Peds study for 63 GeV

Hello,

I have done some study on pedestal tables already stored in DB. This is one of the items listed in Alex's task list for production readiness. Here, I am trying to give summary of my observations for 63 GeV run period, even though I have looked into entire '04 run.

WHAT I DID =========

(i) A macro is used to obtain all numbers for BPRS, BSMDE, BSMDP peds for various entries in DB.

(ii) MEAN and RMS values are extracted for all channels.

(iii) Time variation of those numbers are studied.

(iv) Details studied to find discrepency if any, in peds tables from time to time.

(v) Numbers are compared with with Dimitry and Julia's browser.

WHAT I DID NOT DO =================

(i) No attempt has been made to recalculate peds from data and compare with DB entries.

(ii) No check is made in the code (ADCtoEMaker) if the peds are being applied correctly to eaxch channel. I assumed the discrepency found last year has been fixed.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------

BPRS

During the 63 GeV run period (time between 2004-03-25/19-11-57 and 2004-04-02/09-02-06) there are two entries of peds for BRPS DB.

Those entry dates are: (a) 2004-03-25/19-54-14 and (b) 2004-04-02/07-11-47

Observations ===========

(i) ALL entries in the time stamped (a) is zero for ALL channels both for mean and RMS.

(ii) So, I ended up with one table for the entire run, in that time stamped entry (b),

MEAN =======

(a) has -ve mean values in channel 1946,1967. In both places mean=-327.68.

(note: In march entries we have seen several cases where there are -327.68 for mean and rms both).

(b) Two regions in channel Ids have zero mean , IDs for those zones are: (101-140, 901-980)

The figure to look at for mean variation vs channel number is,

(Mean variation with channel no)

RMS ===

(note: in all plots on rms, value shown is rms*10)

(a) Has zero rms, for channels where mean =0 as well.

(b) in region, 1940-1970, 6 channels have high rms.

(RMS variation with channel no)

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

BSMDE

Time stamped entries in the 63 GeV time window:

(a) 2004-03-26/07-12-42

(b) 2004-04-01/13:57:51

(c) 2004-04-02/07:14:14

-->Out of 3 time stamped regions, only (a) and (c) show entry, for (b) all channel_entries are zero.

--> Even though in SMD table definition, there are array of [18000][3], presumably for 3 time bins, but peds table has entry only for [18000][0].

Observations: ============

MEAN: ====

day0 (i.e (a) in list above):

-> There are holes (i.e channel_zero entry) in the channle nos, 5851-5859, 6001-6841.

--> After 5859 upto 5957, there are zero entries in alternate channels.

day 2 (i.e (c) in list above):

--> Two new holes appeared in id zones 600-750 and 4501-4650 in addition to the ones seen in (a).

Variation of ped_mean with channel_no and for two days are shown in figs below, (note: in 2D plots, X axis is number of channels, while Y axis represents number_of_updates_in_tables, when there are two updates in 63 GeV run pedriod, there are two rows for all channels.)

(Mean variation with day and channel no, shows the holes)

(Mean variation with day and channel no)

RMS ====

Holes exist in the same region as for mean.

some time large RMS seen for few channels.

figs to look at:

(rms variation with day and channel no)

(rms variation with channel no)

----------------------------------------------------------------------------------------------------------------

BSMDP

3 times stamped entries:

(a) 2004-03-26/07-12-01

(b) 2004-04-02/07-13:06

(c) 2004-04-02/08-05-01

--> Tables in (b) and (c) are EXACTLY same in entries for all channels, so I ended up with 2 entries.

Observations: ============

day0 (i.e (a) in the list) --> No zero entry in any channel.

day1 (i.e (b) in list) some zero enries (601-660,4501-4650)

--> one RMS value is -ve, on channel_no =6580 , rms=-327.68.

(Mean variation with day and channel no)

(RMS variation with day and channel no)

(RMS variation with channel no for one table)