Chamber uniformity and HV equalization The usual requirement
Chamber uniformity and HV equalization The usual requirement is: 95% of the gap area in the range G 0/1. 25 < G 0*1. 25 Since the current measured with the radioactive source is proportional to the gas gain and known that for DV=+105 Volt the gain doubles, the requirement above translates in a ± 34 Volt window. • The requirement above is appropriate to guarantee that all detectors are in the range imposed by the minimum efficiency and the maximum cluster size ? • Do we need to set a different HV to each gap ? I will show a preliminary analysis on 36 LNF chambers to provide an answer to these questions. C. Forti, D. Pinci – LHCb Muon Roma I - 27 -jul-04
Plateau width (M 2 -M 5) From testbeams: plateau width=170 V for bigaps; 150 V for 4 -gaps Lower plateau limits: 95% for bigap (2. 53 k. V) / 99% quadrigap (2. 55 k. V) Upper limit: cluster size in quadrigap < 1. 2 HV <2. 7 k. V (calculated from oct’ 03 results on the bigaps) 2. 62 ± 0. 75 k. V 99% Testbeam oct. ’ 03: BIGAP GIF july ’ 04: 4 -GAP
Plateau width (M 1) FROM OCT’ 03 (BIGAP): Upper limit: cluster size =1. 2; in bigap HV~2. 82 k. V FROM GIF: Lower plateau limit: e=99% in double-monogap HV=2. 65 k. V Work point ~ 2. 72 k. V Plateau width ~ 170 V Testbeam oct. ’ 03 : BIGAP GIF july ’ 04: DOUBLE-MONOGAP
Method for uniformity studies For each gap we have the current in 48 x 3 cells For each chamber we have: The gap uniformity is given by the current range of 95% of the cells: < I >/F < I(95%) < < I >*F (where <I> is the average current in the gap) From F we find a voltage range ± DV where DV = (ln. F/ln 2)*105
C D Compensation of the effect of the bent panel Several panels are bent at the center. This effect cannot be recovered during chamber assembly. However, if the bent panel is a layer 2 or 4 (pad-pad panel) in our R 3 chambers, the bigap uniformity is still good. CD Ex. Of chamber 27: the panel between C and D is bent but the effect is not visible on the bigap current.
Compensation of bent panel effect (ch. 27) 95% of the area of the bigap is in a current range equivalent to ± DV around the average current of the bigap < I >. Even with the bent panel the voltage range of CD is very small: DV = 15. 8 V
C D CD No-compensation of bent panel effect Ex. of chamber 31: here the dependences of the current on the wire pad number are similar in gaps C and D, so there is no compensation.
No-compensation of bent panel effect (ch. 31) Even in this “bad” case, we have that 95% of the CD area is within ± 33. 8 Volt from the average current. This is equivalent to the requirement: <I>/1. 25 < I(95%) < <I>*1. 25
Uniformity inside each gap 8/144 gaps of 7 ch. with DV > 50 V 40/144 gaps with DV > 34 V ± 50 Volt ± 34 Volt The gap uniformity does not profit of compensation effect: it provides a pessimistic result.
Uniformity inside each bigap 2/72 bigaps of 2 ch. with DV > 50 V 10/72 bigaps of 7 ch. with DV > 34 V DV > 50 V 2 chambers over 36 DV > 40 V 4 chambers over 36 DV > 34 V 7 chambers over 36
Compensation of 2 adj. Gaps (ch. 7) A B AB
Equalization of the average gap gain For each gap we find the shift DV required to align its own average current with the average current of all the gaps together: <I>=215 n. A HV system with 3 HV: HV 0 , HV 0 ± D D = 29 V from minimization HV system with 5 HV: HV 0 , HV 0 ± D, HV 0 ± 2 D D = 20 V from minimization
Equalization of the average gap gain Without equalization: all gaps at same HV This is the voltage spread DV in each gap to guarantee that 95% of the gap area is within ± DV from the average of all the gap currents <I>=215 n. A It is evident that we must set the HV of each gap to align its gain to the average gain.
Effect of equalization Percentage of bigaps within a certain range from the average current of all bigaps. HV systems with 1, 3, 5 HV values + hospital are compared. ~150 V
Effect of equalization Percentage of BAD bigaps: out of a certain DV range from the average current. In first column: only the effect of intrisic uniformity of the bigap. In other columns: the range DV is defined around the average current of all bigaps (i. e. from the HV working point). 1 HV means that all gaps are set at the same HV (no equalization) Uniformity 1 HV 3 HV 5 HV DV > 34 13. 9 % 43 % 26 % 21 % DV > 50 2. 8 % 10 % 7% 18% - 2. 8% is the effect of the spread of the average bigap currents (next slide). The advantage of 5 HV with respect to 3 HV is poor (few percent).
<I> ± 2 s No equalization Effect of equalization: spread of the <I> of the bigaps Band: <I>/1. 25 < I < <I>*1. 25 Equalization with 3 HV Equalization with 5 HV
Set of HV values including M 1 Working point from testbeams for M 1: HV=2. 72 k. V / M 2 -M 5: 2. 62 k. V Set of HV values in a 3 HV equalization system M 1 2690 2720 2750 M 2 -M 5 2590 2620 2650 In this configuration, 6 HV values + hospital are needed Set of HV values in a 5 HV equalization system (not needed) M 1 2680 2700 2720 2740 2760 M 2 -M 5 2580 2600 2620 2640 2660
Conclusions A preliminary analysis of the source scan data, collected for 36 M 3 R 3 chambers show that: - the requirement 0. 8 -1. 25 (DV= ± 34 V ) on the gap gain spread is not realistic; - a new requirement should be set on the bigap uniformity, of the order of DV=40 V 4 LNF chambers out of 36 to be rebuilt. To guarantee that 95% of the detector area satisfies the requirements on minimum efficiency and max. cluster size, the HV system must provide an alignement of the average gain of each gap to the average of all gap gains. Accounting for the different operating HV of M 1 with respect to M 2 -M 5, this implies that at least 7 different HV values can be set on the system gaps. We cannot exclude that after equalization some parts of the detector will operate at a high gas gain, eventually provoking a faster ageing of these areas. This will be investigated.
Spare transparencies
Effect of collimator Currents in last 16 chambers as function of X, Y. Even if the measurements are in the center of the pads (at ~7 cm from the chamber edge), a correction factor 1. 077 is needed to align the measurements in strips Y=1 and 3 to the central strip Y=2. In uniformity studies, we have rejected data from first 3 and last 3 wire pads, to eliminate border effects due to collimator.
Effect of collimator Currents in all chambers as function of X, Y.
Max cluster size quadrigap (tboct 03) Bigap (measured) 4 -gap (calculated)
Compensation of bent panel effect (ch. 27) Several panels are bent at the center. This effect cannot be recovered during chamber assembly. However, if the bent panel is a layer 2 or 4 (pad-pad panel) in our R 3 chambers, the bigap uniformity is still good.
Efficiency (4 -gap at GIF)
Effect of equalization (including ch. 1, 2, 4, 5)
Effect of equalization Fraction of gaps/bigaps within a certain range from the average current of all gaps/bigaps. gaps Bi-gaps
- Slides: 26