The ADC calibration of all strips relative to each other
in a given chamber will need to be known to less than 1%
in order to obtain a 100 
m resolution from the fine cathode
planes, if the readout pitch is 1 cm. To achieve this, all strips
will have to be pulsed with a common pulser over several pulseheights
spanning the range of the ADC. The strips in a given chamber may
be pulsed in sets of every fourth (or more) strips at a time to avoid
cross talk. If a calibration error less than 1%
is to be achieved, a precision pulser must be used, and the capacitive
coupling of the pulse to the strips must be such that the error
in the known capacitance is less than 1%. This can be achieved
either by using precision capacitors which are uniform to the
sub-1% level or by using capacitive strips whose area is known
to less than 1% and whose distance from the strips is constant.
The North muon arm consists of approximately 16000 channels of
fine cathode strips, and we would like to take calibration data
for approximately 30 different voltage levels and on the order
of 300 data points at each voltage. This gives a total of
1.44x10
data points required for ADC gain calibration.
Twice this number of data points will be required for both the
North and South muon arms to be calibrated.
In addition to the ADC gain calibration, the pedestal levels of
all fine cathode channels must be known.
To do this we would like
to take approximately 1000 data points at zero voltage. This
same measurement will allow us to measure the noise levels of
all channels, which is a critical measure of how well the system
is performing. To do the pedestal calibration a total of
1.6x10
data points will be needed.
The data stored for each fine cathode strip will be: 1) the fit parameters for the ADC gain calibration (one parameter if it is very linear, more likely a few parameters) 2) the pedestal values and 3) the sigmas of the pedestal values. This adds up to approximately 80,000 calibration constants for the fine cathode planes.
The calibrations should probably be performed bi-monthly at a minimum since the performance of the chambers depends so strongly on a good fine cathode calibration and low noise levels (which would be checked with a calibration). It would also be useful to track a small fraction of the channels on a short time scale in between calibrations to make sure that calibrations are not drifting significantly. This would be especially useful if the full calibration can only be performed infrequently.