EMU Chambers Calibration Constants

Four different runs are taken for CFEB calibration and four for AFEB to cover all tests done previously on the chambers.

1. The first CFEB run is designed to pulse all 16 individual strips for each external capacitor 32 times, having a 6.25ns delay inbetween the pulses. In total 512 events are taken that are later analyzed separately. The analysis consists of calculating calibration constants for pedestal, gain and cross-talk for each strip in all layers and chambers present in the run. The software code for the analysis is based on Stan Durkin's note crosstalkdoc.pdf. The unpacker used is Sandbox.MERawFormat and it provides infomation regarding ADC counts, DMB and crate ID. A separate mapping of EMU chambers exists on the online database provided by Igor Vorobiev, which will identify each chamber in the run.

a) For pedestal calculation the first 2 time bins are used, the pulse signal appearing much later in time. A mean value is calculated for pedestal and its RMS for each strip, giving in total 480 numbers/chamber. 480 numbers are stored for pedestal means and 480 for RMS.

b) Gains are obtained by calculating the constants for the line joining two ADC values: the first value is the pedestal value determined in a) , the second is the ADC corresponding to injection of a pulse with amplitude 1V or ~112 fC. 480 numbers are stored for gains, 480 for the intercept and 480 for RMS.

c) Cross-talk is determined by the following formula:

Where is the maximum charge obtained from pulsing a strip, is the charge to the left of the pulsed one and is the one to the right of the pulsed strip. The same algorithm is used as in Stan's note above. 480 numbers are stored for left cross-talk, 480 for right cross-talk and 960 (in total) for values.

These values are sent to the online database using ORACLE tools from Igor and the results can be seen here. From online, data is sent to offline tables using an xml file created by this offline package.

In the tests performed for the chamber commissioning the far cross-talk was calculated as well, however since it was found to be very small, the CFEB calibration constants will not include it.

-- OanaBoeriu - 17 Nov 2005