APV6 Multiregion Module Beam Test (June 1998)

The very first silicon detector module with APV6 readout was constructed in early 1998 at HEPHY. Three chips were amplifying the signals of a single silicon detector with 384 strips. The assembled detector module was tested both in laboratory with a source and in a particle beam at CERN. Detailed information about the setup can be obtained at [62].

Fig.  shows this silicon detector module on a support together with an early repeater and temperature probes.

Figure: The first silicon detector module with APV6 readout.

The multiregion detector had twelve domains of 32 strips each with varying geometry specified in tab. . Each zone was surrounded by a guard ring structure, as shown in the detailed view of fig. .

Table: Strip pitch and implant width values of the multiregion silicon detector.

Zone	0	1	2	3	4	5	6	7	8	9	10	11
Strip pitch $\rm [\mu m]$	60	80	240	120	60	80	240	120	60	80	240	120
Implant width $\rm [\mu m]$	25	40	70	50	20	25	50	35	15	15	30	20

Figure: Detail of the multiregion detector showing the guard rings surrounding each zone. The bond wires on top connect the polysilicon resistors to the HV rail.

The beam data were already analyzed and published [62] shortly after the beam test. However, to get consistent results with later beam tests, the raw data were completely re-analyzed with the same software and parameters that were used for the other tests. Thus, the results presented here are slightly different from what was published earlier, but allow direct comparison to the other APV measurements.

The beam test was performed at the X5 beamline of the SPS accelerator at CERN. It delivers highly relativistic pions and muons with a momentum of $100\,\rm GeV/c$, which have a mean ionization of $6\%$ above the MIP level. This excess is approximately compensated by a small jitter of the trigger relative to the particle crossing. The module was placed in a moderately cooled environment at an ambient temperature of $14^{\circ}\,\rm C$, where all measurements were performed in deconvolution mode.

According to eq. , p. , the strip capacitance is a linear function of $w/p$. Moreover, the amplifier noise, as shown in eq. , p. , linearly depends on the capacitance and thus on $w/p$. The SNR implicitly takes over the inverse functional dependence as shown in fig. . The prediction has been obtained from capacitance values calculated by eq. , an APV6 amplifier noise of $\rm 1125\,e+65\,e/pF$ and a MIP charge of $25000\,\rm e$.

Figure: SNR vs. strip geometry of the multiregion detector.

These results were measured at a bias voltage of $150\,\rm V$. Other runs were performed at $100\,\rm V$, resulting in SNR values between $10.6$ and $12.7$, depending on the zone. The SNR decrease of approximately $10\%$ indicates that the multiregion detector was not fully depleted at $100\,\rm V$.

An angle scan between $0^{\circ}$ and $30^{\circ}$ was performed on two different zones. Due to the longer path of particles in the detector, the deposited energy and thus the signal increases. In relation to the collected charge $Q_0$ at perpendicular incidence ($0^{\circ}$), the charge $Q_{\alpha}$ at a tilt angle $\alpha$ is given by

$$\frac{Q_{\alpha}}{Q_0}=\frac{1}{\cos \alpha}\quad.$$ (5.1)

Moreover, the cluster width increases with the tilt angle. In this case, pure geometry does not fit the experimental results, since the average cluster width is slightly above one already at $0^{\circ}$. An empirical approach, which was successfully applied to a large number of angle scans [63], leads to

$${\rm clw}\propto \sqrt{\tan^2 \alpha + {\rm const}^2} \quad,$$ (5.2)

where $\rm clw$ is the average cluster width and $\rm const$ a fit parameter.

Fig.  shows the dependence of the cluster width on the tilt angle for two different zones with the fit function (eq. ) applied. Due to geometrical reasons, the projection of an inclined particle track covers a larger number of strips if their pitch is smaller, explaining the different cluster width curves.

Figure: The cluster width increases with the tilt angle with a slope depending on the strip pitch.

After the beam test, source tests were performed with the same setup, yielding compatible results.