The MECS consists of three units,
each composed of a grazing incidence Mirror Unit (MU),
and of a position sensitive Gas Scintillation
Proportional Counter (GSPC) located at the focal plane.
The MUs are connected to the GSPCs by a Carbon fiber envelop, about 2 m
long.
MECS overall performance is listed in Table 1 (click here).
The quoted angular resolution values indicate the radii encircling
50% and 80% of the total signal, respectively (,
).
Table 1: MECS overall performance
Each MU is composed of 30 nested coaxial and confocal mirrors.
The mirrors have a double cone geometry to approximate the Wolter I
configuration (Citterio et al. 1985), with diameters ranging from
68 to 162 mm, total length of 300 mm, thickness from 0.2 to 0.4 mm
and focal length of 1850 mm.
The MU design was optimized to have the best response at 6 keV.
A replica technique by nickel electroforming from super-polished
mandrels was used to build up the mirrors (Citterio et al. 1988).
A 1000 Å thick gold
layer provides the X-ray reflecting surface.
The 30 mirrors are nested using two front-end spiders with
eight arms.
The geometrical collecting area of each MU is 123.9 cm;
the spiders and an active anti-ions grid reduce this geometrical area
by about 18%.
The measured radius for 50%, 80% and 90% encircled energy
are of the order of 40, 110, and 210 arcsec at 8 keV, respectively.
A more detailed description of the MU Point Spread Function can be found
in Conti et al. (1993, 1994).
The focal plane detectors are Xenon filled GSPC, working in the range
1.3-10 keV with an energy resolution of 8% at 5.9 keV and a
position resolution of
mm (corresponding to 1 arcmin,
approximately) at the same energy.
The gas cell is composed by a cylindrical ceramic body (96 mm
internal diameter) closed, at the top, by a 50
m thick entrance
beryllium window with 30 mm diameter and, on the bottom, by an UV exit
window made of Suprasil quartz with 80 mm diameter and 5 mm thickness, as
schematically shown in Fig. 1 (click here).
In flight, the getter can be activated to purify the gas, if necessary.
Figure 1: Schematic view of the MECS instrument: gas cell and position
sensitive GSPC
The entrance window is externally supported by a beryllium strongback structure, 0.55 mm thick, consisting of a ring (10 mm inner diameter, 1 mm width) connected to the window border by four ribs, as shown in Fig. 2 (click here).
Figure 2: Geometry of the strongback
An X-ray photon absorbed in the gas cell liberates a cloud of electrons.
A uniform electrical field across the cell drifts the cloud up to the
scintillation region, with an higher electric field, where UV light is
produced through the interaction of the accelerated electrons with the
Xe ions.
The amplitude of the UV signal, detected by a PMT, is proportional to
the energy of the incident X-ray. The duration of the signal, the
so-called Burst Length (BL), depends on the interaction point and it
is used to discriminate genuine X-rays
against induced background events.
BL rejection may be carried out on board and/or on-ground. The BL
rejection mechanism on board is based on a programmable BL acceptance
window (not energy dependent).
Two grids inside the cell separate the absorption/drift region (20 mm
depth) from the scintillation region (17.5 mm depth).
The UV readout system consists of a crossed-wire anode position sensitive
Hamamatsu photomultiplier (PMT) with quantum efficiency of 20%.
The high voltage nominal values are -8 kV for the Be window,
for the scintillation grid, 1000, 992, and 943 V for the PMT of ME1, ME2, and
ME3 units, respectively.
Two collimated calibration sources (nuclear line at 5.95 keV),
with an emission rate of
count per second, are located,
diametrically opposed, near the edge of the Be window.
These inner calibration sources, continuosly visible at the edge of the
Field of View (FOV), allow the monitoring of the detector
gain. Furthermore a passive ion shield is placed in front of the detector.
The focal plane detector characteristics are shown in
Table 2 (click here).
The MECS electronic processing takes place, almost entirely, in the Electronic Unit; only signal buffering is performed near the PMTs (one for each detector).
From each PMT, six signals are transferred to the Electronic Unit:
Trigger, Energy, ,
,
, and
.
All the signals are converted from current to voltage before
to be passed to the Electronic Unit; in particular, the Energy and
the position signals are integrated with a little time constant (few
hundred of nanoseconds).
The Trigger signal is taken from the
dynode whereas the Energy
is taken from the
dynode (the last one).
The
and
signals come out from the 16 anode wires
(connected through a resistor divider) that give information about the
X position. The same is valid for
and
.
In the Electronic Unit there are three analog processing modules (one for each detector) followed by one Event Processor that arranges the data in packets sending them to the Communication Processor.
All the signals incoming from a single PMT (Trigger excepted) undergo
to a baseline restoring and then they are integrated by Gated Integrators
with fixed integration time.
The X and Y positions are calculated (via hardware) from
the original
,
,
and
in accordance to the formulas:
The Burst Length BL signal is obtained from the Energy through a constant fraction, zero crossing and Time-to-Amplitude Converter chain.
There is a programmable selection logic, working on Energy and BL signals, to reject events before the A/D conversion. If during the integration time two trigger pulses are recognised, then the Pile-up Logic will reset the electronics in order to be ready for the next event. An Event Qualification Logic increments the Ratemeter registers, that are read at 1 s rate.
After the A/D conversion the data are saved into a FIFO register together with the event time. Data from the FIFO are read by the Event Processor and submitted to a rejection rule with programmable windows for Energy, BL, X and Y. Finally data are packed and written in a Dual Port RAM from where they will be read by the Communication Processor.