collision hall to fix the problem of serial number corruption. We think some kind of shock is setting channel serial number to a bogus number, 10.

The long known, flipped bit problem was fixed by Dan by reshuffling data lines, so that a different shorter line used the serial-to-RS485 convertor which is different from all the other lines.