DØ Calorimeter Logbook

May 19, 2002 15:15 - From Dean Schamberger
Hi,
  I just wanted to warn you all that I did in fact reprogram the crate
0x4C VBD, to
readout the Controller card data, as well as the 12 T&C cards and the
PIB data.  This
"should" be transparent to everyone.  I did it by hand, while waiting
for the DAQ system
to start working for the store that went in friday evening (store number
in my logbook entry).  I have LOTS of time to check what I was doing
visually....but that does not
mean that I did not make a typing error.  If CALIB stops working this is
something
that might be the cause.
  I did stay around long enough to see the Gustaaf's L3 "bunch/turn"
checking did not
kill every event, so at least some of the data it still OK.

      Dean

June 18 Run 157478 has TCB v26 and TCC v25
Jun 18 11:13 Run 157479
Jun 18 11:46 Switch to single buffer

Jun 19 20:05 Change to multi-buffer
Jun 19 20:52 switched back to single buffer mode. run 157574 on. 
       20:18 20:50 157573 global_CalMuon-7.30 21797 / 11.0 Hz /2.5E30 good (multi-buffer) 
       20:52 23:28 157574 global_CalMuon-7.30 105755 / 11.3 Hz /2.5E30 good smt H disk voltage trip 

Jun 20 08:10 still single buffer 157577 going

Jun 23 15:31 Florencia o shift. 
Crate 4c is not receiving events from time to time. It doesn't 
have front end busy. But it is 2% of the times off. 
This has been happening for more than 2 days now. 
I called expert on call. Although there is no beam the cap would 
like to solve this problem in case we get beam. 

Jun 23 16:45 Florencia Canelli: Leslie changed to single buffer. This made 
             the problem dissapear. But we will continue running on multi buffer. 

Jun 25 19:15 Tuts: As of the next store, the ped suppression window will be changed, so one should not load the prepare for run file until 
new peds have been taken for the next store. Will attempt to take them during the 30min access at 8pm. 

Jun 25
19:14 19:34 158040 global_CalMuon-7.31 22,733/18.3 Hz/2E30 good   Ramping down the magnet (preparing for access)
19:46 19:58 158041 global_CalMuon-7.31 11,542/15.9 Hz/2E30 uncertain  Magnet was ramping off.  Multi-buffer mode.


Jun 25 22:42 Tuts:
Dean has done the following: 

1. we were running with T&C controller board #001, has intermittent multibuffer errors 
2. placed #001 on extender, multi buffer errors went away 
3. changed T&C controller #001 from ver 25 to ver 26 
4. still ran on extender in multibuffer mode w/o errors 
5. replaced #001 with #002 (also ver 26) on extender, board did not respond 
6. replaced #002 with #001 on extender, and programmed onboard flash RAM to remember ver 26, worked fine 
7. removed #001 from extender and installed in crate, now did not respond to L1 accept 
8. put #001 back on extender, running now with internmittent multibuffer errors9. dean leaves, board #001 still running on extender 

Jun 25 22:58 Tuts: Downloaded new cal_prepare for run with new threshold
looks good, so I am downloading 158056 (x8) & 158044 (x1) using cal_prepare_for_run-1.4, which should also use the new limits. 

Jun 26 1:48    3:49    158068  global_CalMuon-7.31     70,750/15.9 Hz/7E30             good       occasional problems from crates 53 & 4A. 4C always red on UMon.


Jun 27 20:32 Leo on shift. Everything (GUIs) looks fine, but no beam. Was running a l2_zero_bias run without Cal crates. 
Now captain decided to run a zero_bias run with L2 and Cal crates. Changed calorimeter to single buffer mode. 

Jun 30
11:16 12:05 158535 global_CalMuon-noL2MU7.4 66,493/23Hz/7E30 setgood   prescales as in previous run. Stop to put cal in single buffer mode for more tests.
12:20 13:10 158537 global_CalMuon-noL2MU7.4 38,073/13.7Hz/7E30 setgood   cal in single buffer mode. CEM(1,5)^2 set to 4 and CJT(2,5)^2 set to 2.   
13:15 13:35 158538 global_CalMuon-noL2MU7.4 11,430/8Hz/7E30 setgood   back in multi buffer mode. CEM(1,5)^2 set to 4 and CJT(2,5)^2 set to 2.      



Jul 5 13:55 Bassler: Latest version of PIB-eproms installed: PibRampB6 version 8: 
- corrected update of current pattern number in Pib memory: check with manually programming a pattern ramp went nicely! 
- wait function implemented that allows to put a wait upto 10000 iterations in between downloads within a ramp. 
not fully tested yet. 
through all the 

Jul 6 17:20 Oleg:
Run 158978 
Level 1 trigger is unstable and sometimes has a high rate. I looked for hot celles a Trigger Towers. 
bad sigma : (17, 61, 15)(46,1,6,1,11) 1.05798 GeV 
high occupancy : (17, 61, 15)(46,1,6,1,11) 0.331 pedestal 0.508609 GeV 
This cell has been killed. 
I run L1CAL Examine. Finally I found Hot Trigger Tower eta=9 and phi=31. I call to Daniel Mendoza 
he came to fix the problem (disconnect some cables). 

Jul 6 20:01 Oleg:
Run 159983 
bad sigma : (-36, 39, 11)(40,5,6,3,7) 1.03747 GeV 
bad sigma : (23, 61, 11)(46,2,1,1,7) 1.18241 GeV 
bad sigma : (24, 20, 15)(44,9,5,2,11) 1.14908 GeV 

Jul 9 14:08
Dean and MingCheng found a noisy +13V power supply in quadrant 41 BLS crate 0. It had ~100mV A/C ripple 
that caused noise in all trigger output in that BLS crate. They swapped it with the last working power supply. 

Jul 9 14:18 MingCheng replaced a daughter card: crate 6 ADC 6 BLS 3 tower 3 

Jul 13 10:22 The TC Mode for crates 0-11 (except 5) are in mode 0x889 (Si Buff Master) rathe rthan the usual 0x89. 
Tried to reset this with ./vme_tc_ctrl with no success. Daniel M. checked one crate(0x40) to see that 
it was ok. 

Jul 13 10:49 Ran start_cal tandc and did a Global T&C set to Normal. Now the crates are in TC Mode 0x88 as normal. 
The previous mode was noticed by the DAQ shift who say 20% front-end busy from the cal crates in 
a zero bias run. 

Jul 15 12:46
One bad BLS board found yesterday in 46-3-7 is replaced. The board is clearly having problems. 
(hot voltage regulator). New board works fine when checking with Cal_elec. 
Cal-elec won't run with version t02.08.00, so I have to swiched back to t01.67.00, all crates are checked. 
--- Mingcheng 

Jul 30 08:51 Supervised Access: 
Bawo has a list of "bad" (noisy,etc) SCA cards. We swapped some of them (BLS board) to localize the problem: 
Crate-ADC-BLS 
2-0-6 w/ 2-2-2 
2-2-0 w/ 2-0-4 
2-1-6 w/ 2-1-0 
2-7-3 w/ 2-6-5 
2-7-4 w/ 2-8-2 
2-7-5 w/ 2-9-1 
8-3-2 w/ 8-2-4 
8-3-4 w/ 8-1-2 
8-8-6 w/ 8-6-4 
A noise study run was taken after these swaps: Run 160775, pedestal subtracted, all preamp and ICD off, 
fixed L1 cell to be 14u, (13u, 15u), # events = 50k 
All BLS boards will be swapped back to its original place in case that Pedestal Calibration still doesn't 
work today. 

Jul 30, 2002 11:28
During access we measured ripples on all BLS power supplies. 
Supply 10-1 (middle supply) had 140mv ripple on 13V power line. 
This supply is swapped with a modified supply. Ripple on thenew 
supply is checked and we found that this one is good. 
We have replaced crate controllers 7-4 and 7-5 (remembered to 
change the jumper position) to fix the temperature monitor problem. 
This seem to be fixed. 

Jul 30 12:21 Supervised Access (2): 
1. Checked all crates with pedestal run; 
2. Replaced one SCA card (crate 4 ADC 11) and one Analog Drive (crate 08 ADC 11); 
3. Cal_elec keeps crashing once in a while => NEED TO BE FIXED !!! 
4. No time for debugging with pulser run. Access is over. 


Jul 31 17:31 Major Configuration Change
 Operator: Dean Schamberger
Configuration change:
 After a few hours of searching, I found the typing error 
in version 27 of the T&C Controller code.  It should fix
the bug that caused crate 0x4C to readout multiple times.
As well as improve the monitoring information.
  We are finally ready to try version 28...stay tuned.

L1CAL Trigger

From Dan:

- For Trigger Towers eta 1:4  switch from the Run I electronics to
the Run II  Term-Attn-Brd's  and the Run II PROM's.

I know this was done by 24-MAY-02.  See the log book for 21:24-MAY-02.



- installation of coverage to 2.4 enabled in trigger

We installed this during the "2 week shutdown" at the beginning of
June.  We came out of this shutdown ready to run with TT eta's 1:12.
I know that we were ready to run this way by 15-JUNE.  See the log
book entry for 11:15-JUNE-02.

BUT I think the first store or two after the shutdown was done with
only TT eta's 1:4 to verify that all was all OK.  My guess is that the
"global cal muon" trigger version number made a significant change on
the 2nd or 3rd store after the 2 week shutdown and that would have
been the first physics use of TT eta's 1:12.



- calibration adjustments made to L1Cal from Bob's first set of constants

I know that this was done sometime in the window  26:28-JUNE-02
My guess would be that I changed it on the 27th.


- change of timing of L1Cal from initial timing measurements

This was done on Thursday  1-AUG-02.



- Repair the problem with Tot Et Ref Set #3

Tot Et Ref Set #3 (which is typically used by COOR during global
Physics running for the 3.0 GeV Tot Et Ref Set) was not correctly
counting the number of TT's over threshold from the time that we
started to use TT eta's 1:12 (I believe that this problem started
right away) until 23-July-02 when it was repaired.  This was the
problem that Tot Et Ref Set #3 always thought that TT -7,2 was over
threshold even when there was no energy in the calorimeter at all.



- Increase the gain on  52  even eta EM TT's in the eta range +- 6:12
  by a factor of  1.45  because these TT's have the  correct  value
  resistor installed at R80.    13-AUG-2002



- Plug back in the TT's   eta -7:-12 phi 9   and   eta -7:-10 phi 10.
  They had all been unplugged because of noise.
  Reduce the gain of  Eta +- 7  EM and Had  to 2/3 of what it was.
  Increase the gain on  3  even eta EM TT's in the eta range +- 6:12
  by a factor of  1.45  because these TT's have the  correct  value
  resistor installed at R80.    This was done 22-AUG-2002



From Dean:
Hi Leslie,
  I think I also logged the FPGA 26 change, but in the captains logbook,
since I was on shift, and did not realize that CAL config changes should
probably be logged in the CAL shift logbook....

  The other "significant" change was going from single to multi-buffer
running.  As I remember the sequence, the cal was in "multi-buffer" mode,
but being protected from a bug by the tracking crates.  Then one day,
they changed to multi-buffer, and the cal crates started missing events.
I think we ran for a few runs like that, until it was clear that we needed
to put CAL in single buffer mode, while I tracked down the bug.  Then I
put in the new code (v26?) and started running in multi-buffer mode for real.

  The only way I know of to find out when the tracking crates started
to run in multi-buffer, is check other detector's logbooks (in case they logged
the event), old DAQ-shifter E-mail, askGustaaf, or guess from when the
CAL crates went back to single buffer mode for a while.  The way to
really know when the CAL was in single buffer mode, is to look at one event
in every run.   The T&C cards report the state of the "single buffer bit"
in the low order bit of the third word.



>From Kehoe:
The extension in triggering to eta 2.4 was first usable in run 159705.  We
had installed it during the preceding 2 weeks and it is in the readout for
a few runs before this.  This was the first time triggering was enabled,
though.