Detector Performance On Ground and In Orbit

For the most recent devices, the best value for the readout (or electronic) noise of the on-chip electronics is 2 e- rms at -60° C, whereas typical values for the previous devices scatter around 5 e- rms, e.g., for the XMM-Newton pnCCD camera system. This includes all noise contributions described in (4). The charge transfer properties of the pnCCDs on XMM-Newton are reasonably good, in the order of a several per cent signal loss from the last to the first pixel over a distance of 3 cm charge transfer. As the charge transfer losses describe the position-dependent energy resolution, it is one of the key parameters for the spectroscopic performance, especially after radiation damage may have occurred. Figure 7.12 shows an55Fe spectrum of a pnCCD in a flat field measurement resulting in a typical energy resolution of 125 eV at an operating temperature of -120° C [17]. The XMM-Newton flight camera was operated at -90° C during calibration on ground with a resolution of about 145 eV (FWHM) over the entire area of 36 cm2. The degradation of energy resolution was mainly caused by the reduction of the charge transfer efficiency. Leakage currents and on-chip JFET properties only played a minor role. The impact of the material properties of silicon and related impurities and their consequences for the operation of scientific grade X-ray pnCCDs including the effects of radiation damage are treated in detail in the references [7,10]. The radiation damage accumulated over the expected life time of XMM-Newton was estimated to be equivalent to a 10 MeV proton fluence of 4-5 x 108 p cm-2. The Figs. 7.9 and 7.10 show the results of the irradiation tests with 10MeV protons: the expected decrease of energy resolution over the 10-year dose is from 146 to 164 eV FWHM at an operating temperature of -100° C. At the actual operating temperature of -90° C, the expected effect of trapping and detrapping at A-centers [19], generated by the radiation, is even more reduced.

In a single photon counting mode, the quantum efficiency was measured with respect to a calibrated solid state detector. Figure 7.11 shows measurements from the synchrotron radiation facilities in Berlin and Orsay. At 525 eV a 5% dip can be seen because of the absorption at the oxygen edge in the SiO2 layers. The same happens at the Si-K edge at 1840 eV showing the fine structure of a typical XAFS spectrum (see insert of Fig. 7.11). For all energies, the quantum efficiency is nicely represented by a model using the photo absorption coefficients from the atomic data tables. The quantum efficiency on the low energy side can be further improved with respect to the measurements shown in Fig. 7.11, by increasing the drift field at the p+n - junction entrance window [4] and by using (100) silicon instead of (111)

Energy [eV]

Fig. 7.9 Energy spectrum from an 55Fe source after different 10 MeV proton fluences of 0 p cm-2 (dotted line), 4.1 x 108p cm-2 (solid line), 6.1 x 108p cm-2 (dashed line), measured at the low (and after irradiation unfavorable) temperature of 142 K. The expected dose over a life time of 10 years is equivalent to approximately 5 x 108 10-MeV p cm 2 for the pnCCD on XMM-Newton [22]

Energy [eV]

Fig. 7.9 Energy spectrum from an 55Fe source after different 10 MeV proton fluences of 0 p cm-2 (dotted line), 4.1 x 108p cm-2 (solid line), 6.1 x 108p cm-2 (dashed line), measured at the low (and after irradiation unfavorable) temperature of 142 K. The expected dose over a life time of 10 years is equivalent to approximately 5 x 108 10-MeV p cm 2 for the pnCCD on XMM-Newton [22]

silicon. The useful dynamic range of the pnCCD camera on XMM-Newton was adjusted to the energy band from 100 eV to 15 keV (see Fig. 7.11).

Split events, i.e., events with electrons spread over more than one pixel, originating from one single photon, were reconstructed and summed to one photon event. In total, about 70% of all events are single pixel events, 28% are two pixel events, and 2% are events with three and four pixels involved. In the case of the XMM-Newton pnCCDs, a single X-ray photon spreads the generated signal charge never over more than four pixels.

The readout electronics of the pnCCD system is described in the references [2,20]. A charge sensing amplifier followed by a fourfold double-correlated sampling stage, multiplexer, and output amplifier (CAMEX64B JFET/CMOS ASIC chip) guide the pnCCD pixel content as a voltage signal to a 10 MHz 12 bit flash ADC system. The whole system, i.e., CCD and CAMEX64B amplifier array

20 0

20 0

2.0x1 0"

Proton Fluence [p/cm!] Fig. 7.10 FWHM of the Mn-Ka-spectrum (5 894 eV) in dependence on proton fluence and temperature. Before proton exposure, the lower operating temperature of 140 K gains better results. After a 10MeV proton fluence of more than 2 x 108/cm2 the higher temperature of 174 K results in a better enrgy resolution. The FWHM is degraded from 135 eV (140 K) to 160eV and 175 eV (174 K) after 4.1 x 108 p/cm2 and 1.9 x 109 p/cm2, respectively. A FWHM of 164eV is expected after the 10-year XMM-Newton mission [22]

Fig. 7.11 Quantum efficiency of the pnCCD as a function of the incident photon energy. The energy scale ranges from 0.15 to 30 keV. The solid line represents a 300 |m thick sensitive volume, the dotted line 500 |m [22]

10.0

10.0

1100 1200 1300 1400 1500 1600 1700 1800

energy [adc counts]

Fig. 7.12 Mn-Ka and Mn-Kp spectrum of an55Fe source. The measured FWHM is 125 eV at -120° C. Note the log scale

1100 1200 1300 1400 1500 1600 1700 1800

energy [adc counts]

Fig. 7.12 Mn-Ka and Mn-Kp spectrum of an55Fe source. The measured FWHM is 125 eV at -120° C. Note the log scale dissipates a power of 0.7 W for the entire camera (768 readout channels), a value that is acceptable in terms of thermal budget on XMM-Newton realized through passive cooling. A further increase of the readout speed can be made only at the expense of further increase of power, or a degradation of the noise performance.

The charge handling capacity of the individual pixels was tested with 5.5 MeV alpha particles from a radioactive 241Am source. Around 106 electrons can be properly transferred in every pixel [22]. The spatial resolution was intensively tested in the PANTER facility with the flight mirror module in front of the focal plane. The first light image of the Large Magellanic Cloud in orbit (see Fig. 7.13), as well as the quantitative analysis of the point spread function have shown a perfect alignment of the telescope system as on ground.

Until now, more than seven years after launch, no instrumental surprise occurred: The energy resolution is almost equal to the ground measurements as is the case for the charge transfer efficiency [23]. To date, the electrical stability of the instrument

EPIC pn

offline analysis

Fig. 7.13 The Large Magellanic Cloud in X-ray colors. The figure shows the first Light image of the pnCCD camera in orbit. The field of view of 30arcmin corresponds approximately to the angular size of the moon. The image shows the area of 30 Doradus, a supernova remnant as an extended source of X-rays. The "north-west" of 30 Dor shows an emission of X-rays up to 5 keV (blue), while the "south-east" rim appears much softer in X-rays (yellow and red). The supernova 1987A is the bright source "south west" of 30 Doradus. About 40 new X-ray objects have been found in this exposure [22]

is perfect. The first light image in Fig. 7.13 qualitatively summarizes the above enthusiastic statements.

0 0

Post a comment