[continues from 4.10.2]

4.10.3 Time (Temperature) dependency

The temperature dependency is fully compensated monitoring the gain of the built-in calibration sources and generating a gain history.

Therefore there are no associated additional calibration files.

The following example shows the typical time dependency of the relative gain. It also shows a temperature HK parameter available to observers (PMT power supply temperature). There is a clear (anti-)correlation with the overall temperature trend (envelope of the temperature profile), i.e. the increase of environment temperature after the first switch on, and the diurnal modulation. There is (fortunately) no correlation with short-term temperature variations due to on-off orbital cycling.

[gain and temperature versus time]

Fig. 4.10.3-I : Example of gain history

The spectra of the calibration sources (referred to detector centre), corresponding to the first and last point of the above M1 gain history, are reported below.

[Fe calibration spectra]

Fig. 4.10.3-II : Example of calibration source spectra, showing typical gain variation

A more precise correlation exists with a different temperature HK parameter, which however is transmitted in the spacecraft (TCU) telemetry, and is not normally available, the DU box temperature (which is not influenced by short term power cycling). SVP observations allowed to verify the existence of a linear trend of gain with temperature.

[gain vs temperature]

Fig. 4.10.3-III : Example of gain-temperature correlation (unit M1), covering almost all Performance Verification phase.
Left frame is for PMT temperature, right frame for DU box temperature. For the PMT case, good temperature values (when instrument is stably on) are in cyan, while gray indicates unstable temperature during switch on.

These are the typical coefficients of the correlation (derived from a "direct" linear fit of gain vs T, and also from a "reverse" fit of T vs gain, indicated in different colours in figure above).

unitgain vs PMT temperaturegain vs DU temperature
M1 g=1.155-0.0104T g=1.196-0.0123T g=1.027-0.00965T g=1.034-0.0104T
M2 g=1.110-0.0079T g=1.186-0.0117T g=1.050-0.00886T g=1.066-0.0104T
M3 g=1.197-0.0066T g=1.224-0.0076T g=1.088-0.00573T g=1.097-0.0063T

4.10.4 Absolute reference

The reference gain to which all gains are normalized is contained in the calibration source characteristics file: the normalization is such that channel 125.5 corresponds to the energy of the 55Fe calibration line (5.894 keV) for all three MECS units, which is only a rough equalization.

4.10.5 The channel boundaries

The photon energy is reported in the telemetry as a raw PHA value (range 0-255). These values are usually corrected during the accumulation for all gain effects and are therefore normalized to a standard scale (sometimes called "PI" or "Pulse Invariant" channels).
The boundaries of the channel in a spectrum have no particular physical meaning and do not enter the convolution procedure, but are used just to plot spectra on an energy scale. They are conventionally produced using the gain relation with an offset O* being the average between O1 and O4.

[Home] [ToC] [Previous] [Next 4.11 cont.]