Results on Hal 25 + ALABUF testing.
23.10.2002.
Vladimir Gromov, Alexei Sokolov.,
Utrecht University, Utrecht, the Netherlands.
Goals of the testing.
1. Characterization of the analog signal processing chain as a whole i.e. measurements of the dynamic range, linearity and equivalent noise.
2. Compatibility of the front-end chip (HAL25) and the analog buffer (ALABUF) chip.
Fig.1 Differential output signals corresponding to 1MIP (22000e).
a). Positive input signal. Dynamic range and linearity.
Fig.2 Differential output signal as a function of charge, injected into the input of HAL25.
Fig.3 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.4 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
b). Negative input signal. Dynamic range and linearity.
Fig.5 Differential output signal as a function of charge, injected into the input of HAL25.
Fig.6 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.7 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
c). Noise at the outputs.
Fig.8 Noise at the outputs.
Conclusions.
1. Dynamic range of the complete analog signal processing chain is +1.6V…-1.34V (±15 MIPs) (see Fig.2, Fig.5). Full dynamic range of the ALABUF chip itself is ±1.85V (see “Report on analogue chip (ALABUF) development in 0.25u CMOS technology for the ALICE Silicon Strip Detector (SSD)”). It let us conclude that ALABUF chip does not restrict dynamic range.
2. Linearity is not easy to estimate within specified dynamic range due to considerable spread between different channels (30%) (see Fig.3, Fig.6). When one linear fit is applied to all the channels, nonlinearity goes to ±20% within 4MIP range (see Fig.4, Fig.7). Such a nonlinearity certainly is not caused by ALABUF chip and most probably comes from HAL25 performance.
3. The observed noise with standard deviation of s=11.4mV (2100 electrons) (see Fig.8) consists of both pure electronic noise and pick-up noise mostly (HAL25 pure electronic noise is 400 electrons). It seems like special precautions must be carried out to get rid of the pick-ups
d). Positive input signal. Dedicated channel calibration.
Channel AIN57. Fitting function(Qin)=0.0045*Qin+30mV
Fig.9 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.10 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
e). Positive input signal. Dedicated channel calibration.
Channel AIN54. Fitting function(Qin)=0.00555*Qin-20mV
Fig.11 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.12 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
f). Positive input signal. Dedicated channel calibration.
Channel AIN55. Fitting function(Qin)=0.0055*Qin-27mV
Fig.13 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.14 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
g). Negative input signal. Dedicated channel calibration.
Channel AIN57. Fitting function(Qin)=-0.0054*Qin-45mV
Fig.15 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.16 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
h). Negative input signal. Dedicated channel calibration.
Channel AIN54. Fitting function(Qin)=-0.006*Qin-20mV
Fig.17 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.18 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
i). Negative input signal. Dedicated channel calibration.
Channel AIN55. Fitting function(Qin)=-0.0054*Qin-32mV
Fig.19 Differential output signal as a function of charge, injected into the input of HAL25 (fine scale).
Fig.20 Nonlinearity as a function of charge, injected into the input of HAL25 (fine scale).
Conclusions #2.
1. Linearity gets better (±5% within 4MIP range) if each channel is individually calibrated (assigned with a slope factor and an offset).