Phenix Note 125-Prototype Measurements

NEXT:Results UP:Cathode Strip Readout

The CSRC prototype chambers used in the test were mechanically identical to the chambers that were designed for the Neutral Meson Spectrometer (NMS) but had the anode and cathodes modified for PHENlX applications. These prototype chambers have an active area of 38 cm x 73.7 cm with overall dimensions of 50.8 cm x 86.4 cm. The anode to cathode gap was 3.18 mm and the cathode foil was 6.35 micron kapton with 100 Angstroms of gold. The cathode strips were parallel to the long dimension.

3.2 Amplifiers

The signals induced on the cathode strips must be amplified and digitized with a signal-to-noise of 100/1 in order to obtained a resolution which is 1% of the cathode strip spacing. For the prototype tests we have used an amplifier which was developed for the NMS at LAMPF. It is based on an in expensive commercial transimpedence amplifier, the NE5212. The single ended input to each channel is protected by a 100

resistor and back to back diodes. The differential outputs were capacitively coupled and transported through 100 feet of shielded twist and flat cable where they are transformer coupled to the single ended inputs of LeCroy FERA ADC's. This provides a relatively noise immune method of transporting and delaying the analog outputs from the amplifiers.

Tests of the linearity and noise levels in this system were performed to determine whether it was suitable for the CSRC tests. Shown in Figure 3.2.1 are the results of tests of the amplifier linearity made using a test input on the amplifier and a tail pulser to simulate chamber inputs. Over the dynamic range tested the amplifier was found to be linear to within 0.1% This is better than needed in the current application.

Noise levels were measured using the same system by measuring the RMS of the digitized peak. The noise level was found to be independent of input pulse height, but as expected, highly dependent on input capacitance. This is shown in Figure 3.2.2. The expected input capacitance for a l cm x 300 cm strip is approximately 330 pf and approaches 500 pf if a chevron configuration is used. Under all configurations the signal-to-noise of this amplifier is less than 100/1 and could be used for the production electronics.

Because these amplifiers were readily available, they were used in all subsequent tests along with the FERA ADC's.

Figure 3.2.1: Linearity test of NE5212

Figure 3.2.2: Noise level vs. capacitance

3.3 Anode Structure

The anode structure used for the NMS chambers was an alternating anode - cathode design with 5.56 mm anode wire spacing. The anode wires are 20

m gold plated tungsten and the cathode wires are 75

m gold plated copper clad aluminum. The anode and cathode wires were wound with 50 g of tension. They are attached to the chamber planes with epoxy and soldered to the art work. This structure gives a simple chamber that has a wide efficiency plateau. This is illustrated in Figure 3.3.1, where the gain and efficiency are plotted as a function of voltage. The wide plateau insures that the chamber is insensitive to small fluctuations in gas composition and minor fluctuations in high voltage as well as variations in the front-end amplifier characteristics such as gain and trigger level.

We have compared this alternating anode-cathode design with an MWPC design, i.e. an anode plane with no cathode field wires. The ratio of the induced cathode pulse to the anode pulse for both designs was measured to determine if there was any significant loss of the induced signal due to capacitive coupling to the field wires. We found that the induced cathode pulse heights were within 5% for the two designs.

Because the anode wires in the alternating anode-cathode design can be disabled without effecting the other anode wires in the plane, all further tests used the alternating design.

Figure 3.3.1: Efficiency vs. High Voltage

3.4 Gain Calibration

The simplest method of computing position is to calculate the center of gravity of the induce charge distribution. The center of gravity is,

(1)

where is the induced charge and is the location of each strip. From this equation it can be shown that the relative accuracy of the charge measurement must be better than 1% for 100 m resolution with 1 cm amplifier spacing, the goal of the present work.

Reaching the goal of better than 1% in the charge measurement means amplifier gains requiring separate gain calibration for each channel and then correcting the measured charges in the software. Since the variations in the amplifier gains can be as much as +/-10%, this type of calibration procedure is necessary. The procedure developed for doing this was to histogram the ratio of the induced cathode charge, , to the anode charge, for the strip with the largest induced pulse. The ratio of the maximum induced strip charge to the anode charge is,

(2)

The maximum occurs when the avalanche is exactly in the middle of the cathode strip. Since the chamber construction is very accurate, all strips should be measuring the same maximum induced charge except for amplifier gain variations. The gains were software adjusted by finding the value of above which 25% of the area of the histogram was contained. This procedure gave a gain calibration with an error of less than 0.5% after a few hours of accumulated cosmic ray data.

3.5 Resolution Measurements

Although equation (l) provides a simple method for computing positions, it gives significant differential nonlinearities, which dominate the resolution. The problem can be solved by using the shape of the induced pulse distribution to map out the differential nonlinearities. That can be done by using the Mathieson distribution or by measuring it. We have decided to measure it. We have used a histogram of dN/df where

, the sum is over all cathode strips, along with the knowledge that the cosmic ray flux,

, is uniform as a function of x. This gives,

We can obtain the induced pulse distribution, , by a simple integration,

The results of this procedure is shown in Figure 3.5.1. All subsequent data use this distribution.

Figure 3.5.1: Measured induced charge distribution

A method for calculating positions is to use the table of and to calculate a weighted sum,

where is given by equation (1), w are a set of weights, and the sum is restricted to only three pads.

Position resolutions have been measured with a number of possible cathode configurations using a setup with three planes of CSRC. Positions from the two outer planes were used to project to the center plane, and the residual was,

For the case where all three chambers are assumed to contribute equally to the resolution,

and are equally spaced, the contribution of each chamber, , is given by

3.6 Cathode Configurations

The design of the cathodes can be in strips or in a chevron configuration. Both have been tested. If the strip width is too large compared to the anode-cathode gap, significant nonlinearities can occur. These nonlinearities can be reduced if a chevron configuration is used in place of the strips. An example of a chevron configuration is shown in Figure 3.6.1. In this example all chevron strips are instrumented with an amplifier. An alternative to instrumenting all strips or chevrons is to use the intermediate cathode developed at BNL. This scheme uses capacitive coupling and the readout occurs every other strip or chevron or every third chevron as illustrated in Figure 3.6.2. We have investigated position resolutions for the capacitive coupling scheme (sometimes called the floating cathode) on chevrons.

Figure 3.6.1: Example of a Chevron configuration[5]

Figure 3.6.2: Examples of capactive coupling or floating cathode techniques. (a) Single intermediate strip cathode, and (b) two intermediate strip cathodes.