cms week – march06
DESCRIPTION
CMS WEEK – MARCH06. REVIEW OF MB4 COMMISSIONING DATA. Giorgia Mila 17-03-06. 1. MAIN TOPICS. CMS WEEK – MARCH06. ANALYSIS TOOLS AND RESULTS CALIBRATION FIRST ALIGNMENT STUDIES. Giorgia Mila 17-03-06. Full analysis. Calibration. Alignment. 2. ANALYSIS TOOLS. CMS WEEK – MARCH06. - PowerPoint PPT PresentationTRANSCRIPT
CMS WEEK – MARCH06
REVIEWOF MB4
COMMISSIONING DATA
Giorgia Mila 17-03-06
1CMS WEEK –
MARCH06
MAIN TOPICS
Giorgia Mila 17-03-06
1. ANALYSIS TOOLS AND RESULTS2. CALIBRATION3. FIRST ALIGNMENT STUDIES
2CMS WEEK –
MARCH06
Full analysisCalibrationAlignment
ANALYSIS TOOLS
For each MB4 chamber placed in the wheel 1, we have observed :• time boxes• occupancies• efficiencies• angular distributions• residual distributionsin order to check anomalies in the cell behavior
In general the chambers work fine but there are some problematic cells
Giorgia Mila 17-03-06
PROBLEMATIC CELLS 3CMS WEEK – CMS WEEK –
MARCH06MARCH06
Wheel 1 SL1L1 SL1L2 SL1L3 SL1L4 SL3L1 SL3L2 SL3L3 SL3L4Sector 2 8,76,94 83 --- 4,23 --- --- --- ---Sector 3 --- --- --- --- --- --- --- ---Sector 5 4 --- --- --- 51 --- --- 3Sector 6 --- --- --- --- --- 1 (L.S.) --- ---Sector 8 --- --- --- --- --- --- --- ---
Sector 12 --- --- --- --- --- --- --- ---
DEAD CELLS & NOISY CELLS
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
EFFICIENCY - the method - 4CMS WEEK – CMS WEEK –
MARCH06MARCH06
We measure the efficiency from the segment extrapolation :we take a reconstructed segment and for the cells it goes through we look if there is a hit or not
We count not only the hits associated to the segment We look only at the segments reconstructed with 7 or 8 hits
inefficient cell
We apply this method on data taken with a HH+HL trigger
EFFICIENCY (I) 5CMS WEEK – CMS WEEK –
MARCH06MARCH06Studying the MB4 efficiency we have found 2 cells which have a bad cathode HV connection :sector 5, SL3 L4 w 36,37 ....
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
dead cell
EFFICIENCY (II) 6CMS WEEK – CMS WEEK –
MARCH06MARCH06... but also an inefficient cell which has a good time box (maybe a strip HV problem – to be investigated) :sector 2, SL1 L1 w21
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
dead cells :
ANGULAR DISTRIBUTION - the method -
7
CMS WEEK – CMS WEEK – MARCH06MARCH06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
We suppose that the measured angular distributions of reconstructed tracks N(φ) for a chamber with an angle φo with respect to the vertical direction are :
φ0
N(φ, φ0)= kφo · f(φ- φ0) · acc(φ)
With an iterative method including a combined fit of the measured angular distributions it is possible to extract:• the cosmics angular distribution f(φ) (parameterized)• kφ0· acc(φ) = angular dependence of the trigger acceptance
acc(φ) = angular dependence of trigger acceptance (unknown) f(φ) = cosmics angular distribution (unknown)
with :
NOTE : the chambers must have the same internal geometry
for more details see my presentation of 8-02-06
ANGULAR DISTRIBUTION (I) 8CMS WEEK – CMS WEEK –
MARCH06MARCH06We have studied the global angular trigger acceptance for the left chambers:(HH+HL trigger)
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
sector 5 sector 8sector 6
Trigger acceptance
-30° 60°-60°
Measuredangular
distributions
1. The trigger acceptance is asymmetric for MB42. Singular acceptances consistent one from the other
NOTE :
ANGULAR DISTRIBUTION (II) 9CMS WEEK – CMS WEEK –
MARCH06MARCH06and the global angular trigger acceptance for the right chambers :
sector 2sector 3 sector 12there is a range of angles where the trigger
acceptance deviates a bit from the global one
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
The trigger angular acceptance changes sign moving from right chambers to left chambers: this confirms that the asymmetry is related to the staggering between the SLs
OCCUPANCY – sector 12 10CMS WEEK – CMS WEEK –
MARCH06MARCH06
sector 12 sector 6
We suspect that the trigger configuration was changed for the commissioning of the chamber placed in sector 12: in fact it has an
occupancy distribution very different from the other MB4 chambers (we don’t see the regular pattern of drops
in the cell occupancy)
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
t0-event correction 11CMS WEEK – CMS WEEK –
MARCH06MARCH06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
We have used the t0-event correction (Anna’s routine) on our data
t0 correction t0 correction on logarithmic scale
These tails give problems on digi computation (we had to reset them to zero)
TDC counts
TDC counts
CALIBRATION 12CMS WEEK – CMS WEEK –
MARCH06MARCH06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
2- We compute the ttrig correction by a k-factor : meantmeantrig ktt
The method to compute the correct k-factor:we plot residuals distributions using different k-factors
for each k-factor we compute the distance from the means of residuals in the left and in the right part of the cell
we fit this distance with a line and we extrapolate the k corresponding to the distance=0
res_posres_neg
3- We extract the drift velocity from the mean-timer distributions
1- We calculate the tmean and σtmean by fitting the rising edge of the time box
mean distance
k0
commissioning data & t0 correction (I)
13CMS WEEK – CMS WEEK –
MARCH06MARCH06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
Using the routine for the t0-event correction we obtain 2 different k-factors for the 2 phi SLs.
For example for the MB4 chamber placed in sector 6:k_SL1=1.84 ( tmean = 1881.68 σtmean = 7.95756 )k_SL3=2.81 ( tmean = 1880.86 σtmean = 7.74864 )
When we recompute the residuals with these two k factors we obtain a point which is very distant from the linear fit
k_SL1
mean distance
14CMS WEEK – CMS WEEK –
MARCH06MARCH06
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
commissioning data & t0 correction (II)
If we compute the k factor which minimize the distance from the means of residuals without the t0 correction routine, we obtain a reasonable calibration:• very similar k factors for the 2 SLs (we use their mean)
For example for the MB4 chamber placed in sector 6:k_SL1=1.79k_SL3=1.74
• a point on the fit (when we recompute the residuals) :
mean distance
k_SL1
15CMS WEEK – CMS WEEK –
MARCH06MARCH06
t0-event correction
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
Residuals : not t0-event correction vs t0-event correction (MB4 chamber : wheel1, sector 6, SL1)
σ=526μm σ=328μm
We calibrate without the t0-event correction, and after we apply the t0-event correction routine to the data.It seem to work well :
The t0-event correction improve the σ of residuals but it remain of about 300 μm ! ... one possibility is that we use different chi2 cuts in the track selection
16CMS WEEK – CMS WEEK –
MARCH06MARCH06
t0-event correction
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
Anna’s talk at the CMS week on 06/12/05
17CMS WEEK – CMS WEEK –
MARCH06MARCH06
ALIGNMENT
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
δ are the misalignments with respect to the nominal positions taking as reference the first layerr are the mean residuals from 8 point fits
NOTE : for MB4 we restrict the study on two dimensions
δ1δ2δ3
δSL
δ4δ5δ6
r1r2 r3r4
r5r6r7r8
In our method we have 7 free parameters ( misalignments δ )
18CMS WEEK – CMS WEEK –
MARCH06MARCH06
ALIGNMENT STUDIES – the method -
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
The residuals are correlated to the 7 misalignments by a linear relation :
SL
C
rrrrrrrr
6
5
4
3
2
1
8
7
6
5
4
3
2
1
where C is a matrix (8,7) and its rank is :
6C
We have 2 ways to compute misalignments :1. fix δSL (from ISR measurements) and calculate the layer misalignments (δ1,2, ... )2. fix the layer misalignments (from CCD measurements at the assembly sites)
and extract δSL
19CMS WEEK – CMS WEEK –
MARCH06MARCH06
PRELIMINARY RESULTS
Full analysisCalibration
Alignment Giorgia Mila 17-03-06
MB4 chamber - sector 2 -
(cm) ISR (cm)δSL 0.00783165 0.003
CCD (cm)
δ SL3 L4 -0.000807 0.0175δ SL3 L3 0.0471147 -0.0148δ SL3 L2 0.0479869 0.0500δ SL1 L4 0.00795136 0.0059δ SL1 L3 0.0107243 0.0110δ SL1 L4 0.00955415 0.0085
NOTE : we are working to understand the mismatches between our results and the measured data
1. We have finalized our tools we want to use to check the chamber performance on commissioning data.
2. We are using the ORCA Calibration Tool to calibrate the chambers3. We are starting to use the t0-event correction routine
(very useful but with some problems to be understood)4. We are trying to check the alignments parameters on
commissioning data in a simple way
20CMS WEEK – CMS WEEK –
MARCH06MARCH06
CONCLUSIONS
Giorgia Mila 17-03-06